ProgrammableMarketing/cinder.ai (opens in a new tab)
PDF

Views

  • Plan
  • Roadmap
  • Research

On this page

  • The brief
  • Start here
  • Idea portfolio
  • Effort vs impact
  • This week · 6
  • This quarter · 10
  • Big bets · 4
  • Wildcards · 2
  • How we can help
Generated Oct 23, 2026
  1. Programmable
  2. Playbooks
  3. Cinder
Cinder logo

growth playbook · Oct 23, 2026

cinder.ai (opens in a new tab)

Cinder growth playbook

Diagnosis

Cinder raised a $41 million Series B in May 2026, plans to double headcount, and is hiring its first AI Growth Marketer to reach generative AI buyers who do not yet know they have a trust and safety problem.

It already has the raw material: original red-team tests of open-weight models, strong case studies from Krea, Black Forest Labs and Character.AI, and evening events in New York.

The gaps are in how that material reaches buyers: the research has no home and one post shows raw developer messages where its tables belong, the homepage loads a 13.9 MB video, the demo form has no option for a launch, and the YouTube channel has 14 subscribers. This playbook gives the research a series, a fast-response routine and an index, speeds up the site, sharpens the story for three kinds of AI buyer and adds launch-timed outreach, events and a self-serve door.

Growth Score

7/10

Biggest Opportunity

Become the source everyone checks when an open-weight model or AI product launches, with published tests, a release checklist and a standing index.

Biggest Risk

Alice (formerly ActiveFence) and other well-funded vendors are moving into AI safety and red teaming, so Cinder must turn its research lead into pipeline quickly.

01

Start here

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

  1. 1

    Make the open-weight red-team tests a research series buyers can find, cite and download, starting with the two missing tables

    Render the two tables in the Alibaba post today and check the other posts for the same message.

    High impact First result in 1 month
  2. 2

    Stop the homepage downloading a 13.9 MB video before buyers can read it

    Switch the hero video to preload="none" with the poster image.

    High impact First result in 2 weeks
  3. 3

    Add "launching a model or AI product" to the demo form and route it to a red-team call

    Add the two options and the date field.

    High impact First result in 2 weeks

02

Your idea portfolio

22 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.

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 · 4Big bets · 10Fill-ins · 3Reconsider · 5
Effort →

Showing 22 of 22 ideas

This week 6

Quick wins with a clear first action.

  • 01

    Cinder's new AI Growth Marketer is asked for "original research or benchmarks that set Cinder apart", and Cinder already has them: the October 21 test of six open-weight models (four gave harmful help in over 40% of responses) and the August test of an abliterated Alibaba model (23.1% violation rate across 2,670 attacks). But the results sit in the same list as hiring posts, end only in a newsletter box, and in the Alibaba post the two places where tables belong show a raw "Unknown block type tableBlock" developer message instead.

    High impactLow effort First result in 1 month AI Growth Marketer

    First action

    Render the two tables in the Alibaba post today and check the other posts for the same message.

    How to measure it

    Red-team scoping calls from research

    Now
    –
    Target
    10 a month
    By
    90 days

    Steps

    1. 1

      Fix the template

      Add the table component and re-publish the Alibaba post.

      Web developer 1 day

    2. 2

      Check every post

      Search all 59 posts for unrendered blocks.

      Web developer 1 day

    3. 3

      Build /research

      One card per test: model, date, method, headline rate.

      Senior Product Designer 1 week

    4. 4

      Make the download

      One-page PDF and category list per test, behind a short form.

      AI Growth Marketer 1 week

    5. 5

      Add the call to action

      "Test your model" books a scoping call with sales.

      Head of Sales 2 days

    6. 6

      Brief press and labs

      Send the series page to safety reporters and model teams.

      AI Growth Marketer 1 week

    7. 7

      Set a cadence

      One new test each month, announced in the newsletter.

      Red team lead monthly

    8. 8

      Measure

      Downloads, calls booked and pipeline by test.

      AI Growth Marketer monthly

    First result
    1 month
    Cost
    Staff time only
    Owner
    AI Growth Marketer

    Smallest test

    Fix the tables and add one gated PDF to the Alibaba post this week.

    Why it works

    Fix the table block in the blog template and re-check every post. Then give the research its own home: a /research page that lists each test with the model, date, method, headline rate and a chart; a one-page PDF and the prompt categories (not the prompts) behind a short form; and a "test your model" call to action that books a red-team scoping call. Publish each new test as part of the series, so press, AI labs and safety teams start to expect the next one.

    Generative AI buyers trust numbers from tests they could run themselves. A named, recurring series is what makes a one-off blog post into a demand asset, and the form turns readers into the pipeline the role is measured on.

    Ready to use

    ChecklistResearch series launch list

    [ ] Table block renders in the blog template
    [ ] No raw developer messages in any post
    [ ] /research page with one card per test
    [ ] Method, date and model on every card
    [ ] One-page PDF per test behind a short form
    [ ] "Test your model" call to action to a scoping call
    [ ] Newsletter issue for each new test
    [ ] Pipeline tracked by test

    Risks and how to handle them

    • Model makers object to results

      Publish the method, share findings with the model team before release, and stick to measured rates.

    • Forms cut readership

      Keep the article open; gate only the PDF and category list.

    Evidence

    • Your websiteRed-team post "What Cinder Learned Red Teaming Alibaba's Latest Open-Weight Model" (August 24, 2026): across 2,670 attacks the original model generated NCII once and the abliterated version violated 23.1% of the time. Two places in the post where tables belong show the text "[@portabletext/react] Unknown block type "tableBlock", specify a component for it in the `components.types` prop" instead (checked twice on October 22, 2026). The post ends with a newsletter sign-up; there is no download of the data or method.Source
    • Your website"How Leading Open-Weight Models Respond to Requests Involving Violent Harm" (October 21, 2026): six open-weight models tested against 209 adversarial requests built from 70 goals, 18 attack techniques and 15 harm domains; "Four of the six gave harmful, actionable help in more than 40% of their responses, and the worst reached 56%. Even the safest gave unsafe responses about 15% of the time."Source
    • HiringAI Growth Marketer duties: lift the demand-gen performance of case studies ("One of our current case studies is already one of our best-performing assets... good enough that customers ask to be featured"); turn the blog, newsletter and founder ghostwriting into "an explosive growth channel" with "original research or benchmarks that set Cinder apart"; decide which events to attend, sponsor or skip; "Spot and act on relevant news and industry moments quickly"; own "attribution, lead scoring, and inbound reporting"; and partner with sales "so marketing-sourced leads actually convert".Source
    • HiringAI Growth Marketer requirements: demand generation "for an AI product specifically" is "the requirement we care about most"; built from zero; experience in a category "where the downside of failure is existential (cybersecurity, identity and authentication, fintech risk, regtech...)". Nice to have: "You've built a research report, index, or open benchmark that became a demand asset", "ABM or intent-data tools", "Field marketing at conference scale", and marketing into "social platforms, marketplaces, dating apps, gaming, fintech apps, or B2C-facing AI products".Source
    • Your websiteResources hub: 65 items (Insights 37, Product 10, Engineering 7, News 3, Case studies 4); a newsletter sign-up "sent only when we have something worth your attention". The red-team research sits in the same list as hiring and culture posts, with no research or benchmark section.Source
  • 02

    Google's phone test shows the homepage weighs 15.7 MB and the main content takes 14.9 seconds to appear; the largest element is a 13.9 MB hero video set to preload in full. Real visitors fail Core Web Vitals too (LCP 4.0 s, INP 286 ms), and the trend is getting slower. Many of the buyers the growth marketer is chasing will first see Cinder on a phone from a link in a post or a newsletter.

    High impactLow effort First result in 2 weeks Web developer

    First action

    Switch the hero video to preload="none" with the poster image.

    How to measure it

    Field LCP on phones

    Now
    –
    Target
    under 2.5 s
    By
    60 days

    Steps

    1. 1

      Change preload

      Poster first, video after load.

      Web developer 1 hour

    2. 2

      Compress

      Shorter loop, lower bitrate, modern codec.

      Senior Product Designer 1 day

    3. 3

      Skip on phones

      Static image on small screens.

      Web developer 2 hours

    4. 4

      Audit other videos

      Every autoplay video on the site.

      Web developer 1 day

    5. 5

      Re-test

      PageSpeed on homepage and solution pages.

      Web developer 1 hour

    6. 6

      Watch field data

      Chrome UX Report each month.

      AI Growth Marketer monthly

    First result
    2 weeks
    Cost
    $0
    Owner
    Web developer

    Smallest test

    Change preload on the hero video today and re-run PageSpeed.

    Why it works

    Show the poster image first and load a compressed, shorter video only after the page is interactive, or not at all on phones. Serve the video in a modern codec at a much lower bitrate, apply the same rule to the other autoplay videos across the site, and re-test the homepage and solution pages.

    The first screen of a security vendor's site is part of its credibility, and faster pages convert more of the paid and social traffic the new role will buy.

    Risks and how to handle them

    • Less motion on the homepage

      Keep the product animation as a lighter loop or a still.

    Evidence

    • Google PageSpeedMobile test of the homepage (October 22, 2026): performance 57/100, largest content paint 14.9 s in the lab, page weight 15.7 MB over 97 requests. The largest element is the hero video, a 13.9 MB MP4 set to preload="auto". Real visitors (Chrome UX Report) fail Core Web Vitals: LCP 4.0 s and INP 286 ms, and the origin's LCP trend is slower.Source
    • Your websiteHomepage: H1 "Deploy AI to Fight Abuse"; a site-wide banner "Deploy open-weight models with frontier safety built in"; proof "3B+ end-users protected", "94% of human review, automated", "100k+ hours saved", "400M+ events processed daily"; a Synthesia CEO quote; one call to action, "Get a demo".Source
    • HiringAI Growth Marketer (New York, posted September 25, 2026, reports to the Head of Sales): "Most companies building with generative AI don't know yet that they have a trust and safety problem... Right now, that story lives only with our current customers." The job is to take existing positioning on "AI red teaming, agentic moderation, and content integrity" and turn it into "a story that lands with a generative AI buyer specifically", reaching "a market that doesn't buy the way traditional enterprise software buyers do". It owns marketing-sourced pipeline, with the target set with the Head of Sales, and works with the Founding Marketer.Source
  • 03

    The demo form offers four reasons: evaluating platforms, replacing a tool, researching, other. None fits the generative AI buyer the growth marketer is hired to reach, who usually arrives with a launch date and a red-team need, not a platform evaluation.

    High impactLow effort First result in 2 weeks Head of Sales

    First action

    Add the two options and the date field.

    How to measure it

    Launch-tagged demos

    Now
    –
    Target
    25% of inbound
    By
    60 days

    Steps

    1. 1

      Add options

      Two new reasons and a launch date.

      Web developer 2 hours

    2. 2

      Set routing

      Launch leads to a scoping call queue.

      Head of Sales 1 day

    3. 3

      Auto follow-up

      Research post and Krea or Black Forest Labs case study.

      AI Growth Marketer 1 day

    4. 4

      Score

      Launch within 90 days gets priority.

      AI Growth Marketer 1 day

    5. 5

      Review

      Close rates by reason each month.

      Head of Sales monthly

    6. 6

      Adjust

      Merge or split reasons as data shows.

      AI Growth Marketer quarterly

    First result
    2 weeks
    Cost
    $0
    Owner
    Head of Sales

    Smallest test

    Add the options today and watch a week of submissions.

    Why it works

    Add the options "Launching a model or AI feature" and "Red teaming or safety evals", with a launch-date field. Route those to a red-team scoping call within a day and send the relevant research post and case study automatically.

    Launch dates create urgency, and asking for one lets sales prioritize the leads with a deadline.

    Risks and how to handle them

    • More fields lower completion

      Make the date optional.

    Evidence

    • Your websiteDemo form at /contact asks for team size and a reason for reaching out with four options: "Evaluating platforms", "Replacing existing tool", "Researching the space", "Other". There is no option for a model or product launch, or for red teaming.Source
    • HiringAI Growth Marketer (New York, posted September 25, 2026, reports to the Head of Sales): "Most companies building with generative AI don't know yet that they have a trust and safety problem... Right now, that story lives only with our current customers." The job is to take existing positioning on "AI red teaming, agentic moderation, and content integrity" and turn it into "a story that lands with a generative AI buyer specifically", reaching "a market that doesn't buy the way traditional enterprise software buyers do". It owns marketing-sourced pipeline, with the target set with the Head of Sales, and works with the Founding Marketer.Source
    • Your websiteGenerative AI solution page: H1 "SHIP FASTER. WITHOUT SHIPPING RISK."; covers AI red teaming, prompt and output screening, QA and custom safety agents; proof ">97% AI Moderation Accuracy", "< 1% NSFW rate across evaluated generations", ">90% Reduction in CSAM and NCII vulnerability", "10X Safer than benchmark industry models at launch"; the only call to action is "Get a demo".Source
  • 04

    Of the pages checked, only the switch and content moderation services pages have a canonical tag, and only one page has structured data. Cinder publishes research meant to be cited by search engines and AI assistants.

    Medium impactLow effort First result in 1 month Web developer

    First action

    Add canonicals in the page template.

    How to measure it

    Indexed pages

    Now
    –
    Target
    100%
    By
    30 days

    Steps

    1. 1

      Canonicals

      Template-level.

      Web developer 2 hours

    2. 2

      Article markup

      Author and date on posts.

      Web developer 1 day

    3. 3

      Organization markup

      Homepage.

      Web developer 1 hour

    4. 4

      FAQ markup

      Pages with FAQs.

      Web developer 2 hours

    5. 5

      Resubmit

      Sitemap in Search Console.

      Web developer 1 hour

    6. 6

      Check

      Coverage report after two weeks.

      AI Growth Marketer 2 weeks

    First result
    1 month
    Cost
    $0
    Owner
    Web developer

    Smallest test

    Ship canonicals this week.

    Why it works

    Add self-referencing canonical tags site-wide, Organization and Article markup on posts (author, date), and FAQ markup where FAQs exist. Re-submit the sitemap and check coverage in Search Console.

    Clear signals help search engines and AI assistants credit the right page when they cite Cinder's findings.

    Risks and how to handle them

    • Template change breaks pages

      Test on staging first.

    Evidence

    • Your websiteSitemap (100 URLs, checked October 22, 2026): 59 blog posts, 5 case studies (Black Forest Labs, Character.AI, Krea, Synthesia, Zello; the Resources page lists only the first four), 6 solution pages, 13 product pages, 2 switch pages (Community Sift, Unitary), an /llms.txt file, no pricing page and no RSS feed. Of the pages checked, only the switch and content moderation services pages carry a canonical tag, and only the content moderation services page has structured data.Source
    • Your websiteRed-team post "What Cinder Learned Red Teaming Alibaba's Latest Open-Weight Model" (August 24, 2026): across 2,670 attacks the original model generated NCII once and the abliterated version violated 23.1% of the time. Two places in the post where tables belong show the text "[@portabletext/react] Unknown block type "tableBlock", specify a component for it in the `components.types` prop" instead (checked twice on October 22, 2026). The post ends with a newsletter sign-up; there is no download of the data or method.Source
    • Your website"How Leading Open-Weight Models Respond to Requests Involving Violent Harm" (October 21, 2026): six open-weight models tested against 209 adversarial requests built from 70 goals, 18 attack techniques and 15 harm domains; "Four of the six gave harmful, actionable help in more than 40% of their responses, and the worst reached 56%. Even the safest gave unsafe responses about 15% of the time."Source
  • 05

    The sitemap has five case studies but the Resources page shows four; Zello is missing. Cinder sells to six segments (generative AI, foundation models, marketplaces, gaming, social and dating, creator economy), but most stories are AI companies.

    Medium impactLow effort First result in 2 weeks Founding Marketer

    First action

    Add Zello to the Resources page.

    How to measure it

    Case study clicks

    Now
    –
    Target
    +30%
    By
    60 days

    Steps

    1. 1

      Add Zello

      Resources page.

      Web developer 1 hour

    2. 2

      Tag stories

      Segment and product.

      Founding Marketer 1 day

    3. 3

      Place on pages

      One per solution page.

      Web developer 1 day

    4. 4

      Find gaps

      Segments without a story.

      Founding Marketer 1 day

    5. 5

      Ask customers

      One per missing segment.

      Customer success 1 month

    6. 6

      Measure

      Clicks and demo requests.

      AI Growth Marketer monthly

    First result
    2 weeks
    Cost
    Staff time only
    Owner
    Founding Marketer

    Smallest test

    List Zello today.

    Why it works

    List Zello with the others, tag each case study by segment and product, and show the right one on each solution page. Note which segments have no story and ask current customers in those segments first.

    Buyers look for a peer, and a missing story is lost proof that already exists.

    Risks and how to handle them

    • Customers decline to be named

      Offer an anonymized story with the numbers.

    Evidence

    • Your websiteSitemap (100 URLs, checked October 22, 2026): 59 blog posts, 5 case studies (Black Forest Labs, Character.AI, Krea, Synthesia, Zello; the Resources page lists only the first four), 6 solution pages, 13 product pages, 2 switch pages (Community Sift, Unitary), an /llms.txt file, no pricing page and no RSS feed. Of the pages checked, only the switch and content moderation services pages carry a canonical tag, and only the content moderation services page has structured data.Source
    • HiringAI Growth Marketer duties: lift the demand-gen performance of case studies ("One of our current case studies is already one of our best-performing assets... good enough that customers ask to be featured"); turn the blog, newsletter and founder ghostwriting into "an explosive growth channel" with "original research or benchmarks that set Cinder apart"; decide which events to attend, sponsor or skip; "Spot and act on relevant news and industry moments quickly"; own "attribution, lead scoring, and inbound reporting"; and partner with sales "so marketing-sourced leads actually convert".Source
    • HiringSenior Product Manager, Trust & Safety (posted July 23, 2026): Cinder is "an AI-native safety platform serving over thirty customers, including some of the largest consumer platforms in the world. Many of them serve tens or hundreds of millions of users"; many users "are domain experts rather than technical ones".Source
  • 06

    The growth marketer must "spot and act on relevant news and industry moments quickly", and the open-weight page says releases "compress the timeline between model availability and real-world misuse". Each release is a news moment Cinder is equipped to answer.

    High impactMedium effort First result in 1 quarter AI Growth Marketer

    First action

    Write the kit and pick the next expected release.

    How to measure it

    Earned mentions

    Now
    –
    Target
    4 a quarter
    By
    2 quarters

    Steps

    1. 1

      Build the kit

      Harness, categories, judge, chart.

      Red team lead 2 weeks

    2. 2

      Watch releases

      Model hubs and lab announcements.

      AI Growth Marketer ongoing

    3. 3

      Run the test

      Within 48 hours.

      Red team lead 2 days

    4. 4

      Publish

      Short note with method.

      AI Growth Marketer 1 day

    5. 5

      Pitch

      Safety reporters and newsletters.

      AI Growth Marketer 1 day

    6. 6

      Share

      Founder posts and the newsletter.

      CEO 1 day

    First result
    1 quarter
    Cost
    Red team time
    Owner
    AI Growth Marketer

    Smallest test

    Run the kit on the next notable release.

    Why it works

    Keep a standing test kit (prompt categories, harness, judge, chart template). Within 72 hours of a notable open-weight release, publish a short note with the headline rate and what self-hosting teams should do, then pitch it to reporters and post it in the founders' channels.

    Being first with numbers makes Cinder the source journalists and model teams call, and each note feeds the research series.

    Risks and how to handle them

    • Rushed results are wrong

      Publish only rates the team has checked, and label notes as early findings.

    Evidence

    • HiringAI Growth Marketer duties: lift the demand-gen performance of case studies ("One of our current case studies is already one of our best-performing assets... good enough that customers ask to be featured"); turn the blog, newsletter and founder ghostwriting into "an explosive growth channel" with "original research or benchmarks that set Cinder apart"; decide which events to attend, sponsor or skip; "Spot and act on relevant news and industry moments quickly"; own "attribution, lead scoring, and inbound reporting"; and partner with sales "so marketing-sourced leads actually convert".Source
    • Your websiteOpen-weight model safety page: pre-release red teaming, safety eval operations, guardrail loops and "Release governance"; "Open-weight releases compress the timeline between model availability and real-world misuse." The Trust Center (Vanta) lists SOC 2 and HIPAA.Source
    • Your website"How Leading Open-Weight Models Respond to Requests Involving Violent Harm" (October 21, 2026): six open-weight models tested against 209 adversarial requests built from 70 goals, 18 attack techniques and 15 harm domains; "Four of the six gave harmful, actionable help in more than 40% of their responses, and the worst reached 56%. Even the safest gave unsafe responses about 15% of the time."Source
    • Your websiteRed-team post "What Cinder Learned Red Teaming Alibaba's Latest Open-Weight Model" (August 24, 2026): across 2,670 attacks the original model generated NCII once and the abliterated version violated 23.1% of the time. Two places in the post where tables belong show the text "[@portabletext/react] Unknown block type "tableBlock", specify a component for it in the `components.types` prop" instead (checked twice on October 22, 2026). The post ends with a newsletter sign-up; there is no download of the data or method.Source

This quarter 10

The core plays for the next three months.

  • 07

    The role says the story must land "with a generative AI buyer specifically". Cinder has one generative AI page and one foundation models page, but labs releasing weights, apps generating media and teams shipping agents worry about different failures.

    High impactMedium effort First result in 1 quarter AI Growth Marketer

    First action

    Interview three customers, one per type.

    How to measure it

    Page-to-demo rate

    Now
    –
    Target
    +25%
    By
    1 quarter

    Steps

    1. 1

      Interview

      One customer per buyer type.

      AI Growth Marketer 2 weeks

    2. 2

      Write

      Three pages with proof.

      Founding Marketer 2 weeks

    3. 3

      Design

      Shared layout.

      Senior Product Designer 1 week

    4. 4

      Match decks

      Three sales deck versions.

      Head of Sales 1 week

    5. 5

      Promote

      Paid social to each audience.

      AI Growth Marketer 1 month

    6. 6

      Compare

      Demo rate by page.

      AI Growth Marketer monthly

    First result
    1 quarter
    Cost
    Staff time and a small ad budget
    Owner
    AI Growth Marketer

    Smallest test

    Publish the model-release page first.

    Why it works

    Write three short pages: model releases (red teaming and release governance), generative media apps (NSFW, NCII and CSAM screening, Krea and Black Forest Labs proof), and agent products (prompt injection, tool-call screening). Give each its own proof and call to action, and use the same three stories in sales decks.

    Buyers act when a page describes their exact risk, and three narrow stories are easier to test than one broad one.

    Risks and how to handle them

    • Too many pages

      Keep one parent page linking the three.

    Evidence

    • HiringAI Growth Marketer (New York, posted September 25, 2026, reports to the Head of Sales): "Most companies building with generative AI don't know yet that they have a trust and safety problem... Right now, that story lives only with our current customers." The job is to take existing positioning on "AI red teaming, agentic moderation, and content integrity" and turn it into "a story that lands with a generative AI buyer specifically", reaching "a market that doesn't buy the way traditional enterprise software buyers do". It owns marketing-sourced pipeline, with the target set with the Head of Sales, and works with the Founding Marketer.Source
    • Your websiteGenerative AI solution page: H1 "SHIP FASTER. WITHOUT SHIPPING RISK."; covers AI red teaming, prompt and output screening, QA and custom safety agents; proof ">97% AI Moderation Accuracy", "< 1% NSFW rate across evaluated generations", ">90% Reduction in CSAM and NCII vulnerability", "10X Safer than benchmark industry models at launch"; the only call to action is "Get a demo".Source
    • Your websiteOpen-weight model safety page: pre-release red teaming, safety eval operations, guardrail loops and "Release governance"; "Open-weight releases compress the timeline between model availability and real-world misuse." The Trust Center (Vanta) lists SOC 2 and HIPAA.Source
    • Your websiteHomepage: H1 "Deploy AI to Fight Abuse"; a site-wide banner "Deploy open-weight models with frontier safety built in"; proof "3B+ end-users protected", "94% of human review, automated", "100k+ hours saved", "400M+ events processed daily"; a Synthesia CEO quote; one call to action, "Get a demo".Source
  • 08

    The role wants case studies so good "customers ask to be featured". Krea and Black Forest Labs released models after Cinder's testing. A public mark of that work gives customers something they want to show.

    High impactMedium effort First result in 2 quarters Founding Marketer

    First action

    Draft the badge rules and ask one recent customer.

    How to measure it

    Public marks

    Now
    –
    Target
    6
    By
    2 quarters

    Steps

    1. 1

      Set rules

      What qualifies and what is claimed.

      Red team lead 2 weeks

    2. 2

      Design

      Badge and summary page.

      Senior Product Designer 1 week

    3. 3

      Legal review

      Wording of claims.

      Legal 1 week

    4. 4

      Offer

      To recent customers.

      Customer success 1 month

    5. 5

      Publish

      Summary and case study.

      Founding Marketer ongoing

    6. 6

      Track

      Referral traffic and demos.

      AI Growth Marketer monthly

    First result
    2 quarters
    Cost
    Design time
    Owner
    Founding Marketer

    Smallest test

    Offer the mark to one customer.

    Why it works

    Offer customers who complete a red-team engagement a badge and a short public summary (what was tested, the headline improvement) for their model card or launch post, linking to a page about the engagement. Turn each into a short case study.

    Customers get a trust signal for their own launch; Cinder gets a link and a reference every time.

    Risks and how to handle them

    • Badge read as a guarantee

      State the scope and date of testing.

    Evidence

    • HiringAI Growth Marketer duties: lift the demand-gen performance of case studies ("One of our current case studies is already one of our best-performing assets... good enough that customers ask to be featured"); turn the blog, newsletter and founder ghostwriting into "an explosive growth channel" with "original research or benchmarks that set Cinder apart"; decide which events to attend, sponsor or skip; "Spot and act on relevant news and industry moments quickly"; own "attribution, lead scoring, and inbound reporting"; and partner with sales "so marketing-sourced leads actually convert".Source
    • Your websiteGenerative AI solution page: H1 "SHIP FASTER. WITHOUT SHIPPING RISK."; covers AI red teaming, prompt and output screening, QA and custom safety agents; proof ">97% AI Moderation Accuracy", "< 1% NSFW rate across evaluated generations", ">90% Reduction in CSAM and NCII vulnerability", "10X Safer than benchmark industry models at launch"; the only call to action is "Get a demo".Source
    • Your websiteSitemap (100 URLs, checked October 22, 2026): 59 blog posts, 5 case studies (Black Forest Labs, Character.AI, Krea, Synthesia, Zello; the Resources page lists only the first four), 6 solution pages, 13 product pages, 2 switch pages (Community Sift, Unitary), an /llms.txt file, no pricing page and no RSS feed. Of the pages checked, only the switch and content moderation services pages carry a canonical tag, and only the content moderation services page has structured data.Source
  • 09

    The role owns "attribution, lead scoring, and inbound reporting", and the form already asks for team size and reason. Without an agreed score, marketing pipeline is argued about rather than counted.

    High impactMedium effort First result in 1 quarter AI Growth Marketer

    First action

    Write the score with the Head of Sales.

    How to measure it

    Leads with source and score

    Now
    –
    Target
    100%
    By
    1 quarter

    Steps

    1. 1

      Agree the score

      Fields and weights.

      AI Growth Marketer 1 week

    2. 2

      Tag sources

      UTMs on every link.

      AI Growth Marketer 1 week

    3. 3

      CRM fields

      Score and touches.

      AI Growth Marketer 1 week

    4. 4

      Handoff rule

      Score threshold to sales.

      Head of Sales 1 day

    5. 5

      Weekly view

      Pipeline by channel.

      AI Growth Marketer weekly

    6. 6

      Review

      Score against closed deals.

      Head of Sales quarterly

    First result
    1 quarter
    Cost
    Staff time only
    Owner
    AI Growth Marketer

    Smallest test

    Tag every link in the next newsletter.

    Why it works

    Agree a score with sales (launch date, segment, team size, research downloads, event attendance), track first and last touch on every lead, and publish one weekly pipeline view by channel.

    A shared definition makes the pipeline target real and shows where the next dollar should go.

    Risks and how to handle them

    • Score ignores dark social

      Add a "how did you hear" field.

    Evidence

    • HiringAI Growth Marketer duties: lift the demand-gen performance of case studies ("One of our current case studies is already one of our best-performing assets... good enough that customers ask to be featured"); turn the blog, newsletter and founder ghostwriting into "an explosive growth channel" with "original research or benchmarks that set Cinder apart"; decide which events to attend, sponsor or skip; "Spot and act on relevant news and industry moments quickly"; own "attribution, lead scoring, and inbound reporting"; and partner with sales "so marketing-sourced leads actually convert".Source
    • Your websiteDemo form at /contact asks for team size and a reason for reaching out with four options: "Evaluating platforms", "Replacing existing tool", "Researching the space", "Other". There is no option for a model or product launch, or for red teaming.Source
    • HiringAI Growth Marketer (New York, posted September 25, 2026, reports to the Head of Sales): "Most companies building with generative AI don't know yet that they have a trust and safety problem... Right now, that story lives only with our current customers." The job is to take existing positioning on "AI red teaming, agentic moderation, and content integrity" and turn it into "a story that lands with a generative AI buyer specifically", reaching "a market that doesn't buy the way traditional enterprise software buyers do". It owns marketing-sourced pipeline, with the target set with the Head of Sales, and works with the Founding Marketer.Source
  • 10

    The role says most generative AI companies "don't know yet that they have a trust and safety problem". They show it in public before they know it: a new model on a hub, an app climbing the charts, a hiring post for a first trust and safety role.

    High impactMedium effort First result in 1 quarter AI Growth Marketer

    First action

    Pull last month's new image and video apps with funding.

    How to measure it

    Meetings booked

    Now
    –
    Target
    8 a month
    By
    1 quarter

    Steps

    1. 1

      Pick signals

      Launches, first T&S hires, funding.

      AI Growth Marketer 1 week

    2. 2

      Build the list

      Weekly refresh.

      AI Growth Marketer weekly

    3. 3

      Match stories

      One finding per segment.

      Founding Marketer 1 week

    4. 4

      Send

      Short personal notes.

      Head of Sales weekly

    5. 5

      Follow up

      Scoping calls.

      Head of Sales ongoing

    6. 6

      Measure

      Meetings by signal.

      AI Growth Marketer monthly

    First result
    1 quarter
    Cost
    Data tool cost
    Owner
    AI Growth Marketer

    Smallest test

    Write to ten teams that launched last month.

    Why it works

    Build a weekly list from those signals (new models and apps, first T&S hires, funding rounds), match it to the right story, and send short outreach with a relevant research finding. Sales follows up on replies.

    Timing beats volume: a team days from launch is open to a red-team offer.

    Risks and how to handle them

    • Too broad a list

      Start with generative media apps only.

    Evidence

    • HiringAI Growth Marketer (New York, posted September 25, 2026, reports to the Head of Sales): "Most companies building with generative AI don't know yet that they have a trust and safety problem... Right now, that story lives only with our current customers." The job is to take existing positioning on "AI red teaming, agentic moderation, and content integrity" and turn it into "a story that lands with a generative AI buyer specifically", reaching "a market that doesn't buy the way traditional enterprise software buyers do". It owns marketing-sourced pipeline, with the target set with the Head of Sales, and works with the Founding Marketer.Source
    • HiringAI Growth Marketer requirements: demand generation "for an AI product specifically" is "the requirement we care about most"; built from zero; experience in a category "where the downside of failure is existential (cybersecurity, identity and authentication, fintech risk, regtech...)". Nice to have: "You've built a research report, index, or open benchmark that became a demand asset", "ABM or intent-data tools", "Field marketing at conference scale", and marketing into "social platforms, marketplaces, dating apps, gaming, fintech apps, or B2C-facing AI products".Source
    • Your websiteSeries B press release (May 12, 2026): $41 million led by Radical Ventures with Accel, Y Combinator, M12 and others; a New York headquarters; plans "to double headcount in the next 12 months"; customers named include OpenAI, Spotify, depop and Black Forest Labs.Source
  • 11

    Cinder ran two well-attended evenings at its New York office in 2026 (synthetic NCII, applicant fraud) and has a TrustCon page with its own form. The role must decide which events to attend, sponsor or skip.

    Medium impactMedium effort First result in 2 quarters AI Growth Marketer

    First action

    Set dates for the next three evenings.

    How to measure it

    Opportunities from events

    Now
    –
    Target
    5 per event
    By
    2 quarters

    Steps

    1. 1

      Pick topics

      Three topics tied to products.

      AI Growth Marketer 1 week

    2. 2

      Invite

      Target accounts and customers.

      Head of Sales 3 weeks

    3. 3

      Record

      One talk per event.

      AI Growth Marketer per event

    4. 4

      Publish

      Post and video.

      Founding Marketer 1 week

    5. 5

      Follow up

      Within two days.

      Head of Sales 2 days

    6. 6

      Decide

      Which conferences get side events.

      AI Growth Marketer quarterly

    First result
    2 quarters
    Cost
    Venue and catering
    Owner
    AI Growth Marketer

    Smallest test

    Schedule the next evening.

    Why it works

    Run the evenings every other month on one topic, record a talk for YouTube and a post, invite target accounts, and add attendees to a tracked list. Hold one side event at each major conference instead of a bigger booth.

    Small rooms with the right buyers create more pipeline than large sponsorships, and each event produces content.

    Risks and how to handle them

    • Same faces each time

      Rotate topics and co-hosts.

    Evidence

    • Your websiteEvents: a TrustCon 2026 page with its own contact form, and two "A Night with Cinder" evenings at the New York headquarters in 2026 (synthetic NCII in May, "Know Your Applicant" on North Korean IT-worker hiring fraud in June); Cinder sells a Know Your Applicant service.Source
    • HiringAI Growth Marketer duties: lift the demand-gen performance of case studies ("One of our current case studies is already one of our best-performing assets... good enough that customers ask to be featured"); turn the blog, newsletter and founder ghostwriting into "an explosive growth channel" with "original research or benchmarks that set Cinder apart"; decide which events to attend, sponsor or skip; "Spot and act on relevant news and industry moments quickly"; own "attribution, lead scoring, and inbound reporting"; and partner with sales "so marketing-sourced leads actually convert".Source
    • YouTubeThe Cinder YouTube channel linked from the site has 14 subscribers, 3 videos and 813 total views.Source
  • 12

    The role covers ghostwriting for the founders, who spent years countering terrorism, child exploitation and nation-state actors at Meta and in government. The newsletter only sends "when we have something worth your attention".

    High impactLow effort First result in 1 month CEO

    First action

    Draft four posts from existing research.

    How to measure it

    Meetings attributed

    Now
    –
    Target
    4 a month
    By
    2 quarters

    Steps

    1. 1

      Pick findings

      Four a month.

      AI Growth Marketer monthly

    2. 2

      Draft

      Short posts with a chart.

      AI Growth Marketer weekly

    3. 3

      Approve

      Founder review.

      CEO weekly

    4. 4

      Post

      Same day each week.

      CEO weekly

    5. 5

      Roll up

      Monthly newsletter.

      AI Growth Marketer monthly

    6. 6

      Measure

      Replies and meetings.

      AI Growth Marketer monthly

    First result
    1 month
    Cost
    Staff time only
    Owner
    CEO

    Smallest test

    Publish one founder post this week.

    Why it works

    Each week, turn one finding (from research, a customer pattern or a release note) into a founder post with a chart, and roll the month's posts into one newsletter. Track followers, replies and meetings.

    Buyers in this category follow people more than brands, and a steady cadence builds the audience the research needs.

    Risks and how to handle them

    • Posts feel ghostwritten

      Keep the founder's own opinion in every one.

    Evidence

    • HiringAI Growth Marketer duties: lift the demand-gen performance of case studies ("One of our current case studies is already one of our best-performing assets... good enough that customers ask to be featured"); turn the blog, newsletter and founder ghostwriting into "an explosive growth channel" with "original research or benchmarks that set Cinder apart"; decide which events to attend, sponsor or skip; "Spot and act on relevant news and industry moments quickly"; own "attribution, lead scoring, and inbound reporting"; and partner with sales "so marketing-sourced leads actually convert".Source
    • Your websiteResources hub: 65 items (Insights 37, Product 10, Engineering 7, News 3, Case studies 4); a newsletter sign-up "sent only when we have something worth your attention". The red-team research sits in the same list as hiring and culture posts, with no research or benchmark section.Source
    • Your website"How Leading Open-Weight Models Respond to Requests Involving Violent Harm" (October 21, 2026): six open-weight models tested against 209 adversarial requests built from 70 goals, 18 attack techniques and 15 harm domains; "Four of the six gave harmful, actionable help in more than 40% of their responses, and the worst reached 56%. Even the safest gave unsafe responses about 15% of the time."Source
    • Your websiteAbout page: "Our founders spent years countering terrorism, child exploitation, and nation-state actors at Meta and in government."Source
  • 13

    Cinder has two switch pages (Community Sift and Unitary) and a post on build versus buy. Alice (formerly ActiveFence) just raised $140M for AI safety, and Checkstep leads on cost cuts and compliance.

    Medium impactMedium effort First result in 1 quarter Founding Marketer

    First action

    Write the build-versus-buy page first.

    How to measure it

    Comparison-page demos

    Now
    –
    Target
    6 a month
    By
    1 quarter

    Steps

    1. 1

      Pick comparisons

      Five most common in deals.

      Head of Sales 1 week

    2. 2

      Check facts

      Rivals' current claims.

      Founding Marketer 1 week

    3. 3

      Write

      Fair tables and switch steps.

      Founding Marketer 2 weeks

    4. 4

      Legal review

      Claims and trademarks.

      Legal 1 week

    5. 5

      Publish

      With canonical tags.

      Web developer 1 day

    6. 6

      Refresh

      Quarterly.

      Founding Marketer quarterly

    First result
    1 quarter
    Cost
    Staff time only
    Owner
    Founding Marketer

    Smallest test

    Publish the build-versus-buy page.

    Why it works

    Write fair comparison pages for in-house tools, a BPO, Alice, Hive and Checkstep: what each does well, where Cinder differs (agents plus reviewers in one system, shadow mode, proof on your own content), and how a switch works.

    Buyers search for alternatives before talking to sales, and the shadow-mode promise is a strong differentiator.

    Risks and how to handle them

    • Out-of-date claims

      Date each page and cite sources.

    Evidence

    • Your websiteSitemap (100 URLs, checked October 22, 2026): 59 blog posts, 5 case studies (Black Forest Labs, Character.AI, Krea, Synthesia, Zello; the Resources page lists only the first four), 6 solution pages, 13 product pages, 2 switch pages (Community Sift, Unitary), an /llms.txt file, no pricing page and no RSS feed. Of the pages checked, only the switch and content moderation services pages carry a canonical tag, and only the content moderation services page has structured data.Source
    • Competitor researchRivals' own homepages: Alice (formerly ActiveFence) announces it "raises $140M for frontier AI safety and security" and sells red teaming and custom guardrails; Hive sells APIs of pre-trained content-understanding models, with a Pricing page in its main navigation; Checkstep claims "98% accuracy", "Cut moderation costs 90%" and DSA and OSA compliance.Source
    • Your websiteContent moderation services page: "A BPO sells reviewer hours. Cinder pairs reviewers with agents that take over routine work over time"; Cinder "runs in shadow mode next to your current vendor"; the switch pages promise to "prove it on your content" before anything changes.Source
  • 14

    Cinder's YouTube channel has 3 videos and 813 views. Red-team findings are visual (attack, response, fix) and safety engineers watch technical walkthroughs.

    Medium impactMedium effort First result in 1 quarter AI Growth Marketer

    First action

    Record one walkthrough of the violent-harm test.

    How to measure it

    Monthly views

    Now
    –
    Target
    5,000
    By
    2 quarters

    Steps

    1. 1

      Script

      Attack, response, fix.

      Red team lead 1 week

    2. 2

      Record

      Screen and voice.

      AI Growth Marketer 1 day

    3. 3

      Review

      Remove harmful detail.

      Red team lead 1 day

    4. 4

      Publish

      YouTube and research page.

      AI Growth Marketer 1 day

    5. 5

      Cut clips

      Social versions.

      AI Growth Marketer 1 day

    6. 6

      Measure

      Views and clicks.

      AI Growth Marketer monthly

    First result
    1 quarter
    Cost
    Staff time only
    Owner
    AI Growth Marketer

    Smallest test

    Publish one walkthrough.

    Why it works

    Record a five-minute walkthrough for each research post (with harmful content blurred), cut 30-second clips for social, and embed the video on the research page.

    Video spreads research further than text and gives sales something to send.

    Risks and how to handle them

    • Showing attacks teaches attackers

      Show categories and rates, never working prompts.

    Evidence

    • YouTubeThe Cinder YouTube channel linked from the site has 14 subscribers, 3 videos and 813 total views.Source
    • Your website"How Leading Open-Weight Models Respond to Requests Involving Violent Harm" (October 21, 2026): six open-weight models tested against 209 adversarial requests built from 70 goals, 18 attack techniques and 15 harm domains; "Four of the six gave harmful, actionable help in more than 40% of their responses, and the worst reached 56%. Even the safest gave unsafe responses about 15% of the time."Source
    • Your websiteRed-team post "What Cinder Learned Red Teaming Alibaba's Latest Open-Weight Model" (August 24, 2026): across 2,670 attacks the original model generated NCII once and the abliterated version violated 23.1% of the time. Two places in the post where tables belong show the text "[@portabletext/react] Unknown block type "tableBlock", specify a component for it in the `components.types` prop" instead (checked twice on October 22, 2026). The post ends with a newsletter sign-up; there is no download of the data or method.Source
  • 15

    Cinder is hiring a Forward Deployed Engineer in London and works with the Appeals Centre Europe; Checkstep leads its pitch with DSA and OSA compliance. Cinder's site has no page for European duties.

    Medium impactMedium effort First result in 2 quarters Head of Sales

    First action

    List the duties Cinder already supports.

    How to measure it

    EU and UK demos

    Now
    –
    Target
    6 a month
    By
    2 quarters

    Steps

    1. 1

      Map duties

      Feature by duty.

      Senior Product Manager, Trust & Safety 2 weeks

    2. 2

      Legal review

      Accurate wording.

      Legal 2 weeks

    3. 3

      Write the page

      With a London contact.

      Founding Marketer 1 week

    4. 4

      Workshop offer

      Run by the FDE.

      Forward Deployed Engineer 1 month

    5. 5

      Promote

      UK and EU outreach.

      Head of Sales 1 month

    6. 6

      Measure

      Demos by region.

      AI Growth Marketer monthly

    First result
    2 quarters
    Cost
    Legal review time
    Owner
    Head of Sales

    Smallest test

    Run one workshop in London.

    Why it works

    Publish a page mapping DSA and Online Safety Act duties (notices, statements of reasons, appeals, risk assessments) to Cinder features, with a London contact and a workshop offer run by the FDE.

    Compliance deadlines are a reason to buy now, and a local team lowers the risk for European buyers.

    Risks and how to handle them

    • Legal advice risk

      Describe features, not legal conclusions.

    Evidence

    • HiringForward Deployed Engineer (London, posted September 28, 2026): the technical owner after a deal closes; builds "proofs of concept that demonstrate value and de-risk the path to production", connects Cinder to each customer's stack, runs "technical workshops", and turns patterns across customers into "reusable tools". An AI/ML Engineer (posted August 27) will turn "the enormous volume of customer decisions" into better models.Source
    • Competitor researchRivals' own homepages: Alice (formerly ActiveFence) announces it "raises $140M for frontier AI safety and security" and sells red teaming and custom guardrails; Hive sells APIs of pre-trained content-understanding models, with a Pricing page in its main navigation; Checkstep claims "98% accuracy", "Cut moderation costs 90%" and DSA and OSA compliance.Source
    • Your websiteBlog posts on partnerships and integrations: StopNCII, the Internet Watch Foundation, the Christchurch Call, the Appeals Centre Europe, ROOST (open-source safety), an NCMEC CyberTipline integration and the OpenAI moderation API inside Cinder.Source
  • 16

    Cinder's open-weight page describes release governance (resolved, accepted and open risks with owners and sign-off). Teams releasing weights have no standard checklist, and a free tool reaches them before they need a vendor.

    High impactMedium effort First result in 1 quarter Senior Product Designer

    First action

    Write the checklist as a document first.

    How to measure it

    Calls from the checklist

    Now
    –
    Target
    5 a month
    By
    1 quarter

    Steps

    1. 1

      Write items

      From real engagements.

      Red team lead 2 weeks

    2. 2

      Design

      Simple interactive page.

      Senior Product Designer 2 weeks

    3. 3

      Build

      Export to a report.

      Web developer 2 weeks

    4. 4

      Launch

      Research newsletter and founder post.

      AI Growth Marketer 1 week

    5. 5

      Follow up

      Exports with company emails.

      Head of Sales ongoing

    6. 6

      Improve

      From user feedback.

      Senior Product Designer quarterly

    First result
    1 quarter
    Cost
    Build time
    Owner
    Senior Product Designer

    Smallest test

    Share the document version with five teams.

    Why it works

    Publish an interactive checklist: risk areas, tests run, results, owner and sign-off. Users can export it as a launch report; a "have Cinder run this" button books a scoping call.

    A useful tool earns links and repeat visits from exactly the teams Cinder sells to.

    Risks and how to handle them

    • Seen as a full safety guarantee

      Label it a starting point.

    Evidence

    • Your websiteOpen-weight model safety page: pre-release red teaming, safety eval operations, guardrail loops and "Release governance"; "Open-weight releases compress the timeline between model availability and real-world misuse." The Trust Center (Vanta) lists SOC 2 and HIPAA.Source
    • HiringAI Growth Marketer (New York, posted September 25, 2026, reports to the Head of Sales): "Most companies building with generative AI don't know yet that they have a trust and safety problem... Right now, that story lives only with our current customers." The job is to take existing positioning on "AI red teaming, agentic moderation, and content integrity" and turn it into "a story that lands with a generative AI buyer specifically", reaching "a market that doesn't buy the way traditional enterprise software buyers do". It owns marketing-sourced pipeline, with the target set with the Head of Sales, and works with the Founding Marketer.Source
    • Your websiteGenerative AI solution page: H1 "SHIP FASTER. WITHOUT SHIPPING RISK."; covers AI red teaming, prompt and output screening, QA and custom safety agents; proof ">97% AI Moderation Accuracy", "< 1% NSFW rate across evaluated generations", ">90% Reduction in CSAM and NCII vulnerability", "10X Safer than benchmark industry models at launch"; the only call to action is "Get a demo".Source

Big bets 4

Larger plays that can change the trajectory.

  • 17

    Cinder has run comparable tests across six open-weight models, and the role prizes "a research report, index, or open benchmark that became a demand asset". Alice and others compete for the same AI safety budget with bigger rounds.

    High impactHigh effort First result in 3 quarters Red team lead

    First action

    Agree a fixed method from the two tests already run.

    How to measure it

    Inbound red-team deals

    Now
    –
    Target
    +50%
    By
    3 quarters

    Steps

    1. 1

      Fix the method

      Domains, techniques, judge.

      Red team lead 1 month

    2. 2

      External review

      Two outside researchers.

      Red team lead 1 month

    3. 3

      Score models

      Current notable releases.

      Red team lead 1 month

    4. 4

      Publish

      Index page with method.

      AI Growth Marketer 2 weeks

    5. 5

      Update

      Each new release.

      Red team lead ongoing

    6. 6

      Sell with it

      Every red-team deck.

      Head of Sales ongoing

    First result
    3 quarters
    Cost
    Red team and compute time
    Owner
    Red team lead

    Smallest test

    Publish the six-model results as version one.

    Why it works

    Publish a standing index: each notable open-weight model scored on the same harm domains with a published method, re-run on new versions and on common fine-tunes. Invite outside reviewers to check the method.

    An index becomes the reference press, labs and buyers check before launch, which no funding round can buy.

    Risks and how to handle them

    • Disputes with model makers

      Share results before publishing and accept corrections.

    Evidence

    • Your website"How Leading Open-Weight Models Respond to Requests Involving Violent Harm" (October 21, 2026): six open-weight models tested against 209 adversarial requests built from 70 goals, 18 attack techniques and 15 harm domains; "Four of the six gave harmful, actionable help in more than 40% of their responses, and the worst reached 56%. Even the safest gave unsafe responses about 15% of the time."Source
    • Your websiteRed-team post "What Cinder Learned Red Teaming Alibaba's Latest Open-Weight Model" (August 24, 2026): across 2,670 attacks the original model generated NCII once and the abliterated version violated 23.1% of the time. Two places in the post where tables belong show the text "[@portabletext/react] Unknown block type "tableBlock", specify a component for it in the `components.types` prop" instead (checked twice on October 22, 2026). The post ends with a newsletter sign-up; there is no download of the data or method.Source
    • HiringAI Growth Marketer requirements: demand generation "for an AI product specifically" is "the requirement we care about most"; built from zero; experience in a category "where the downside of failure is existential (cybersecurity, identity and authentication, fintech risk, regtech...)". Nice to have: "You've built a research report, index, or open benchmark that became a demand asset", "ABM or intent-data tools", "Field marketing at conference scale", and marketing into "social platforms, marketplaces, dating apps, gaming, fintech apps, or B2C-facing AI products".Source
    • Competitor researchRivals' own homepages: Alice (formerly ActiveFence) announces it "raises $140M for frontier AI safety and security" and sells red teaming and custom guardrails; Hive sells APIs of pre-trained content-understanding models, with a Pricing page in its main navigation; Checkstep claims "98% accuracy", "Cut moderation costs 90%" and DSA and OSA compliance.Source
  • 18

    The role says this market "doesn't buy the way traditional enterprise software buyers do". Hive sells content-understanding APIs with public pricing; Cinder's only door is a demo.

    High impactHigh effort First result in 3 quarters Senior Product Manager

    First action

    Interview ten AI startups about how they buy safety today.

    How to measure it

    Self-serve to enterprise conversions

    Now
    –
    Target
    10
    By
    4 quarters

    Steps

    1. 1

      Interview

      Ten startups.

      Senior Product Manager, Trust & Safety 1 month

    2. 2

      Define the tier

      Limits and price.

      Senior Product Manager, Trust & Safety 1 month

    3. 3

      Build

      Sign-up, keys, billing.

      Engineering 1 quarter

    4. 4

      Templates

      Common policies.

      Senior Product Manager, Trust & Safety 1 month

    5. 5

      Launch

      Research audience first.

      AI Growth Marketer 1 month

    6. 6

      Track

      Usage signals for sales.

      Head of Sales ongoing

    First result
    3 quarters
    Cost
    Engineering time
    Owner
    Senior Product Manager, Trust & Safety

    Smallest test

    Offer API keys to five startups by hand.

    Why it works

    Offer a usage-priced screening API with Cinder's policy templates for small AI teams, with a path to agents, QA and red teaming as they grow. Track which accounts grow into enterprise contracts.

    Startups that adopt early become the enterprise customers of tomorrow, and usage shows when to call.

    Risks and how to handle them

    • Support load

      Docs and community support only on the free tier.

    Evidence

    • HiringAI Growth Marketer (New York, posted September 25, 2026, reports to the Head of Sales): "Most companies building with generative AI don't know yet that they have a trust and safety problem... Right now, that story lives only with our current customers." The job is to take existing positioning on "AI red teaming, agentic moderation, and content integrity" and turn it into "a story that lands with a generative AI buyer specifically", reaching "a market that doesn't buy the way traditional enterprise software buyers do". It owns marketing-sourced pipeline, with the target set with the Head of Sales, and works with the Founding Marketer.Source
    • Your websiteGenerative AI solution page: H1 "SHIP FASTER. WITHOUT SHIPPING RISK."; covers AI red teaming, prompt and output screening, QA and custom safety agents; proof ">97% AI Moderation Accuracy", "< 1% NSFW rate across evaluated generations", ">90% Reduction in CSAM and NCII vulnerability", "10X Safer than benchmark industry models at launch"; the only call to action is "Get a demo".Source
    • Competitor researchRivals' own homepages: Alice (formerly ActiveFence) announces it "raises $140M for frontier AI safety and security" and sells red teaming and custom guardrails; Hive sells APIs of pre-trained content-understanding models, with a Pricing page in its main navigation; Checkstep claims "98% accuracy", "Cut moderation costs 90%" and DSA and OSA compliance.Source
    • Your websiteSitemap (100 URLs, checked October 22, 2026): 59 blog posts, 5 case studies (Black Forest Labs, Character.AI, Krea, Synthesia, Zello; the Resources page lists only the first four), 6 solution pages, 13 product pages, 2 switch pages (Community Sift, Unitary), an /llms.txt file, no pricing page and no RSS feed. Of the pages checked, only the switch and content moderation services pages carry a canonical tag, and only the content moderation services page has structured data.Source
  • 19

    Open-weight releases reach users through model hubs and inference providers within hours. Cinder's open-weight page sells release governance, but each lab must be found one at a time.

    High impactHigh effort First result in 3 quarters Head of Sales

    First action

    List five hosting and inference platforms and their safety offers.

    How to measure it

    Partner opportunities

    Now
    –
    Target
    10
    By
    3 quarters

    Steps

    1. 1

      List partners

      Hubs and inference providers.

      Head of Sales 2 weeks

    2. 2

      Design the offer

      Add-on or referral.

      Senior Product Manager, Trust & Safety 1 month

    3. 3

      Pilot

      One partner.

      Head of Sales 1 quarter

    4. 4

      Integrate

      API hook.

      Forward Deployed Engineer 1 month

    5. 5

      Co-market

      Joint post and webinar.

      AI Growth Marketer 1 month

    6. 6

      Measure

      Pipeline by partner.

      AI Growth Marketer monthly

    First result
    3 quarters
    Cost
    Partner team time
    Owner
    Head of Sales

    Smallest test

    Pitch one inference provider.

    Why it works

    Offer inference and hosting platforms a screening add-on or a red-team referral for models they host, with revenue share. Start with one platform and one model family.

    One partner can put Cinder in front of hundreds of launches.

    Risks and how to handle them

    • Partner builds its own tools

      Offer depth they cannot build: reviewers and red team.

    Evidence

    • Your websiteOpen-weight model safety page: pre-release red teaming, safety eval operations, guardrail loops and "Release governance"; "Open-weight releases compress the timeline between model availability and real-world misuse." The Trust Center (Vanta) lists SOC 2 and HIPAA.Source
    • HiringAI Growth Marketer (New York, posted September 25, 2026, reports to the Head of Sales): "Most companies building with generative AI don't know yet that they have a trust and safety problem... Right now, that story lives only with our current customers." The job is to take existing positioning on "AI red teaming, agentic moderation, and content integrity" and turn it into "a story that lands with a generative AI buyer specifically", reaching "a market that doesn't buy the way traditional enterprise software buyers do". It owns marketing-sourced pipeline, with the target set with the Head of Sales, and works with the Founding Marketer.Source
    • Your websiteBlog posts on partnerships and integrations: StopNCII, the Internet Watch Foundation, the Christchurch Call, the Appeals Centre Europe, ROOST (open-source safety), an NCMEC CyberTipline integration and the OpenAI moderation API inside Cinder.Source
  • 20

    Cinder processes "400M+ events processed daily" for over thirty customers, many with tens or hundreds of millions of users. Aggregated, that is a view of how abuse is changing that no one else has.

    High impactHigh effort First result in 4 quarters AI/ML Engineer

    First action

    Pick ten questions buyers ask most.

    How to measure it

    Downloads

    Now
    –
    Target
    3,000
    By
    4 quarters

    Steps

    1. 1

      Pick questions

      Ten buyer questions.

      AI Growth Marketer 1 month

    2. 2

      Customer approval

      Data use terms.

      Legal 1 quarter

    3. 3

      Analyze

      Aggregated trends.

      AI/ML Engineer 1 quarter

    4. 4

      Write

      Report and summary.

      Founding Marketer 1 month

    5. 5

      Launch

      Press and an event.

      AI Growth Marketer 1 month

    6. 6

      Use in sales

      Segment findings.

      Head of Sales ongoing

    First result
    4 quarters
    Cost
    Analyst time
    Owner
    AI/ML Engineer

    Smallest test

    Publish one aggregated chart with customer approval.

    Why it works

    Publish a yearly report with aggregated, anonymized trends: which abuse types grew, how much AI-generated content is reported, how fast agents handle cases. Get customer approval and suppress small groups.

    Original data earns press and citations and gives every sales call a reason to start.

    Risks and how to handle them

    • Customer data concerns

      Aggregate only and get written approval.

    Evidence

    • Your websiteHomepage: H1 "Deploy AI to Fight Abuse"; a site-wide banner "Deploy open-weight models with frontier safety built in"; proof "3B+ end-users protected", "94% of human review, automated", "100k+ hours saved", "400M+ events processed daily"; a Synthesia CEO quote; one call to action, "Get a demo".Source
    • HiringSenior Product Manager, Trust & Safety (posted July 23, 2026): Cinder is "an AI-native safety platform serving over thirty customers, including some of the largest consumer platforms in the world. Many of them serve tens or hundreds of millions of users"; many users "are domain experts rather than technical ones".Source
    • HiringForward Deployed Engineer (London, posted September 28, 2026): the technical owner after a deal closes; builds "proofs of concept that demonstrate value and de-risk the path to production", connects Cinder to each customer's stack, runs "technical workshops", and turns patterns across customers into "reusable tools". An AI/ML Engineer (posted August 27) will turn "the enormous volume of customer decisions" into better models.Source

Wildcards 2

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

  • 21

    Cinder sells a Know Your Applicant service and hosted an evening on North Korean IT workers in hiring pipelines. That buyer (talent, security, IT) is different from trust and safety and every tech company has the problem.

    Medium impactMedium effort First result in 2 quarters Head of Sales

    First action

    Write the guide from the June event.

    How to measure it

    KYA opportunities

    Now
    –
    Target
    6
    By
    2 quarters

    Steps

    1. 1

      Write the guide

      Hiring fraud signals.

      AI Growth Marketer 2 weeks

    2. 2

      Build the page

      Talent and security audience.

      Founding Marketer 1 week

    3. 3

      Webinar

      With a recruiting leader.

      AI Growth Marketer 1 month

    4. 4

      Promote

      Recruiting and security channels.

      AI Growth Marketer 1 month

    5. 5

      Sell

      Short pilots.

      Head of Sales ongoing

    6. 6

      Track

      Cross-sell later.

      Head of Sales quarterly

    First result
    2 quarters
    Cost
    Staff time only
    Owner
    Head of Sales

    Smallest test

    Run the webinar.

    Why it works

    Build a page and short guide for talent and security leaders, run a webinar on hiring fraud signals, and list it in recruiting and security channels. Track whether these accounts later buy other products.

    A shared, urgent problem opens doors at companies that do not yet have a trust and safety team.

    Risks and how to handle them

    • Distraction from the core

      Keep it a small, measured side channel.

    Evidence

    • Your websiteEvents: a TrustCon 2026 page with its own contact form, and two "A Night with Cinder" evenings at the New York headquarters in 2026 (synthetic NCII in May, "Know Your Applicant" on North Korean IT-worker hiring fraud in June); Cinder sells a Know Your Applicant service.Source
    • Your websiteSeries B press release (May 12, 2026): $41 million led by Radical Ventures with Accel, Y Combinator, M12 and others; a New York headquarters; plans "to double headcount in the next 12 months"; customers named include OpenAI, Spotify, depop and Black Forest Labs.Source
  • 22

    Cinder builds wellness tooling (routing, blurring, exposure limits) for moderators and red teamers exposed to the worst content. Safety leaders care deeply about this, and few vendors talk about it.

    Medium impactLow effort First result in 1 quarter Senior Product Designer

    First action

    Add the wellness controls to the standard demo.

    How to measure it

    Deals citing wellness

    Now
    –
    Target
    20%
    By
    2 quarters

    Steps

    1. 1

      Write the guide

      Exposure management.

      Founding Marketer 2 weeks

    2. 2

      Demo it

      Standard demo flow.

      Head of Sales 1 week

    3. 3

      Customer report

      Quarterly exposure metrics.

      Senior Product Manager, Trust & Safety 1 month

    4. 4

      Promote

      Founder post and newsletter.

      AI Growth Marketer 1 week

    5. 5

      Event

      One evening on moderator wellbeing.

      AI Growth Marketer 1 quarter

    6. 6

      Measure

      Deals citing it.

      Head of Sales quarterly

    First result
    1 quarter
    Cost
    Staff time only
    Owner
    Senior Product Designer

    Smallest test

    Publish the guide.

    Why it works

    Publish an exposure-management guide and a short benchmark of practices, show the wellness controls in every demo, and report exposure metrics to customers each quarter.

    It speaks to the buyer as a manager of people, and it is a retention point once teams rely on it.

    Risks and how to handle them

    • Seen as a soft topic

      Tie it to turnover and accuracy.

    Evidence

    • Your websiteReviewer and red team wellness page: routing, blurring and grayscale, permissions and exposure limits for moderators and AI red teamers "exposed to CSAM, NCII, gore, violent extremism, and the worst outputs of generative models".Source
    • HiringSenior Product Manager, Trust & Safety (posted July 23, 2026): Cinder is "an AI-native safety platform serving over thirty customers, including some of the largest consumer platforms in the world. Many of them serve tens or hundreds of millions of users"; many users "are domain experts rather than technical ones".Source
    • HiringAI Growth Marketer requirements: demand generation "for an AI product specifically" is "the requirement we care about most"; built from zero; experience in a category "where the downside of failure is existential (cybersecurity, identity and authentication, fintech risk, regtech...)". Nice to have: "You've built a research report, index, or open benchmark that became a demand asset", "ABM or intent-data tools", "Field marketing at conference scale", and marketing into "social platforms, marketplaces, dating apps, gaming, fintech apps, or B2C-facing AI products".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.

ProgrammableInfo

Turn the research into a list you can act on

These ideas start with finding the right companies. Info finds them and the right person at each, researches every one and drafts the outreach, then asks before it sends.

  • Build this list in Info for idea #10 (opens programmable.info in a new tab)

New accounts: 100,000 free credits with code BETA100K

Find the companies (opens programmable.info in a new tab)
ProgrammableDigital

Show it, don't just describe it

These ideas need something to show: a demo, a video, an animation, graphics or a deck. Digital turns screenshots or the site into all of them, in the company's own brand.

Make a demo (opens programmable.digital in a new tab)
More example playbooks and reviews

More growth playbooks

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

  • Dando La Hora growth strategyA Chilean online watch store and watch community: Casio vintage, G-Shock, Seiko, Orient, Citizen, Tissot and niche brands, plus straps…22 growth ideas
  • Givelo growth strategyPremium cycling apparel designed and made in Medellín, Colombia, with Italian and French fabrics: jerseys, bib shorts, jackets…22 growth ideas
  • Liftoff growth strategyAn AI-powered performance marketing platform for mobile apps: user acquisition (Accelerate), publisher monetization (Monetize, Vungle…21 growth ideas
  • Caracole growth strategyA Greensboro, North Carolina maker of contemporary luxury furniture for living, dining and bedroom, sold through authorized dealers and to…22 growth ideas
  • The Pokémon Company International growth strategyManages the Pokémon brand outside Asia: licensing and marketing, the Pokémon Trading Card Game, the animated series, Pokémon Center…21 growth ideas
  • Lily-Lark growth strategyHandcrafted bamboo sun parasols with UPF 50+ canopies printed with original contemporary art, in two lines (Lark from $93 and Signature…23 growth ideas
  • AutoScout24 Austria growth strategyRuns autoscout24.at (and gebrauchtwagen.at), the Austrian sites of Europe's largest online car marketplace: new and used cars, motorbikes…22 growth ideas
  • Enhesa growth strategyExpert-authored, AI-enhanced regulatory intelligence for EHS, product, chemical and corporate sustainability compliance, including…21 growth ideas

See every playbook and website review

Independent research. Programmable wrote this playbook from public sources, as they stood on October 23, 2026, and cites them throughout. Cinder did not commission, review or endorse it, and Programmable is not affiliated with Cinder. 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 Cinder runs its business.

Something wrong, or do you represent Cinder? 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 306 ideas

    Research series and table fix, homepage video, launch option on the form, canonicals and structured data, Zello and segment proof, the 72-hour release note.

  2. Days 31 to 6012 ideas

    Three buyer stories, the public mark, attribution, launch-signal outreach, events series, founder posts, switch pages, videos, the Europe page and the release checklist; start KYA and wellness.

  3. Days 61 to 904 ideas

    Commit to the big bets: the safety index, self-serve screening, hosting partners and the yearly abuse report.

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 agents that clear routine cases so moderators focus on hard ones. The most cited problem: growing moderation backlogs.

  • Red teaming before a model launch
  • Launch deadlines with unknown safety risks
  • In their words: “shrinking the haystack”

Market

Competitive landscape

4 competitors compared: Alice (formerly ActiveFence), Hive, Checkstep and In-house tools and BPOs.

  • Gap to exploit: A public, repeatable safety index for open-weight models
  • Alice (formerly ActiveFence): Rebrand still settling
  • Hive: Models rather than operations

Hiring signals

Cinder is building demand generation for the first time.

  • AI Growth Marketer: Marketing-sourced pipeline from generative AI buyers; research, case studies, events. Ideas 1, 3, 6-12, 17.
  • Senior Product Designer: Multiple product lines and a growing brand and GTM footprint. Ideas 2, 16.
  • Senior Product Manager, Trust & Safety: Over thirty customers; products from discovery to adoption. Ideas 15, 18.

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 firstNew accounts get 100,000 free enrichment credits with code BETA100K (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)
  • ProgrammableMediaAn AI note taker, interpreter and webinar host for your calls (opens programmable.media 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