ProgrammableMarketing/quintagroup.com (opens in a new tab)
PDF

Views

  • Plan
  • Roadmap
  • Research

On this page

  • The brief
  • Start here
  • Idea portfolio
  • Effort vs impact
  • This week · 8
  • This quarter · 11
  • Big bets · 3
  • Wildcards · 1
  • How we can help
Generated Nov 16, 2026
  1. Programmable
  2. Playbooks
  3. Quintagroup
Quintagroup logo

growth playbook · Nov 16, 2026

quintagroup.com (opens in a new tab)

Quintagroup growth playbook

Diagnosis

Quintagroup has unusual proof for a mid-sized development company: the open-source toolkit behind ProZorro, auctions that sold more than USD 700 million of assets, ISO 27001:2022, and a working SaaS auction product. In 2026 it added a set of well-written service pages for agentic AI, self-hosted AI, answer engine optimization and an Auction Delivery Framework, each promising a fixed-price quote within five business days.

The site around those pages has not caught up. Every promise ends in a generic four-field contact form that answers "as soon as possible"; the homepage still sells Python, JS and DevOps and never mentions AI; 36 case studies carry no results or dates; the newest Clutch review is from 2021; the blog went quiet between 2023 and 2026; and the experience figures and the AI framework differ from page to page.

This playbook starts with an estimate desk that makes the five-day promise real and becomes the first AI case study, then cleans up visitor identification, proof, the auction demo and the conflicting numbers, and builds toward procurement-specific AI offers and a public procurement data report.

Growth Score

6/10

Biggest Opportunity

Use Quintagroup's own agentic AI and auction products on its own funnel: a structured intake and drafting desk behind the five-business-day quote, and a working Atreus sandbox behind every auction demo request, so the proof and the sales process are the same thing.

Biggest Risk

Larger Python houses already sell AI with measured results and a hundred recent reviews, while Quintagroup's AI pages cite no case study and its public proof is dated; buyers who compare may never send the contact form.

01

Start here

The three moves to make first. Each one is written up in full below.

  1. 1

    Put an estimate desk behind the five-business-day promise: service-specific intake, a drafted brief, an analyst's approval

    List the last 20 enquiries for the AI, AEO and auction services and count how many needed a follow-up question before an estimate could start.

    High impact First result in 3 weeks
  2. 2

    Answer every auction demo request with a working Atreus sandbox of the auction type the buyer chose

    Set up one reverse English sample auction on Atreus and send it to the next demo request by hand.

    High impact First result in 2 weeks
  3. 3

    Keep one visitor-identification tool instead of three, and turn its daily list into researched account briefs

    Export last month's identified companies from all three tools and compare the overlap and which one sales actually opened.

    High impact First result in 2 weeks

02

Your idea portfolio

23 ideas, grouped by when to run them. Open any idea for how to run it: the steps, who to approach, copy you can use, and how to measure it.

  • 7 ideas are marked Unexpected: plays you are unlikely to have tried.

    1 of these are things Programmable can build and run for you.

Every idea placed by the effort it takes against the impact it could have. Quick wins sit top left; select a number to open that idea. The filters below narrow the list by status, timing, category and what Programmable can build.

Effort vs impactEach number is an idea. Select one to open it.
Impact →
Quick wins · 1Big bets · 6Fill-ins · 11Reconsider · 5
Effort →

Showing 23 of 23 ideas

This week 8

Quick wins with a clear first action.

  • 01

    Quintagroup's newest pages each promise a fixed-price quote within five business days, and the agentic AI page says that quote needs the buyer's target workflow and current tech stack; the AEO page asks for a URL list and target queries. But every "Contact Quintagroup" button lands on the general contact form: name, e-mail, phone and one comment box, no file upload, and a reply "as soon as possible". The only form on the site that asks real qualifying questions is the auction demo form. So the clock on a five-day promise starts with an e-mail asking for the information the page already said was required.

    High impactMedium effort First result in 3 weeks Business Analyst

    First action

    List the last 20 enquiries for the AI, AEO and auction services and count how many needed a follow-up question before an estimate could start.

    How to measure it

    Proposals inside the five-business-day promise

    Now
    –
    Target
    90% of qualified enquiries
    By
    90 days

    Steps

    1. 1

      Measure today's cycle

      For the last 20 enquiries, record time to first reply, number of clarifying e-mails and days to proposal. This is the baseline the desk is judged against.

      Business Analyst 2 days

    2. 2

      Write the four intakes

      Five to seven questions per service, taken from what each page says an estimate needs: workflow and tech stack, URL list and queries, team size and hosting, auction type and purpose. Add a file upload to each.

      Business Analyst 3 days

    3. 3

      Build the forms and point the buttons at them

      Create the forms in the existing Plone form tool, keep the reCAPTCHA, and change each service page's "Contact Quintagroup" button to its own intake. Submissions go to the sales mailbox and to a structured record.

      Senior Python Developer 3 days

    4. 4

      Build the drafting agent

      Inputs: the form answers, the uploaded file, the enquirer's public website and the case library. Outputs: a one-page discovery brief, open questions, the two closest cases and a proposal skeleton in the service's own framework. It writes drafts into the team's document space and never e-mails a buyer.

      Senior Python Developer 1 week

    5. 5

      Set the approval step

      A Business Analyst reviews, corrects and approves each draft; scope, price and timeline are always set by a person. Every edit is logged so the agent's weak spots are visible.

      Business Analyst 2 days

    6. 6

      Add the clock

      A simple board shows each enquiry's age in business days, who holds it, and an alert on day three if no proposal is in review.

      Project Manager 2 days

    7. 7

      Review after 30 days

      Compare first-reply time, clarifying e-mails and days to proposal with the baseline; check how much of each draft the analyst kept.

      Head of Sales monthly

    8. 8

      Write it up as a case study

      Publish the workflow, the approval points and the measured before and after as the first agentic AI case on the site.

      Digital Marketing Manager 1 week

    First result
    3 weeks
    Cost
    Two to three weeks of internal development and analyst time; model usage costs are small at this volume
    Owner
    Business Analyst

    Smallest test

    Add the workflow and tech-stack questions to one page's form (agentic AI development) and draft the next five briefs by hand in the proposed format.

    Why it works

    Give each promised service its own short intake (agentic AI: the workflow, the systems it touches, volumes; AEO: URLs and queries; self-hosted AI: team size and hosting; auctions: type and purpose, as the demo form already does). Behind the forms, build an estimate desk: an agent that reads the submission and any uploaded brief, researches the company from its public site, finds the two closest Quintagroup cases, and drafts a discovery brief, open questions and a proposal skeleton in the framework the page describes. A Business Analyst edits and approves every draft before anything is sent. The desk tracks hours from submission to first reply and days to proposal against the published promise. It is also Quintagroup's own agentic AI running on a real workflow, which the AI pages currently lack as proof.

    Buyers who are promised five days notice when day one is spent on a clarifying e-mail. Collecting the required inputs up front shortens the cycle, and a drafted brief lets a small pre-sales team answer more enquiries at the same quality. Running it in-house gives the sales team a live example to show.

    Ready to use

    ChecklistIntake questions for the agentic AI form

    [ ] Which workflow should the agent handle, in one or two sentences?
    [ ] Which systems does it touch today (CRM, ERP, ticketing, internal APIs)?
    [ ] Roughly how many items a week go through it?
    [ ] Who approves the outcome today?
    [ ] What must never happen without a person's sign-off?
    [ ] Where must data stay (own servers, a specific cloud region)?
    [ ] Upload a process description or sample documents (optional)

    Risks and how to handle them

    • Longer forms reduce the number of enquiries

      Keep only name and e-mail required, make the rest optional, and leave the short general form available.

    • A draft contains a wrong assumption about the buyer

      Nothing leaves without analyst approval, and the brief marks which statements came from the form and which from public research.

    • Buyers upload confidential documents

      Process files inside the company's own environment under the ISO 27001 controls and state this next to the upload field.

    Evidence

    • Your websiteNew service pages promise a quote in five business days: agentic AI development offers a "fixed-cost proposal within 5 business days" and its FAQ says a fixed-cost estimate needs the target workflow and current tech stack; the AEO page asks for "your URL list and target query set for a fixed-price estimate within 5 business days"; the self-hosted AI and Auction Delivery Framework pages make the same promise. Each "Contact Quintagroup" button links to the general /contact form.Source
    • Your websiteContact form: Name, E-Mail, Phone, Comment ("Tell us what you are looking for..."), Country Code and a reCAPTCHA; no service, workflow or tech-stack question, no file upload and no calendar booking; the reply is promised "as soon as possible". The Estimate project form adds one file upload and a "Project details" box, with the same "as soon as possible". Two US phone numbers and info@ are the other routes; no live chat is shown.Source
    • Your websiteRequest an Auction demo form: asks the auction type (English, Reverse, Multivariable, Hybrid Dutch, Dutch, Sealed Bid, Spectrum, SMR, Clock, Combinatorial, Other) and the purpose (Selling, Buying, Both); the page offers no live demo link, sandbox or video, and says the team will be in touch "as soon as possible".Source
    • HiringBusiness Analyst vacancy: "Gathering and analyzing data in support of business cases, proposed projects, and system requirements" and "Managing reports". Project manager vacancy: "Monitoring progress and managing reports" and "Being responsible for resources and time planning". Both list public procurement experience as a plus.Source
  • 02

    The homepage loads three tools that do the same job, naming the companies behind anonymous visits: Leadfeeder, Happierleads and Salespanel. Three overlapping lists usually mean nobody owns any of them. The value is not in the list; it is in what happens the morning after a procurement agency or an auction house reads the sealed-bid page.

    High impactMedium effort First result in 2 weeks Head of Sales

    First action

    Export last month's identified companies from all three tools and compare the overlap and which one sales actually opened.

    How to measure it

    Conversations from identified accounts

    Now
    –
    Target
    8 a month
    By
    90 days

    Steps

    1. 1

      Compare the three tools

      Overlap, match quality, cost and who has logged in during the last month.

      Digital Marketing Manager 2 days

    2. 2

      Keep one, remove two

      Remove the unused tags from Google Tag Manager and cancel the subscriptions; record the owner of the remaining tool.

      Digital Marketing Manager 1 day

    3. 3

      Define a good account

      Sectors, size and page patterns worth a salesperson's time (for example two visits to auction or e-procurement pages); an exclusion list for vendors, recruiters and universities.

      Head of Sales 1 day

    4. 4

      Build the daily brief

      Inputs: the tool's export or API, pages viewed, the company's public website. Output: a five-line brief with what the company does, why the pages they read matter, the closest case study and a suggested contact role. Delivered to the sales mailbox and the CRM.

      Senior Python Developer 1 week

    5. 5

      Draft, never send

      The automation writes a first e-mail draft per account; a salesperson approves, edits or discards it. Nothing is sent automatically.

      Head of Sales 2 days

    6. 6

      Respect privacy rules

      Company-level data only, an updated privacy notice, and no contact for people who have opted out.

      Head of Sales 1 day

    7. 7

      Measure weekly

      Briefs produced, briefs acted on, replies and meetings; drop filters that produce no meetings.

      Head of Sales weekly

    First result
    2 weeks
    Cost
    One to two weeks of development; two subscriptions saved
    Owner
    Head of Sales

    Smallest test

    For one week, hand-write briefs for the five best identified accounts each day and see how many the sales team acts on.

    Why it works

    Pick one tool, remove the other two, and name an owner. Then build a small automation: each morning it takes the identified companies and the pages they read, filters out students, vendors and job seekers, researches each remaining company from public sources, matches it to the closest service and case study, and drafts a short brief and a first e-mail for a salesperson. A person decides whether to send anything and edits every message.

    Intent data loses value within days. A single owned list with research attached gets used, and removing two scripts also lightens every page.

    Risks and how to handle them

    • Visitor matching is often wrong for small companies and home networks

      Only act on accounts with two or more visits and a plausible sector.

    • Outreach that reveals tracking feels intrusive

      Lead with a relevant case study, never with "we saw you on our site".

    Evidence

    • Site researchThird-party scripts on the homepage (11 services, about 2.2 MB): three visitor-identification tools at once (Leadfeeder, Happierleads and Salespanel), Google Tag Manager, Google Analytics, a Facebook script and a request to Tawk.to although no chat is shown. The site runs on Plone with Google Workspace mail and an RSS feed.Source
    • HiringDigital Marketing Manager vacancy: "Planning digital marketing campaigns, including web, SEO/SEM, social media, and display advertising"; analysing market trends, sales and performance indicators; preparing presentations and brochures; optimising English articles for SEO around given key phrases; running social media; "Solid knowledge of website analytics tools"; one year of experience asked.Source
    • Your websiteContact form: Name, E-Mail, Phone, Comment ("Tell us what you are looking for..."), Country Code and a reCAPTCHA; no service, workflow or tech-stack question, no file upload and no calendar booking; the reply is promised "as soon as possible". The Estimate project form adds one file upload and a "Project details" box, with the same "as soon as possible". Two US phone numbers and info@ are the other routes; no live chat is shown.Source
  • 03

    Quintagroup's strongest numbers sit on the About page: ProZorro.Sale enabled sales worth more than USD 700 million and ProZorro helped save USD 2 billion. The case studies themselves say what was built and with which technology, but the two read in full give no result, date, team size, duration or client quote. A competitor of the same kind leads with "14,000+ staff hours saved every year".

    High impactMedium effort First result in 1 month Digital Marketing Manager

    First action

    Pick the six cases and ask each delivery lead for the year, duration, team size and one measured outcome.

    How to measure it

    Case-study pages with a result figure

    Now
    –
    Target
    12 of 36
    By
    60 days

    Steps

    1. 1

      Choose six cases

      Those closest to auctions, procurement, API and AI work.

      Head of Sales 1 day

    2. 2

      Collect the facts

      Year, duration, team size and outcome figures from delivery leads and public sources.

      Project Manager 1 week

    3. 3

      Ask for quotes

      One or two sentences from each client contact, with approval in writing.

      Head of Sales 2 weeks

    4. 4

      Add the header block

      A reusable Plone template with the same fields on every case.

      Senior Python Developer 2 days

    5. 5

      Move the big numbers

      Put the USD 700 million and USD 2 billion figures at the top of their own case pages and on the homepage.

      Digital Marketing Manager 1 day

    6. 6

      Add filters

      Industry and technology tags on the case list.

      Senior Python Developer 3 days

    First result
    1 month
    Cost
    Staff time only
    Owner
    Digital Marketing Manager

    Smallest test

    Rewrite one case (Prozorro.sale) with year, scale and outcome, and watch time on page and enquiries for a month.

    Build the prospect list. The list compiled from public sources, verified, deduplicated and delivered to your CRM. Usually live within days.

    Book a working session

    Why it works

    Add a standard header to every case study: year, duration, team size, the client's sector, and one to three outcome figures, with a named or role-level quote where the client agrees. Start with the six cases most relevant to current selling (ProZorro, Prozorro.sale, Atreus, the Mastercard API project, the insurance image-recognition service and the tender data project). Add filters by industry and technology to the list of 36.

    Buyers of custom software read case studies for evidence that a similar problem was solved and what changed. Dated, measured cases are also what AI answer engines quote.

    Risks and how to handle them

    • Clients will not approve figures

      Use ranges, publicly reported numbers or anonymised sector descriptions.

    Evidence

    • Your websiteCases: 36 case studies on three pages of 12, with no filter by industry or technology. The two case pages read in full (Mastercard API enhancement and the Atreus auction platform) give no results, dates, team size, duration or client quote. The list runs from recent work (an insurance image-recognition service, "The Atlas Project" that collects auction and tender data with Python, Selenium and AWS) back to Plone-era websites and a Plone themes web store.Source
    • Your websiteAbout us: "on the international market since 2003"; ISO 27001:2022 certified; ProZorro.Sale enabled sales worth more than USD 700 million and ProZorro helped save USD 2 billion through reverse auctions; auctions developed for "over 5 years". Other pages say differently: the auctions page claims "9 years of experience", the services index "over 16 years of experience", and the footer reads "© 2002-2026". No team size, office or leadership is shown.Source
    • Competitor researchRivals' own homepages: STX Next, "We turn isolated AI experiments into resilient operational systems for the Industrial and Financial sectors", 450+ specialists, five stars across 101 Clutch reviews and results such as "14,000+ staff hours saved every year"; Django Stars, "AI-Powered Software Development & Consulting Company", 100+ developers and a 92.7 NPS; Auction Software, "OVER 500 AUCTION SITES CREATED", "Schedule a Demo", no pricing or free trial shown.Source
  • 05

    The auction demo form already asks the two questions that matter, auction type and whether the buyer is selling or buying, and then says someone will be in touch "as soon as possible". Meanwhile Quintagroup runs Atreus, a live auction platform with a free tier of 10 auctions a month, and its own framework page promises a demo environment within one to two weeks of kickoff. The demo exists today; the form does not hand it over.

    High impactLow effort First result in 2 weeks Head of Sales

    First action

    Set up one reverse English sample auction on Atreus and send it to the next demo request by hand.

    How to measure it

    Demo request to scoping call rate

    Now
    –
    Target
    50%
    By
    90 days

    Steps

    1. 1

      Build four sample auctions

      One per Atreus type with realistic lots, rules and two bidder accounts.

      Atreus Product Owner 3 days

    2. 2

      Record walkthroughs

      Three minutes per type, organiser and bidder view.

      Atreus Product Owner 2 days

    3. 3

      Write the replies

      One reply per auction type and purpose, with sandbox access, the video and a booking link.

      Head of Sales 1 day

    4. 4

      Send the same day

      Form answers pick the reply; a salesperson checks and sends within one working day.

      Head of Sales ongoing

    5. 5

      Put the videos on the pages

      Embed each walkthrough on the matching auction section so visitors see it before the form.

      Digital Marketing Manager 1 day

    6. 6

      Track

      Sandbox logins, bids placed and calls booked per request.

      Head of Sales monthly

    First result
    2 weeks
    Cost
    About a week of product and sales time
    Owner
    Head of Sales

    Smallest test

    Send one sample auction to the next three demo requests and compare how many book a call.

    Why it works

    Prepare one ready-made sample auction for each type Atreus supports (forward English, reverse English, Hybrid Dutch, Texas), with sample lots and two bidder logins. When a demo request arrives, the reply sent the same day includes access to the matching sandbox, a three-minute recorded walkthrough and a link to book a call. Types Atreus does not cover (spectrum, combinatorial) get the recorded walkthrough and a call offer.

    A buyer who can place a bid in a working auction on day one compares every other vendor's "schedule a demo" against it.

    Risks and how to handle them

    • The sandbox looks simpler than a custom build

      Say in the reply that it shows the engine, and that rules, design and integrations are tailored.

    Evidence

    • Your websiteRequest an Auction demo form: asks the auction type (English, Reverse, Multivariable, Hybrid Dutch, Dutch, Sealed Bid, Spectrum, SMR, Clock, Combinatorial, Other) and the purpose (Selling, Buying, Both); the page offers no live demo link, sandbox or video, and says the team will be in touch "as soon as possible".Source
    • Your websiteAuctions: English, reverse, Hybrid Dutch, multivariable, spectrum (SMR, clock, combinatorial) and real-time bidding, with sub-pages for sealed bid and Odoo auction software; demo button "Fill in this form to get a demo". The Auction Delivery Framework page says "Within one to two weeks of kickoff, Quintagroup provides a functional demo environment", discovery takes "three to five days" and production 4-6 weeks, and calls Atreus.auction "the working proof".Source
    • Atreus siteAtreus (atreus.auction), "Atreus SaaS Auction Platform": Free Trial with 10 auctions a month, Premium with 100 auctions a month at $499/month or $4,990/year, Enterprise on request with sealed bid, spectrum and real-time bidding; buttons "Try it free" and "Request a demo"; an Odoo CRM integration; the homepage read does not name Quintagroup or link to quintagroup.com, while Quintagroup's Atreus case links to the free registration.Source
  • 06

    The site gives different answers to simple questions. Auction experience is "over 5 years" on About us and "9 years" on the auctions page. The services index says "over 16 years of experience" while the company has traded since 2003 and the footer starts at 2002. The Agentic AI Deployment Framework has five phases on one page and six on another, with different delivery times. A careful buyer, or an AI assistant summarising the company, will pick one at random.

    Medium impactLow effort First result in 1 week Digital Marketing Manager

    First action

    Search the site for "years" and "phases" and list every conflicting statement.

    How to measure it

    Pages with conflicting company facts

    Now
    –
    Target
    0
    By
    30 days

    Steps

    1. 1

      List the conflicts

      Site search for years, since, phases and weeks.

      Digital Marketing Manager 1 day

    2. 2

      Decide the facts

      Founder signs off one fact sheet.

      CEO 1 day

    3. 3

      Reconcile the AI framework

      One phase count and one timeline range across both agentic pages.

      Head of Sales 1 day

    4. 4

      Fix the pages

      About, services index, auctions, AI pages and the footer.

      Digital Marketing Manager 1 day

    5. 5

      Switch to dates

      Replace year counts with "since" dates.

      Digital Marketing Manager half a day

    6. 6

      Match outside profiles

      Clutch, LinkedIn and the Atreus site use the same fact sheet.

      Digital Marketing Manager half a day

    First result
    1 week
    Cost
    $0
    Owner
    Digital Marketing Manager

    Smallest test

    Fix the two agentic AI pages today so they describe the same framework.

    Why it works

    Write a one-page fact sheet (founding year, years in auctions and e-procurement, team size, certification, headline project figures, each framework's phases and timelines) and correct every page to match. Use dates rather than counts where possible ("since 2003", "auctions since 2016") so pages do not go stale.

    Consistency is cheap credibility, and the AEO service's own advice is that answer engines favour clear, consistent facts.

    Risks and how to handle them

    • New pages reintroduce old numbers

      Keep the fact sheet as the source for every new page.

    Evidence

    • Your websiteAbout us: "on the international market since 2003"; ISO 27001:2022 certified; ProZorro.Sale enabled sales worth more than USD 700 million and ProZorro helped save USD 2 billion through reverse auctions; auctions developed for "over 5 years". Other pages say differently: the auctions page claims "9 years of experience", the services index "over 16 years of experience", and the footer reads "© 2002-2026". No team size, office or leadership is shown.Source
    • Your websiteAI section: /services/ai lists seven pages (consulting, development, generative, self-hosted platform, AI services, agentic AI solutions, agentic AI development); the hub and both agentic pages cite no case study. The two agentic pages disagree: one gives the "Agentic AI Deployment Framework (AADF)" five phases, 3 to 4 weeks for one workflow and 6 to 10 for multi-agent; the other gives six phases, "4-6 weeks" and "6-8 weeks". Example agents include invoice reconciliation, a compliance review of bid papers against procurement rules, and support resolution.Source
    • Your websiteServices index: 24 services, among them Blockchain, Internet of Things and PaaS; intro "Quintagroup is an IT solutions company with over 16 years of experience"; Prototyping links to the contact page; it lists "Machine Learning & AI" but not the newer agentic AI, self-hosted AI, AEO or Auction Delivery Framework pages.Source
  • 08

    The sitemap lists 4,847 URLs for a site with 24 services, 36 cases and 92 blog posts. The blog section alone has 349 entries. Much of the difference is Plone's image wrapper pages: URLs ending in "/view" whose only content is a file name as the heading and "Click to view full-size image". Image folders appear as pages titled "Blog Images" and "Service Images", and they show up in the site's feed as if they were articles.

    Medium impactLow effort First result in 1 month Senior Python Developer

    First action

    Count sitemap URLs ending in "/view" and check five of them in Search Console.

    How to measure it

    Sitemap size

    Now
    –
    Target
    Under 600 URLs, all content pages
    By
    30 days

    Steps

    1. 1

      Count by type

      Sitemap URLs by content type and path.

      Senior Python Developer half a day

    2. 2

      Check indexing

      Search Console coverage for /view pages.

      Digital Marketing Manager half a day

    3. 3

      Filter the sitemap

      Exclude Image, File and Folder types in the Plone sitemap settings.

      Senior Python Developer 1 day

    4. 4

      Add noindex

      On image and file wrapper views.

      Senior Python Developer 1 day

    5. 5

      Clean the feed

      Only articles, cases and service pages in the RSS feed.

      Senior Python Developer half a day

    6. 6

      Resubmit and watch

      Resubmit the sitemap and review indexed pages after four weeks.

      Digital Marketing Manager 1 month

    First result
    1 month
    Cost
    $0
    Owner
    Senior Python Developer

    Smallest test

    Exclude the Image type from the sitemap and compare the URL count before and after.

    Why it works

    Exclude image, file and folder content types from the Plone sitemap, mark their wrapper pages noindex, and keep the feed to real articles and service pages. Then check in Search Console how many of the wrapper pages were indexed and watch them drop out.

    Search engines and AI crawlers spend their visits on the pages that matter, and the site's feed stops announcing image folders. It also removes an easy objection when the company pitches its own AEO audit.

    Risks and how to handle them

    • Images lose image-search visibility

      The image files stay crawlable from the articles that use them.

    Evidence

    • Site researchSitemap (4,847 URLs, checked November 15, 2026): /services 3,290, /cms 967, /blog 349 (the blog lists 92 posts), /portfolio 91 and /solutions 49 (both last changed in 2019). It includes image wrapper pages ending in "/view", for example a blog image page whose heading is the file name "benefits-PDF-Word-Excel-generation.jpg" with "Click to view full-size image… Size: 649KB", and folders titled "Blog Images" and "Service Images".Source
    • Your websiteAnswer Engine Optimization page: sells a fixed-scope audit and an "Answer Readiness Framework (ARF)" of five phases: content audit, question-format headings (at least three per article), rewritten openings, a FAQ block of 4-6 pairs with FAQPage markup, and an AI citation layer; one page takes about 48 hours and a 10-15 page set 4-6 weeks; the one example (a Swiss insurer) has no figures.Source
  • 09

    The homepage title and headline sell a Python, JS and DevOps shop, the claim hundreds of agencies make. The figures just below (2,000,000 open procedures a year, 10,000 auctions a day) show what is actually rare about the company, and the newest AI pages are not in the menu, the homepage or the services index. Larger rivals now lead with AI and a result.

    Medium impactLow effort First result in 2 weeks Digital Marketing Manager

    First action

    Draft three headline options and test them with five recent clients.

    How to measure it

    Homepage click-through to service and case pages

    Now
    –
    Target
    +20%
    By
    60 days

    Steps

    1. 1

      Draft headlines

      Three options built on the procurement figures.

      Digital Marketing Manager 1 day

    2. 2

      Rewrite title and description

      Specialism first, technologies second.

      Digital Marketing Manager half a day

    3. 3

      Add AI and Auctions to the menu

      And to the services index.

      Senior Python Developer half a day

    4. 4

      Add an AI block

      Three lines and a link to the AI hub on the homepage.

      Digital Marketing Manager 1 day

    5. 5

      Add two quotes

      With role and sector.

      Head of Sales 1 week

    6. 6

      Measure

      Clicks from the homepage and search click-through for the brand.

      Digital Marketing Manager monthly

    First result
    2 weeks
    Cost
    $0
    Owner
    Digital Marketing Manager

    Smallest test

    Change the headline only and compare clicks to the auctions and e-procurement pages for two weeks.

    Why it works

    Rewrite the title and headline around the specialism ("Auction and e-procurement software, built in Python, proven on national systems"), add a line and a link for the AI services, add AI and Auctions to the menu and the services index, and place two client quotes near the top.

    A specific claim with a number beats a technology list, and pages that are not linked from the menu get little internal traffic.

    Risks and how to handle them

    • Python search traffic drops

      Keep Python in the title after the specialism and keep the Python service pages unchanged.

    Evidence

    • Your websiteHomepage: title "Python software development company - experts in VueJS, React and DevOps"; headline "We are Python JS & DevOps Professionals"; figures "2 000 000 open procedures a year", "10 000 auctions a day", "250 000 active procedures"; projects Prozorro, Prozorro sale, Mtender and Rialto; logos include "European Bank", Transparency International, Prozorro, Mastercard, Xperi and Tektelic; no testimonials and no mention of AI; menu: Home, Expertise, Services, Cases, About us, Blog, Contact us, Estimate project.Source
    • Your websiteServices index: 24 services, among them Blockchain, Internet of Things and PaaS; intro "Quintagroup is an IT solutions company with over 16 years of experience"; Prototyping links to the contact page; it lists "Machine Learning & AI" but not the newer agentic AI, self-hosted AI, AEO or Auction Delivery Framework pages.Source
    • Competitor researchRivals' own homepages: STX Next, "We turn isolated AI experiments into resilient operational systems for the Industrial and Financial sectors", 450+ specialists, five stars across 101 Clutch reviews and results such as "14,000+ staff hours saved every year"; Django Stars, "AI-Powered Software Development & Consulting Company", 100+ developers and a 92.7 NPS; Auction Software, "OVER 500 AUCTION SITES CREATED", "Schedule a Demo", no pricing or free trial shown.Source
  • 15

    Every form on the site ends with the same line: the team will be in touch "as soon as possible". There is no calendar, no chat and no stated reply time, only two US phone numbers for a team based in Lviv. Several service pages invite visitors to "schedule" a discovery session or assessment, but there is nothing to schedule with.

    Medium impactLow effort First result in 1 week Head of Sales

    First action

    Set up one shared calendar with ten slots a week and put it on the contact thank-you page.

    How to measure it

    Form-to-call rate

    Now
    –
    Target
    +25%
    By
    60 days

    Steps

    1. 1

      Choose the tool

      A scheduling tool that works with Google Workspace calendars.

      Head of Sales half a day

    2. 2

      Open slots

      Ten a week covering Europe and US East Coast mornings.

      Head of Sales half a day

    3. 3

      Add to thank-you pages

      After contact, estimate and demo forms.

      Senior Python Developer 1 day

    4. 4

      State the reply time

      Replace "as soon as possible" with one business day on every form.

      Digital Marketing Manager half a day

    5. 5

      Send a prep note

      A short confirmation with what to bring to the call.

      Head of Sales half a day

    6. 6

      Track

      Bookings, no-shows and time from form to call.

      Head of Sales monthly

    First result
    1 week
    Cost
    A low monthly tool fee
    Owner
    Head of Sales

    Smallest test

    Add a booking link to the reply e-mail for one week and count how many use it.

    Why it works

    Add a booking calendar with 30-minute discovery slots across European and US East Coast hours, shown on the thank-you page after every form and on the contact page. Replace "as soon as possible" with a stated reply time of one business day and meet it.

    Letting a buyer choose a time removes the first round of e-mail and a stated reply time sets an expectation that rivals leave open.

    Risks and how to handle them

    • Unqualified bookings fill the calendar

      Show the calendar only after a form is submitted.

    Evidence

    • Your websiteContact form: Name, E-Mail, Phone, Comment ("Tell us what you are looking for..."), Country Code and a reCAPTCHA; no service, workflow or tech-stack question, no file upload and no calendar booking; the reply is promised "as soon as possible". The Estimate project form adds one file upload and a "Project details" box, with the same "as soon as possible". Two US phone numbers and info@ are the other routes; no live chat is shown.Source
    • Your websiteNew service pages promise a quote in five business days: agentic AI development offers a "fixed-cost proposal within 5 business days" and its FAQ says a fixed-cost estimate needs the target workflow and current tech stack; the AEO page asks for "your URL list and target query set for a fixed-price estimate within 5 business days"; the self-hosted AI and Auction Delivery Framework pages make the same promise. Each "Contact Quintagroup" button links to the general /contact form.Source

This quarter 11

The core plays for the next three months.

  • 04

    Quintagroup sells an Answer Readiness Framework: question-format headings, rewritten openings, FAQ blocks with markup, about 48 hours a page. Its own blog has 92 posts, 23 of them from December 2022 to August 2023, then almost nothing until two posts in early 2026, and no authors. The marketing vacancy describes optimising articles by hand around given key phrases. The company has the method and the backlog, and the work is repetitive.

    Medium impactMedium effort First result in 6 weeks Digital Marketing Manager

    First action

    Pull search clicks for all 92 posts and sort them into refresh, merge and retire.

    How to measure it

    Posts restructured and republished

    Now
    –
    Target
    30 posts
    By
    90 days

    Steps

    1. 1

      Sort the 92 posts

      Search clicks and topic fit decide refresh, merge or retire.

      Digital Marketing Manager 3 days

    2. 2

      Turn the ARF into a checklist

      The five phases as pass or fail checks a script can apply to a page.

      Digital Marketing Manager 2 days

    3. 3

      Build the audit

      Inputs: the RSS feed, the sitemap, Search Console data. Output: a table of posts with scores and the missing elements.

      Senior Python Developer 4 days

    4. 4

      Build the drafting step

      For each chosen post, a draft with question headings, a direct opening answer, four to six FAQ pairs, markup and flagged outdated facts. Drafts land in a review folder, not on the site.

      Senior Python Developer 1 week

    5. 5

      Edit and approve

      An editor checks facts, adds the author and a subject expert's sign-off, and publishes in Plone.

      Digital Marketing Manager ongoing

    6. 6

      Redirect the retired

      Posts with no value point to the closest current page.

      Senior Python Developer 1 day

    7. 7

      Measure

      Clicks, impressions and AI-assistant citations for refreshed posts against untouched ones.

      Digital Marketing Manager monthly

    First result
    6 weeks
    Cost
    About two weeks of development and an editor's time
    Owner
    Digital Marketing Manager

    Smallest test

    Restructure five posts by hand with the ARF and compare their clicks after six weeks.

    Why it works

    Build a content pipeline that takes each post from the RSS feed and the sitemap, scores it against the ARF checklist (question headings, a direct answer in the first paragraph, FAQ pairs, dated facts, author), checks search data for whether it still earns visits, and drafts the restructured version with a FAQ block and markup. An editor approves, corrects and publishes each one in Plone, with an author and an updated date. Posts with no traffic and no relevance are listed for redirect rather than rewritten.

    Refreshing existing pages is usually faster than writing new ones, and applying the published method to the company's own site gives the AEO service a measurable example, which its one current example lacks.

    Risks and how to handle them

    • Drafted text reads as generic

      Require one concrete project detail from an engineer in every refreshed post.

    • Rankings dip after restructuring

      Refresh in batches of five and keep URLs unchanged.

    Evidence

    • Your websiteBlog: 92 posts, no authors shown, no newsletter signup. Newest: "Powerful PDF, Word, and Excel Generation from Your CMS with Quintagroup" (Mar 02, 2026) and "Staff Augmentation VS Outsourcing" (Feb 11, 2026); before those, one post in April 2025 and then August 2023. 23 posts date from December 2022 to August 2023, many on hiring topics ("How to Hire Vue.js Developers", "Cypress Automation Developers — How to Hire").Source
    • Your websiteAnswer Engine Optimization page: sells a fixed-scope audit and an "Answer Readiness Framework (ARF)" of five phases: content audit, question-format headings (at least three per article), rewritten openings, a FAQ block of 4-6 pairs with FAQPage markup, and an AI citation layer; one page takes about 48 hours and a 10-15 page set 4-6 weeks; the one example (a Swiss insurer) has no figures.Source
    • HiringDigital Marketing Manager vacancy: "Planning digital marketing campaigns, including web, SEO/SEM, social media, and display advertising"; analysing market trends, sales and performance indicators; preparing presentations and brochures; optimising English articles for SEO around given key phrases; running social media; "Solid knowledge of website analytics tools"; one year of experience asked.Source
  • 07

    One of Quintagroup's case studies describes a system it built to collect auction, offer and tender data automatically with Python, Selenium and AWS. Its best-known work is public e-procurement, bought through public tenders and donor-funded programmes. The company has built the tool for a client that its own sales team needs.

    High impactMedium effort First result in 5 weeks Head of Sales

    First action

    List the ten portals and feeds where past opportunities came from and the keywords that described them.

    How to measure it

    Qualified tenders reviewed per quarter

    Now
    –
    Target
    25 reviewed, 5 pursued
    By
    2 quarters

    Steps

    1. 1

      Define relevance

      Keywords, classification codes, budget floor and regions the company can serve.

      Head of Sales 2 days

    2. 2

      Choose the sources

      Public tender portals with open data or feeds; respect each site's terms and skip any that forbid automated reading.

      Business Analyst 2 days

    3. 3

      Build the collector

      Inputs: feeds and open contracting data. Storage: one table of notices with deadline, buyer type, value and documents.

      Senior Python Developer 1 week

    4. 4

      Build the scoring and summary

      Each notice gets a fit score against the reference list and a half-page summary of requirements and required references.

      Senior Python Developer 1 week

    5. 5

      Send the weekly digest

      To the sales lead, with a bid, partner or skip decision recorded by a person for each notice.

      Head of Sales weekly

    6. 6

      Build the partner list

      Local integrators that have won similar tenders, for consortium bids.

      Head of Sales 2 weeks

    7. 7

      Review hit rate

      Notices flagged, pursued, shortlisted and won; tune the keywords quarterly.

      Head of Sales quarterly

    First result
    5 weeks
    Cost
    Three to four weeks of development; small hosting cost
    Owner
    Head of Sales

    Smallest test

    Check three portals by hand every Monday for a month and count the relevant notices.

    Why it works

    Build a monitor that reads public tender portals and open contracting feeds each day for notices about e-procurement systems, e-auction modules, asset-sale platforms and open contracting data work, scores each against Quintagroup's references, and sends a weekly digest with deadline, budget, buyer type, required references and a suggested bid or partner decision. A person decides which to pursue. Notices needing a local prime contractor are flagged for partner outreach.

    Tenders in this niche are few and specific, and the reference list (a national procurement system and a national asset-sale system) is rare. Seeing every relevant notice early matters more than volume.

    Risks and how to handle them

    • Public bids take months and many are lost

      Pursue only notices where the references match directly, and partner where a local prime is needed.

    • Portal terms restrict automated access

      Use official feeds and open data only.

    Evidence

    • Your websiteCases: 36 case studies on three pages of 12, with no filter by industry or technology. The two case pages read in full (Mastercard API enhancement and the Atreus auction platform) give no results, dates, team size, duration or client quote. The list runs from recent work (an insurance image-recognition service, "The Atlas Project" that collects auction and tender data with Python, Selenium and AWS) back to Plone-era websites and a Plone themes web store.Source
    • Your websiteE-procurement: the open-source OpenProcurement toolkit, an e-Auction module, a test stand and OCDS ("Open Contracting Data Standard was incorporated in the ProZorro system's core"); projects ProZorro, Prozorro.sale, Rialto and MTender; nine brochures (OpenProcurement, ProZorro and a short version) each in English, French and Italian, while the page itself is in English; "Request a demo" links to the general /contact form.Source
    • Your websiteHomepage: title "Python software development company - experts in VueJS, React and DevOps"; headline "We are Python JS & DevOps Professionals"; figures "2 000 000 open procedures a year", "10 000 auctions a day", "250 000 active procedures"; projects Prozorro, Prozorro sale, Mtender and Rialto; logos include "European Bank", Transparency International, Prozorro, Mastercard, Xperi and Tektelic; no testimonials and no mention of AI; menu: Home, Expertise, Services, Cases, About us, Blog, Contact us, Estimate project.Source
  • 10

    Quintagroup's Clutch profile shows 4.5 from 14 reviews, and the most recent is dated May 28, 2021. A buyer comparing shortlists sees a rival with 101 reviews. More than five years without a new review reads as more than five years without a happy client, which the 2026 work contradicts.

    Medium impactLow effort First result in 1 month Project Manager

    First action

    List every project that shipped a milestone in the last 12 months and the client contact for each.

    How to measure it

    New Clutch reviews

    Now
    –
    Target
    8 new reviews
    By
    2 quarters

    Steps

    1. 1

      List recent milestones

      Projects, contacts and what was delivered in the last 12 months.

      Project Manager 2 days

    2. 2

      Write the request

      Short, personal, with the review link and the project summary.

      Head of Sales 1 day

    3. 3

      Build the trigger

      Input: a milestone or close status in the project system. Output: a drafted request in the project manager's mailbox and a reminder task.

      Senior Python Developer 3 days

    4. 4

      Approve and send

      The project manager edits, sends or skips each request; one drafted follow-up after two weeks.

      Project Manager ongoing

    5. 5

      Reuse the answers

      With permission, quotes go onto case pages and the homepage.

      Digital Marketing Manager ongoing

    6. 6

      Reply publicly

      A short reply to each review, including the mixed ones.

      Head of Sales ongoing

    7. 7

      Track

      Requests sent, reviews published and referrals received per quarter.

      Head of Sales quarterly

    First result
    1 month
    Cost
    A few days of development
    Owner
    Project Manager

    Smallest test

    Send five personal requests to the happiest recent clients this week.

    Why it works

    Build a small follow-up automation tied to project milestones. When a project manager marks a release or a project close in the project system, it drafts a personal review request for the client contact, with the project summary pre-filled, and reminds the project manager to send it; two weeks later it drafts one polite follow-up. The same trigger asks for a two-sentence quote for the case study and a referral introduction. The project manager approves each message and can skip any client.

    Reviews are given when asked at a moment of success. Tying the ask to a milestone makes it routine instead of a campaign that never starts.

    Risks and how to handle them

    • Public-sector clients cannot review

      Ask for a reference letter or a quote for the case study instead.

    • A critical review appears

      Ask after successful milestones and reply constructively to every review.

    Evidence

    • ReviewsClutch profile: 4.5 from 14 reviews; the most recent review is dated May 28, 2021; $25 to $49 an hour, minimum project $10,000+, 50 to 249 employees, founded 2003, L'viv, Ukraine. The homepage links the profile with an icon only.Source
    • HiringBusiness Analyst vacancy: "Gathering and analyzing data in support of business cases, proposed projects, and system requirements" and "Managing reports". Project manager vacancy: "Monitoring progress and managing reports" and "Being responsible for resources and time planning". Both list public procurement experience as a plus.Source
    • Competitor researchRivals' own homepages: STX Next, "We turn isolated AI experiments into resilient operational systems for the Industrial and Financial sectors", 450+ specialists, five stars across 101 Clutch reviews and results such as "14,000+ staff hours saved every year"; Django Stars, "AI-Powered Software Development & Consulting Company", 100+ developers and a 92.7 NPS; Auction Software, "OVER 500 AUCTION SITES CREATED", "Schedule a Demo", no pricing or free trial shown.Source
  • 11

    The Atreus homepage read for this playbook does not name Quintagroup or link to quintagroup.com, even though Quintagroup's own framework page calls Atreus "the working proof". An auction buyer on Atreus cannot see that the team behind it built auctions for a national asset-sale system, and Enterprise visitors who need sealed bid or spectrum formats are a custom-project lead for the parent company.

    Medium impactLow effort First result in 3 weeks Atreus Product Owner

    First action

    Add a short "Built by" block with three proof figures to the Atreus homepage and pricing page.

    How to measure it

    Atreus Enterprise enquiries passed to sales

    Now
    –
    Target
    5 a month
    By
    90 days

    Steps

    1. 1

      Write the proof block

      Who built it, since when, and the national-scale figures.

      Digital Marketing Manager 1 day

    2. 2

      Add it to Atreus

      Homepage, pricing and footer.

      Atreus Product Owner 2 days

    3. 3

      Link both ways

      Auction pages on quintagroup.com to Atreus "Try it free"; Atreus Enterprise to custom auction development.

      Digital Marketing Manager 1 day

    4. 4

      Share one intake

      Atreus demo and Enterprise forms ask auction type and purpose, like the main site's form.

      Atreus Product Owner 2 days

    5. 5

      Route to sales

      One mailbox and one record for both sites' auction enquiries.

      Head of Sales 1 day

    6. 6

      Measure

      Enquiries by site and how many become custom projects.

      Head of Sales monthly

    First result
    3 weeks
    Cost
    $0
    Owner
    Atreus Product Owner

    Smallest test

    Add the "Built by" block to the pricing page only and watch Enterprise enquiries for a month.

    Why it works

    Add a "Built by Quintagroup" section to Atreus with the procurement and asset-sale figures and the ISO 27001:2022 certification, link both ways between the two sites, and route Atreus Enterprise and "Request a demo" enquiries into the same sales process and intake as the auction pages on quintagroup.com.

    A SaaS product with no visible company behind it is a risk for a buyer running regulated or high-value auctions. The parent's references remove that risk at no cost.

    Risks and how to handle them

    • A services brand makes the SaaS look like a side project

      Keep the Atreus brand first and present Quintagroup as the engineering company behind it.

    Evidence

    • Atreus siteAtreus (atreus.auction), "Atreus SaaS Auction Platform": Free Trial with 10 auctions a month, Premium with 100 auctions a month at $499/month or $4,990/year, Enterprise on request with sealed bid, spectrum and real-time bidding; buttons "Try it free" and "Request a demo"; an Odoo CRM integration; the homepage read does not name Quintagroup or link to quintagroup.com, while Quintagroup's Atreus case links to the free registration.Source
    • Your websiteAuctions: English, reverse, Hybrid Dutch, multivariable, spectrum (SMR, clock, combinatorial) and real-time bidding, with sub-pages for sealed bid and Odoo auction software; demo button "Fill in this form to get a demo". The Auction Delivery Framework page says "Within one to two weeks of kickoff, Quintagroup provides a functional demo environment", discovery takes "three to five days" and production 4-6 weeks, and calls Atreus.auction "the working proof".Source
  • 12

    Atreus gives away 10 auctions a month and charges $499 a month for 100. Between "Try it free" and "Order" there is a clear set of moments: registered but no auction created, first auction run, the monthly limit approaching, and a format requested that only Enterprise has. Each one is a reason to write to the account, and each can be read from the product's own data.

    Medium impactMedium effort First result in 6 weeks Atreus Product Owner

    First action

    Pull the last 90 days of registrations and count how many created an auction, completed one and hit the limit.

    How to measure it

    Trial to first completed auction rate

    Now
    –
    Target
    40%
    By
    2 quarters

    Steps

    1. 1

      Map the funnel

      Registrations, first auction created, first completed, limit reached, upgraded, for the last 90 days.

      Atreus Product Owner 2 days

    2. 2

      Define the events

      Five triggers read from the Atreus database or API.

      Senior Python Developer 2 days

    3. 3

      Write the messages

      One short e-mail per trigger with one action each.

      Digital Marketing Manager 3 days

    4. 4

      Build the sender

      A daily job matches accounts to triggers and sends approved templates through the existing mail service, with unsubscribe and frequency limits.

      Senior Python Developer 1 week

    5. 5

      Alert sales

      High-intent accounts create a task with usage details; a person writes or approves the outreach.

      Head of Sales 2 days

    6. 6

      Hold out a control

      One in five new accounts gets no messages for the first two months, to measure the difference.

      Atreus Product Owner 2 months

    7. 7

      Review

      Activation and upgrade rates by trigger; rewrite or remove messages that do nothing.

      Atreus Product Owner monthly

    First result
    6 weeks
    Cost
    About two weeks of development
    Owner
    Atreus Product Owner

    Smallest test

    E-mail by hand every account that registered last month and has no auction, offering a sample auction.

    Why it works

    Build lifecycle follow-ups driven by Atreus account events. Inputs: registration date, auctions created and completed, bidders invited and the type chosen. Outputs: a setup guide on day one, a sample auction for accounts with none after three days, a note after the first completed auction, an upgrade prompt at seven of ten monthly auctions, and an alert to sales when an account asks about sealed bid or runs high-value lots. Product messages go out automatically from approved templates; anything from a named salesperson is drafted for approval.

    Free trials convert when users reach a first success quickly and are asked to upgrade at the moment the limit matters.

    Risks and how to handle them

    • Too few trials to measure

      Run the control for longer and judge on first completed auction, which happens sooner than upgrade.

    Evidence

    • Atreus siteAtreus (atreus.auction), "Atreus SaaS Auction Platform": Free Trial with 10 auctions a month, Premium with 100 auctions a month at $499/month or $4,990/year, Enterprise on request with sealed bid, spectrum and real-time bidding; buttons "Try it free" and "Request a demo"; an Odoo CRM integration; the homepage read does not name Quintagroup or link to quintagroup.com, while Quintagroup's Atreus case links to the free registration.Source
    • Your websiteRequest an Auction demo form: asks the auction type (English, Reverse, Multivariable, Hybrid Dutch, Dutch, Sealed Bid, Spectrum, SMR, Clock, Combinatorial, Other) and the purpose (Selling, Buying, Both); the page offers no live demo link, sandbox or video, and says the team will be in touch "as soon as possible".Source
  • 13

    Quintagroup's price signals already exist in public: Clutch shows $25 to $49 an hour and a $10,000 minimum, and Atreus lists $499 a month. But the new pages that promise fixed-price work (a single-workflow agent, an AEO audit, a self-hosted AI setup, an auction pilot) give no figure at all. The AEO page even says the work "often costs less than one month's paid search spend" without a number.

    Medium impactLow effort First result in 1 month Head of Sales

    First action

    Work out the delivery cost and margin of the last three engagements of each type.

    How to measure it

    Packaged-offer enquiries

    Now
    –
    Target
    10 a month
    By
    90 days

    Steps

    1. 1

      Cost the packages

      Hours and margin from past work.

      Head of Sales 1 week

    2. 2

      Fix the scopes

      What is in, what is out, and the timeline already stated on each page.

      Business Analyst 1 week

    3. 3

      Set "from" prices

      One entry price per package, with what raises it.

      CEO 2 days

    4. 4

      Publish on the pages

      A short offer box above the form.

      Digital Marketing Manager 2 days

    5. 5

      Brief sales

      How to move from the package to a larger scope.

      Head of Sales 1 day

    6. 6

      Review

      Enquiry quality and win rate with and without the price shown.

      Head of Sales quarterly

    First result
    1 month
    Cost
    $0
    Owner
    Head of Sales

    Smallest test

    Add a "from" price to the AEO audit page only and compare enquiries for a month.

    Why it works

    Define three or four packaged entry offers with a scope, a timeline taken from the existing pages and a "from" price: an AEO audit of 10 to 15 pages, a single-workflow agent pilot, a basic self-hosted AI setup, and an auction pilot built on the framework. Put the price and what is included on each service page next to the intake form.

    A buyer with a budget can qualify themselves in seconds, and a visible entry price turns the five-day proposal into a confirmation rather than a first number. Rivals in this market mostly hide prices.

    Risks and how to handle them

    • A low entry price anchors larger deals

      State clearly what the package excludes and show a typical full-project range.

    Builds on

    Evidence

    • ReviewsClutch profile: 4.5 from 14 reviews; the most recent review is dated May 28, 2021; $25 to $49 an hour, minimum project $10,000+, 50 to 249 employees, founded 2003, L'viv, Ukraine. The homepage links the profile with an icon only.Source
    • Your websiteNew service pages promise a quote in five business days: agentic AI development offers a "fixed-cost proposal within 5 business days" and its FAQ says a fixed-cost estimate needs the target workflow and current tech stack; the AEO page asks for "your URL list and target query set for a fixed-price estimate within 5 business days"; the self-hosted AI and Auction Delivery Framework pages make the same promise. Each "Contact Quintagroup" button links to the general /contact form.Source
    • Atreus siteAtreus (atreus.auction), "Atreus SaaS Auction Platform": Free Trial with 10 auctions a month, Premium with 100 auctions a month at $499/month or $4,990/year, Enterprise on request with sealed bid, spectrum and real-time bidding; buttons "Try it free" and "Request a demo"; an Odoo CRM integration; the homepage read does not name Quintagroup or link to quintagroup.com, while Quintagroup's Atreus case links to the free registration.Source
  • 14

    The only AI result on the site is two lines inside the self-hosted platform page: a FinTech client replaced four AI subscriptions and cut onboarding "from 2 weeks to 3 days". The AI hub and both agentic pages cite no case at all, and none of the 36 case studies is about this work. The best evidence is buried where the cases list cannot show it.

    Medium impactLow effort First result in 1 month Digital Marketing Manager

    First action

    Ask the client whether the two figures can be published with their sector and size.

    How to measure it

    AI enquiries that cite the case study

    Now
    –
    Target
    1 in 3 AI enquiries
    By
    90 days

    Steps

    1. 1

      Get permission

      Named or anonymised, with figures.

      Head of Sales 1 week

    2. 2

      Interview the team

      What was built, what changed, what was hard.

      Digital Marketing Manager 2 days

    3. 3

      Write the case

      Standard header, timeline, architecture sketch and outcomes.

      Digital Marketing Manager 1 week

    4. 4

      Add a quote

      Two sentences from the client's lead.

      Head of Sales 1 week

    5. 5

      Link it everywhere

      AI hub, each AI page, homepage and the cases list.

      Digital Marketing Manager 1 day

    6. 6

      Plan the next two

      The estimate desk from idea 1 and the next client engagement.

      Head of Sales 1 day

    First result
    1 month
    Cost
    Staff time only
    Owner
    Digital Marketing Manager

    Smallest test

    Publish a 300-word version this week and link it from the self-hosted AI page.

    Why it works

    Write that engagement up as a full case study in the new format: the starting situation, the deployment on the client's own cloud, the seven-phase process as it actually ran, the timeline, the two measured outcomes and a client quote. Link it from the AI hub, every AI page and the homepage, and make a second one as soon as another AI engagement finishes.

    For a new service line the first documented result does most of the selling, and buyers with sensitive data want to see exactly how data stayed inside the client's environment.

    Risks and how to handle them

    • The client declines to be named

      Publish it anonymised by sector and size, with the figures.

    Evidence

    • Your websiteSelf-hosted AI platform page: built on LibreChat, an "open-source AI assistant platform that runs entirely on a company's own infrastructure"; Quintagroup says it has deployed these since 2024 under a seven-phase "Private AI Deployment Framework (PADF)"; a FinTech example replaced four AI subscriptions and cut onboarding "from 2 weeks to 3 days"; a basic setup "can be live in 2 weeks".Source
    • Your websiteAI section: /services/ai lists seven pages (consulting, development, generative, self-hosted platform, AI services, agentic AI solutions, agentic AI development); the hub and both agentic pages cite no case study. The two agentic pages disagree: one gives the "Agentic AI Deployment Framework (AADF)" five phases, 3 to 4 weeks for one workflow and 6 to 10 for multi-agent; the other gives six phases, "4-6 weeks" and "6-8 weeks". Example agents include invoice reconciliation, a compliance review of bid papers against procurement rules, and support resolution.Source
  • 16

    The marketing vacancy asks one person with a year of experience to plan campaigns across web, search, social and display, write brochures, optimise articles and also analyse market trends, sales and performance indicators. The data for that analysis sits in several places: Google Analytics, Tag Manager, a visitor-identification tool, the forms mailbox and Search Console. Pulling it together by hand each week is the task most likely to be dropped.

    Medium impactMedium effort First result in 3 weeks Digital Marketing Manager

    First action

    Write down the eight numbers the founder actually wants to see each Monday.

    How to measure it

    Hours spent compiling the weekly report

    Now
    –
    Target
    Under 30 minutes a week
    By
    60 days

    Steps

    1. 1

      Choose the eight numbers

      Enquiries by service, proposals sent, organic visits to service pages, identified accounts, Atreus trials and similar.

      CEO 1 day

    2. 2

      Fix the tracking gaps

      Each form fires one named event with its service; check that events are not double counted.

      Digital Marketing Manager 3 days

    3. 3

      Build the collector

      Scheduled pulls from the analytics, search and visitor tools' APIs and the form records into one table.

      Senior Python Developer 1 week

    4. 4

      Build the summary

      Week-on-week changes and three observations written in plain language, with links to the underlying data.

      Senior Python Developer 3 days

    5. 5

      Review before sending

      The marketing manager checks the numbers, corrects the observations and adds proposed actions.

      Digital Marketing Manager weekly

    6. 6

      Decide on Monday

      Fifteen minutes with the founder and sales lead; one action a week.

      CEO weekly

    7. 7

      Check accuracy monthly

      Compare report figures with the sources by hand.

      Digital Marketing Manager monthly

    First result
    3 weeks
    Cost
    About two weeks of development
    Owner
    Digital Marketing Manager

    Smallest test

    Fill the one-page template by hand for two Mondays and see which numbers lead to a decision.

    Why it works

    Build a weekly report job. Inputs: Google Analytics and Search Console data, form submissions by form and service, identified companies by page group, and Atreus registrations. Output: a one-page report every Monday with the numbers, the changes from last week, the pages gaining or losing visits, and three plain-language observations. The marketing manager reviews it, adds decisions and sends it to the founder and sales lead.

    A report that arrives without effort gets read, and a marketing team of one spends its time on the decisions instead of the spreadsheet.

    Risks and how to handle them

    • Low volumes make weekly changes noisy

      Show four-week averages next to weekly numbers.

    Builds on

    Evidence

    • HiringDigital Marketing Manager vacancy: "Planning digital marketing campaigns, including web, SEO/SEM, social media, and display advertising"; analysing market trends, sales and performance indicators; preparing presentations and brochures; optimising English articles for SEO around given key phrases; running social media; "Solid knowledge of website analytics tools"; one year of experience asked.Source
    • Site researchThird-party scripts on the homepage (11 services, about 2.2 MB): three visitor-identification tools at once (Leadfeeder, Happierleads and Salespanel), Google Tag Manager, Google Analytics, a Facebook script and a request to Tawk.to although no chat is shown. The site runs on Plone with Google Workspace mail and an RSS feed.Source
    • PageSpeedGoogle PageSpeed (November 15, 2026): real visitors pass Core Web Vitals (LCP 1.7 s, INP 54 ms, CLS 0.01 at the 75th percentile, LCP about 20% faster than before). The phone lab test scores 68/100 (desktop 92) with a 6.5 s LCP whose element is the cookie notice; the homepage weighs about 6.3 MB with 792 KB of unused JavaScript; two accessibility issues (colour contrast, links without a name).Source
  • 17

    Real visitors already pass Core Web Vitals on quintagroup.com (LCP 1.7 seconds, and getting faster), so speed is not hurting the site today. The phone lab test still scores 68 because the largest element it measures is the cookie notice and because 792 KB of JavaScript, much of it third-party, is loaded and not used. A DevOps company's own lab score is something technical buyers do check.

    Low impactLow effort First result in 2 weeks Senior Python Developer

    First action

    List every tag in Google Tag Manager with its owner and the last time anyone used its data.

    How to measure it

    Mobile lab performance score

    Now
    –
    Target
    85+
    By
    60 days

    Steps

    1. 1

      Audit the tags

      Owner and use for each of the 11 third-party services.

      Digital Marketing Manager 1 day

    2. 2

      Remove the unused

      Duplicate visitor tools and the leftover chat request.

      Digital Marketing Manager 1 day

    3. 3

      Rework the cookie notice

      A smaller bar that is not the largest element.

      Senior Python Developer 1 day

    4. 4

      Scope the reCAPTCHA

      Load it only where a form is shown.

      Senior Python Developer 1 day

    5. 5

      Cache static files

      Long cache lifetimes for images, fonts and scripts.

      DevOps Engineer 1 day

    6. 6

      Fix accessibility

      Contrast and link names.

      Senior Python Developer 1 day

    First result
    2 weeks
    Cost
    $0
    Owner
    Senior Python Developer

    Smallest test

    Pause two unused tags and rerun the phone test.

    Why it works

    Treat this as housekeeping, not an emergency. Render the cookie notice so it is not the largest painted element, load the reCAPTCHA only on pages with a form, remove the tags for tools no longer used (see idea 2 and the leftover chat request), set long cache lifetimes for static files, and fix the two accessibility issues.

    Fewer scripts mean fewer things to own and maintain, and a clean score removes a small objection during technical due diligence.

    Risks and how to handle them

    • Removing a tag breaks a report someone relies on

      Pause each tag for two weeks before deleting it.

    Builds on

    Evidence

    • PageSpeedGoogle PageSpeed (November 15, 2026): real visitors pass Core Web Vitals (LCP 1.7 s, INP 54 ms, CLS 0.01 at the 75th percentile, LCP about 20% faster than before). The phone lab test scores 68/100 (desktop 92) with a 6.5 s LCP whose element is the cookie notice; the homepage weighs about 6.3 MB with 792 KB of unused JavaScript; two accessibility issues (colour contrast, links without a name).Source
    • Site researchThird-party scripts on the homepage (11 services, about 2.2 MB): three visitor-identification tools at once (Leadfeeder, Happierleads and Salespanel), Google Tag Manager, Google Analytics, a Facebook script and a request to Tawk.to although no chat is shown. The site runs on Plone with Google Workspace mail and an RSS feed.Source
  • 18

    Quintagroup already paid to produce nine procurement brochures in English, French and Italian, which shows where it expected demand. The translations only exist as PDF downloads on an English page. A procurement official searching in French or Italian for an open-source e-procurement system will not find a PDF that sits behind an English page.

    Medium impactLow effort First result in 2 months Digital Marketing Manager

    First action

    Check Search Console for impressions from French- and Italian-speaking countries on the e-procurement pages.

    How to measure it

    Non-English procurement enquiries

    Now
    –
    Target
    3 a month
    By
    2 quarters

    Steps

    1. 1

      Check demand

      Impressions and downloads by country for the brochures.

      Digital Marketing Manager 1 day

    2. 2

      Extract the text

      From the three brochures in each language.

      Digital Marketing Manager 2 days

    3. 3

      Have a native speaker review

      Update dates and figures.

      Digital Marketing Manager 1 week

    4. 4

      Build two pages

      With language tags linking the versions.

      Senior Python Developer 3 days

    5. 5

      Translate the demo form

      And the first reply.

      Head of Sales 2 days

    6. 6

      Measure

      Visits, downloads and enquiries by language.

      Digital Marketing Manager quarterly

    First result
    2 months
    Cost
    A native-speaker review of two pages
    Owner
    Digital Marketing Manager

    Smallest test

    Publish the French page only and watch impressions for six weeks.

    Why it works

    Publish one landing page in French and one in Italian from the existing brochure text: what OpenProcurement is, the ProZorro results, the modules and a demo request in that language. Link the language versions to each other properly and route those requests to a salesperson who can reply in the language or with a translated first answer.

    The translation cost is already spent, and competition for procurement-software searches in French and Italian is thinner than in English.

    Risks and how to handle them

    • No one on the team can sell in the language

      Say on the page that calls are held in English, and use a local partner for tenders.

    Evidence

    • Your websiteE-procurement: the open-source OpenProcurement toolkit, an e-Auction module, a test stand and OCDS ("Open Contracting Data Standard was incorporated in the ProZorro system's core"); projects ProZorro, Prozorro.sale, Rialto and MTender; nine brochures (OpenProcurement, ProZorro and a short version) each in English, French and Italian, while the page itself is in English; "Request a demo" links to the general /contact form.Source
    • Your websiteHomepage: title "Python software development company - experts in VueJS, React and DevOps"; headline "We are Python JS & DevOps Professionals"; figures "2 000 000 open procedures a year", "10 000 auctions a day", "250 000 active procedures"; projects Prozorro, Prozorro sale, Mtender and Rialto; logos include "European Bank", Transparency International, Prozorro, Mastercard, Xperi and Tektelic; no testimonials and no mention of AI; menu: Home, Expertise, Services, Cases, About us, Blog, Contact us, Estimate project.Source
  • 20

    Quintagroup holds ISO 27001:2022 and sells to public bodies and financial firms, the buyers who send long vendor and security questionnaires and formal tender documents. The Business Analyst role is described as gathering and analysing data for "business cases" and "proposed projects". Answering the same hundred questions about hosting, access control, staffing and references by hand is slow work for the people who should be scoping projects.

    Medium impactMedium effort First result in 1 month Business Analyst

    First action

    Gather the last five completed questionnaires and count how many questions overlap.

    How to measure it

    Median hours per questionnaire

    Now
    –
    Target
    Half the current time
    By
    2 quarters

    Steps

    1. 1

      Collect past responses

      Questionnaires, tender responses, policies and reference texts from the last two years.

      Business Analyst 1 week

    2. 2

      Build the library

      One approved answer per recurring question, with owner and review date, in the team's document space.

      Business Analyst 1 week

    3. 3

      Build the matcher

      Input: a new questionnaire as a spreadsheet or document. Output: a draft with matched answers, their sources and a list of unmatched questions.

      Senior Python Developer 1 week

    4. 4

      Set approvals

      Security answers approved by the certification owner; commercial answers by the sales lead; nothing is submitted without sign-off.

      Head of Sales 1 day

    5. 5

      Feed new answers back

      Each completed response updates the library.

      Business Analyst ongoing

    6. 6

      Measure

      Hours per response, share of questions matched and bids submitted per quarter.

      Head of Sales quarterly

    First result
    1 month
    Cost
    Two to three weeks of analyst and development time
    Owner
    Business Analyst

    Smallest test

    Build the library for the security section only and use it on the next questionnaire.

    Why it works

    Collect the last two years of completed questionnaires, tender responses, policies and reference descriptions into one approved answer library with an owner and a review date for each answer. Build a drafting tool: it takes a new questionnaire or tender document, matches each question to approved answers, fills a first draft with the source of every answer, and lists the questions with no match. An analyst completes and signs off every response; security answers are approved by the person responsible for the certification.

    Most questionnaire questions repeat. Starting from approved answers shortens response time and keeps statements consistent across bids, which matters when contracts refer back to them.

    Risks and how to handle them

    • An outdated answer is reused

      Review dates on every answer and a block on answers past their date.

    • Client-specific details leak between responses

      Strip client names from library answers and keep references in a separate approved list.

    Evidence

    • Your websiteAbout us: "on the international market since 2003"; ISO 27001:2022 certified; ProZorro.Sale enabled sales worth more than USD 700 million and ProZorro helped save USD 2 billion through reverse auctions; auctions developed for "over 5 years". Other pages say differently: the auctions page claims "9 years of experience", the services index "over 16 years of experience", and the footer reads "© 2002-2026". No team size, office or leadership is shown.Source
    • HiringBusiness Analyst vacancy: "Gathering and analyzing data in support of business cases, proposed projects, and system requirements" and "Managing reports". Project manager vacancy: "Monitoring progress and managing reports" and "Being responsible for resources and time planning". Both list public procurement experience as a plus.Source

Big bets 3

Larger plays that can change the trajectory.

  • 19

    Quintagroup helped put the Open Contracting Data Standard into the core of a national procurement system and builds business intelligence for public procurement. Open contracting data is published by many governments and is free to analyse. Few engineering companies can read it as fluently, and none of Quintagroup's content uses it.

    High impactHigh effort First result in 2 quarters CEO

    First action

    Pick five countries with good open contracting data and compute three measures for each as a pilot.

    How to measure it

    Conversations with procurement agencies started by the report

    Now
    –
    Target
    15 in the first year
    By
    12 months

    Steps

    1. 1

      Pilot five countries

      Three measures each, to test data quality.

      Business Analyst 3 weeks

    2. 2

      Fix the method

      Definitions, exclusions and caveats, reviewed by an outside procurement expert.

      Business Analyst 2 weeks

    3. 3

      Build the pipeline

      Scheduled loads of public data into one warehouse.

      Senior Python Developer 1 month

    4. 4

      Build the dashboard

      Public, embeddable, with country pages.

      Senior Python Developer 1 month

    5. 5

      Write the report

      Ten findings, each with a chart.

      Digital Marketing Manager 3 weeks

    6. 6

      Launch

      Briefings for agencies, transparency organisations and trade press.

      CEO 2 weeks

    7. 7

      Follow up

      Offer each country's agency a private walkthrough of its own figures.

      Head of Sales 1 quarter

    First result
    2 quarters
    Cost
    About three person-months plus design
    Owner
    CEO

    Smallest test

    Publish one blog post comparing single-bid rates in three countries and see who shares it.

    Why it works

    Publish a yearly report that compares public procurement systems on measures that can be computed from open data: bidders per tender, share of single-bid tenders, time from notice to award, savings in reverse auctions. Include a method note, an interactive dashboard and country pages. Send it to procurement agencies, development programmes and the press before sales conversations.

    Original data earns links, press and invitations in a small, relationship-driven market, and it demonstrates the exact analytics skill the company sells.

    Risks and how to handle them

    • A ranking embarrasses a potential client

      Present measures and trends, not a league table, and share figures with agencies before publishing.

    • Data quality differs by country

      Publish the caveats and exclude countries with incomplete data.

    Evidence

    • Your websiteE-procurement: the open-source OpenProcurement toolkit, an e-Auction module, a test stand and OCDS ("Open Contracting Data Standard was incorporated in the ProZorro system's core"); projects ProZorro, Prozorro.sale, Rialto and MTender; nine brochures (OpenProcurement, ProZorro and a short version) each in English, French and Italian, while the page itself is in English; "Request a demo" links to the general /contact form.Source
    • Your websiteAbout us: "on the international market since 2003"; ISO 27001:2022 certified; ProZorro.Sale enabled sales worth more than USD 700 million and ProZorro helped save USD 2 billion through reverse auctions; auctions developed for "over 5 years". Other pages say differently: the auctions page claims "9 years of experience", the services index "over 16 years of experience", and the footer reads "© 2002-2026". No team size, office or leadership is shown.Source
  • 21

    Among the example agents on Quintagroup's AI pages is one that checks bid papers against procurement rules and vendor registries and produces an exception report. Every other AI agency can describe an invoice or support agent. Very few can credibly build a procurement one, and Quintagroup's toolkit already holds the structured tender data such an agent needs.

    High impactHigh effort First result in 2 quarters CEO

    First action

    Ask two existing procurement clients which bid checks take their reviewers the most time.

    How to measure it

    Paid pilots signed

    Now
    –
    Target
    2
    By
    3 quarters

    Steps

    1. 1

      Interview reviewers

      Which checks, how long, which errors recur.

      Business Analyst 3 weeks

    2. 2

      Scope version one

      Completeness and formal conformity only; no scoring or ranking of bids.

      Business Analyst 2 weeks

    3. 3

      Build on test data

      Using the toolkit's test environment and public tender documents.

      Senior Python Developer 6 weeks

    4. 4

      Design the review screen

      Each flag shows the rule, the bid passage and a confirm or dismiss action; all actions logged.

      UI/UX Designer 2 weeks

    5. 5

      Run one pilot

      Side by side with human review; measure hours and missed issues.

      Project Manager 2 months

    6. 6

      Write the product page

      Scope, safeguards, pilot result and a fixed pilot price.

      Digital Marketing Manager 2 weeks

    7. 7

      Take it to the market

      Existing procurement contacts first, then tenders found by the monitor in idea 7.

      Head of Sales 1 quarter

    First result
    2 quarters
    Cost
    A small team for one to two quarters
    Owner
    CEO

    Smallest test

    Run the agent on 20 closed public tenders and compare its flags with the published outcomes.

    Why it works

    Turn that example into a named, scoped offer: a bid compliance review agent that plugs into OpenProcurement-based and other e-procurement systems, checks each submitted bid for completeness and rule conformity, and hands a reviewer an exception list with the source clause for each flag. The evaluation decision always stays with the procurement officer. Pilot it with one existing procurement client, measure reviewer hours per tender before and after, and sell it as a fixed-price pilot.

    A vertical product sells on domain trust, which is the one thing larger AI-first rivals cannot copy from Quintagroup quickly. It also gives the AI service line and the procurement line the same story.

    Risks and how to handle them

    • Procurement law requires human decisions and audit trails

      The agent only flags; a named officer decides, and every flag and action is logged.

    • Rules differ by jurisdiction

      Start with one client's rule set and make rules configurable.

    Builds on

    Evidence

    • Your websiteAI section: /services/ai lists seven pages (consulting, development, generative, self-hosted platform, AI services, agentic AI solutions, agentic AI development); the hub and both agentic pages cite no case study. The two agentic pages disagree: one gives the "Agentic AI Deployment Framework (AADF)" five phases, 3 to 4 weeks for one workflow and 6 to 10 for multi-agent; the other gives six phases, "4-6 weeks" and "6-8 weeks". Example agents include invoice reconciliation, a compliance review of bid papers against procurement rules, and support resolution.Source
    • Your websiteE-procurement: the open-source OpenProcurement toolkit, an e-Auction module, a test stand and OCDS ("Open Contracting Data Standard was incorporated in the ProZorro system's core"); projects ProZorro, Prozorro.sale, Rialto and MTender; nine brochures (OpenProcurement, ProZorro and a short version) each in English, French and Italian, while the page itself is in English; "Request a demo" links to the general /contact form.Source
  • 23

    Quintagroup has three assets that point at the same channel: an Odoo development practice, a service page for Odoo auction software, and an Atreus integration with Odoo CRM. Companies that run Odoo look for add-ons in Odoo's own app marketplace and through Odoo partners, not on a development company's website.

    Medium impactHigh effort First result in 2 quarters Atreus Product Owner

    First action

    Search the Odoo app marketplace for auction modules and note prices, ratings and supported versions.

    How to measure it

    Module installs

    Now
    –
    Target
    100 installs and 3 partner deals
    By
    12 months

    Steps

    1. 1

      Survey the marketplace

      Existing auction modules, prices and gaps.

      Atreus Product Owner 1 week

    2. 2

      Scope the module

      The smallest useful auction flow inside Odoo, with Atreus behind it.

      Business Analyst 2 weeks

    3. 3

      Build and test

      For the two most used Odoo versions.

      Senior Python Developer 2 months

    4. 4

      Write the listing

      Screenshots, a video and the upgrade path.

      Digital Marketing Manager 1 week

    5. 5

      Recruit partners

      Five Odoo implementers with clients in procurement, liquidation or leasing.

      Head of Sales 1 quarter

    6. 6

      Review

      Installs, upgrades and partner deals after two quarters.

      CEO 2 quarters

    First result
    2 quarters
    Cost
    Two to three person-months plus marketplace fees
    Owner
    Atreus Product Owner

    Smallest test

    Publish a landing page for the module with a waiting list and show it to five Odoo partners.

    Why it works

    Package the Atreus connector, and a light sealed-bid or reverse auction module, as an Odoo app with a free tier that leads to Atreus Premium or a custom project. First check whether a current listing exists and how rival auction modules are priced and rated. Then recruit a handful of Odoo implementation partners who can resell it with a revenue share.

    A marketplace listing puts the product in front of buyers at the moment they search for the function, and partners bring sales reach a company of this size cannot staff.

    Risks and how to handle them

    • Each Odoo version needs maintenance

      Support two versions and keep the logic in Atreus, with a thin connector in Odoo.

    • Demand for auctions inside Odoo is small

      Run the marketplace survey and five partner conversations before building.

    Builds on

    Evidence

    • Atreus siteAtreus (atreus.auction), "Atreus SaaS Auction Platform": Free Trial with 10 auctions a month, Premium with 100 auctions a month at $499/month or $4,990/year, Enterprise on request with sealed bid, spectrum and real-time bidding; buttons "Try it free" and "Request a demo"; an Odoo CRM integration; the homepage read does not name Quintagroup or link to quintagroup.com, while Quintagroup's Atreus case links to the free registration.Source
    • Your websiteAuctions: English, reverse, Hybrid Dutch, multivariable, spectrum (SMR, clock, combinatorial) and real-time bidding, with sub-pages for sealed bid and Odoo auction software; demo button "Fill in this form to get a demo". The Auction Delivery Framework page says "Within one to two weeks of kickoff, Quintagroup provides a functional demo environment", discovery takes "three to five days" and production 4-6 weeks, and calls Atreus.auction "the working proof".Source

Wildcards 1

Unexpected experiments, each with a cheap way to test it.

  • 22

    The sitemap still carries a /portfolio section of 91 URLs and a /solutions section of 49, both last changed in 2019, with products such as intranet, conference management and learning content systems. The third page of case studies is Plone-era website work. Together they tell search engines and AI assistants that Quintagroup builds Plone websites, while the company now sells auctions, procurement systems and AI.

    Low impactLow effort First result in 2 months Digital Marketing Manager

    First action

    Export 12 months of search clicks for /portfolio, /solutions and the oldest 12 cases.

    How to measure it

    Old-section URLs reviewed and resolved

    Now
    –
    Target
    All 140
    By
    90 days

    Steps

    1. 1

      Export traffic

      Clicks and links per old URL.

      Digital Marketing Manager 1 day

    2. 2

      Sort

      Keep, update, archive or redirect.

      Digital Marketing Manager 3 days

    3. 3

      Update the keepers

      Dates, screenshots and a link to the current service.

      Digital Marketing Manager 1 week

    4. 4

      Create the archive

      Labelled, out of the sitemap and the menus.

      Senior Python Developer 2 days

    5. 5

      Redirect

      Dead ends to the closest live page.

      Senior Python Developer 1 day

    6. 6

      Recheck

      How AI assistants and search describe the company after two months.

      Digital Marketing Manager 2 months

    First result
    2 months
    Cost
    $0
    Owner
    Digital Marketing Manager

    Smallest test

    Archive the /solutions section only and watch for any drop in enquiries.

    Why it works

    Review the two old sections and the oldest cases by traffic. Keep and update anything that still earns visits or supports the Plone service, move the rest to a clearly labelled archive that is not in the sitemap, and redirect dead-end pages to the closest current service or case.

    What a site contains in bulk shapes how it is summarised. Trimming the old majority lets the newer pages define the company.

    Risks and how to handle them

    • Plone clients still find the company through these pages

      Keep and refresh any page with real traffic and link it to the Plone service.

    Builds on

    Evidence

    • Site researchSitemap (4,847 URLs, checked November 15, 2026): /services 3,290, /cms 967, /blog 349 (the blog lists 92 posts), /portfolio 91 and /solutions 49 (both last changed in 2019). It includes image wrapper pages ending in "/view", for example a blog image page whose heading is the file name "benefits-PDF-Word-Excel-generation.jpg" with "Click to view full-size image… Size: 649KB", and folders titled "Blog Images" and "Service Images".Source
    • Your websiteCases: 36 case studies on three pages of 12, with no filter by industry or technology. The two case pages read in full (Mastercard API enhancement and the Atreus auction platform) give no results, dates, team size, duration or client quote. The list runs from recent work (an insurance image-recognition service, "The Atlas Project" that collects auction and tender data with Python, Selenium and AWS) back to Plone-era websites and a Plone themes web store.Source

03

How Programmable can help

Where an idea in this playbook is something Programmable builds or runs, it's listed here with what we'd do. Everything else is yours to run.

Have it built

ProgrammableAgency

1 of the 23 ideas are things Programmable builds and runs for its customers: pages, outreach, prospect lists, booking flows and monitors. Pick one, or send the whole list.

Prospect lists1The list compiled from public sources, verified, deduplicated and delivered to your CRM.

A 30-minute session with the engineer who would build it, not a sales call. You leave with a plan whether or not you hire us.

Book a working session

Usually live within days.

More example playbooks and reviews

More growth playbooks

Other companies we have researched, each read from what they are hiring for.

  • Hover growth strategyTurns a smartphone scan or eight photos of a home into a measured 3D model, then into designs with real products, estimates, proposals and…22 growth ideas
  • Summer growth strategyA public benefit corporation that helps people lower and forgive student debt and plan for college costs: it matches borrowers to 150+…23 growth ideas
  • Dudis Design growth strategyHandmade, made-to-order baby and toddler clothing: lounge sets, rompers, leggings, bummies, headbands and accessories in hundreds of…21 growth ideas
  • Adult Toy Box growth strategyAn online store for adult intimate products, lingerie, costumes and related care items from established brands, shipped in plain packaging…24 growth ideas
  • Uber Freight growth strategyLogistics technology and services for shippers: managed transportation (3PL and 4PL), freight capacity across truckload, LTL, intermodal…22 growth ideas
  • InHawk IT Solutions growth strategySoftware for colleges and universities: student and campus management, digital evaluation of exams, OMR services, fund-based accounting…22 growth ideas
  • Shive-Hattery growth strategyA privately held architecture and engineering firm founded in 1895: architecture, interiors, civil, structural, mechanical, electrical…20 growth ideas
  • Planful growth strategyA financial performance management platform for planning, budgeting, close, consolidation and reporting, with Planful AI (Analyst…23 growth ideas

See every playbook and website review

Independent research. Programmable wrote this playbook from public sources, as they stood on November 16, 2026, and cites them throughout. Quintagroup did not commission, review or endorse it, and Programmable is not affiliated with Quintagroup. Company and product names are trademarks of their owners, used only to identify them. The ideas are suggestions and opinion, not professional, legal or financial advice, and not statements about how Quintagroup runs its business.

Something wrong, or do you represent Quintagroup? Email [email protected] and we will correct it or take it down.

30/60/90 roadmap

What to focus on in each month, and the ideas that carry it.

  1. Days 1 to 308 ideas

    The estimate desk, one visitor-identification tool with account briefs, case-study results, the auction sandbox, consistent numbers, the sitemap, the homepage message and direct call booking.

    Programmable can ship 1 of these in days.

  2. Days 31 to 6012 ideas

    Blog refresh pipeline, tender monitoring, review requests, Atreus cross-linking and trial follow-ups, published price ranges, the first AI case study, the weekly growth report, French and Italian pages, the questionnaire library and the old-page clean-up.

  3. Days 61 to 903 ideas

    Commit to the big bets: the open procurement data report, the bid compliance agent as a packaged product, and an Odoo auction listing.

The research behind the plan

3 research areas, each summarised in a few lines. Open any of them for the full detail.

Customers

What customers say

Customers value deep auction and e-procurement know-how proven on national systems. The most cited problem: no price or timeline until a form is answered.

  • Affordable senior Python, Django and Plone engineering
  • Hard to judge results because case studies have no numbers or dates
  • In their words: “reverse auction”

Market

Competitive landscape

3 competitors compared: STX Next, Django Stars and Auction Software.

  • Gap to exploit: Agentic AI built specifically for procurement and auction workflows
  • STX Next: No auction or public procurement specialism
  • Django Stars: Generalist across fintech, logistics, healthcare and travel

Hiring signals

The careers page lists six vacancies without posting dates, so read them as direction rather than urgency.

  • Digital Marketing Manager: One person for campaigns, SEO, social, brochures and analysis. Ideas 2, 4, 16.
  • Business Analyst: Business cases, proposed projects and requirements; procurement experience a plus. Ideas 1, 20, 21.
  • Project manager: Progress monitoring and reports. Idea 10.

From the team that wrote this playbook

All products
  • ProgrammableAgencyWorking AI systems in days, not quarters (opens programmable.agency in a new tab)
  • ProgrammableInfoAn agent that finds, researches and asks first (opens programmable.info in a new tab)
  • ProgrammableDigitalScreenshots in, an animated product demo out (opens programmable.digital in a new tab)
  • ProgrammableWatchKnow the day a supplier falls out of compliance (opens programmable.watch in a new tab)
  • ProgrammableTalkGet registrants for your events, and know who’s coming (opens programmable.talk in a new tab)
  • ProgrammablePageText what you need, your customer gets a page that works (opens programmable.page in a new tab)
  • ProgrammableCommunityForums with labelled AI members, never empty (opens programmable.community in a new tab)
ProgrammableMarketing
ExamplesLeaderboardsProductsTermsPrivacyCookiesDPALegalContact

© 2026 AI-Powered Growth Intelligence