ProgrammableMarketing/aside.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 · 9
  • Big bets · 3
  • Wildcards · 2
Generated Nov 10, 2026
  1. Programmable
  2. Playbooks
  3. Aside
Aside logo

growth playbook · Nov 10, 2026

aside.com (opens in a new tab)

Aside growth playbook

Diagnosis

Aside is a Y Combinator Fall 2025 company with a team of 8. It launched in June 2026 to 20K+ users in two weeks, now reports more than 50k users growing 25% week over week, and shipped Windows in September. It is hiring its first GTM lead plus product, harness, iOS and design engineers.

The product story is strong (benchmark wins, a password manager built for agents, local memory), but the website around it has 17 URLs. Visitors are left to work out the unclear parts themselves: the pricing page sells usage in credits without saying what a credit buys, there are no comparison or switching pages (a rival browser publishes pages about Aside instead), onboarding suggests no first task, and there are no customer stories anywhere.

This playbook fixes those first: credits in plain tasks, searches about Aside answered on aside.com, a switching path, a first-task gallery and trust answers. It then builds the channels the GTM role lists (user stories, referrals, founder clips, developer distribution, use-case pages) and three bigger bets: reproducible benchmarks, phone approvals and a team plan.

Growth Score

7/10

Biggest Opportunity

Turn a fast-growing, viral install base into paying, retained users by making the value of a credit, the first task and the safety model obvious, and by owning every search about Aside on aside.com.

Biggest Risk

Comet, Dia and free open-source agent browsers compete for the same early adopters. Credits that feel unpredictable or a bad first task can lose users after the install, before they ever pay.

01

Start here

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

  1. 1

    Say what 500 credits buys: real task costs and yearly prices on the pricing page

    Pull median credits per task type from the last 30 days of usage data for eight common tasks.

    High impact First result in 2 weeks
  2. 2

    Start every new user with a gallery of first tasks taken from the use-case posts

    Pull the six sales prompts and the best two from each other post into a list.

    High impact First result in 3 weeks
  3. 3

    Answer the searches about Aside on aside.com, before a rival browser does

    Check Search Console for queries containing “aside” plus vs, review, pricing or safe.

    High impact First result in 1 month

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 · 2Big bets · 9Fill-ins · 6Reconsider · 5
Effort →

Showing 22 of 22 ideas

This week 8

Quick wins with a clear first action.

  • 01

    Aside's pricing page sells usage in a unit it never defines. Free gives “500 credits per month”, Pro “3x Free usage”, Max “40x Free usage”, and the help article says only that “Credit rates differ by model and speed”, without saying what a task costs. The Yearly switch says “-20% off” but shows no yearly amounts, and there is no FAQ. So a user deciding between $20 and $200 cannot tell whether either covers their week, and outside pages, including a rival browser's, are the ones explaining Aside's plans.

    High impactLow effort First result in 2 weeks GTM

    First action

    Pull median credits per task type from the last 30 days of usage data for eight common tasks.

    How to measure it

    Pricing-page upgrade rate

    Now
    –
    Target
    +20%
    By
    60 days

    Steps

    1. 1

      Measure real usage

      Median and 90th-percentile credits for eight common task types on the default model.

      MTS, Product 2 days

    2. 2

      Pick the tasks

      Choose tasks from the five use-case posts that users recognise.

      GTM 1 day

    3. 3

      Write the table and FAQ

      Credit ranges, monthly counts per plan, what happens when credits run out, own-subscription use and add-on carry-over.

      GTM 2 days

    4. 4

      Show yearly prices

      Display yearly dollar amounts and per-month equivalents when Yearly is selected.

      Founding Designer 1 day

    5. 5

      Link from the app

      Add a “What do credits buy?” link in Settings > Plan & Usage.

      MTS, Product 1 day

    6. 6

      Measure

      Upgrade rate and support questions about credits, before and after.

      GTM 4 weeks

    First result
    2 weeks
    Cost
    Staff time only
    Owner
    GTM

    Smallest test

    Add a static five-row table and the yearly amounts this week, then compare upgrade rate with the previous fortnight.

    Why it works

    Add a “What 500 credits gets you” table to /pricing with 6-8 real tasks from the use-case posts (a research brief, a CRM update, a weekly invoice download routine, an Ultrabrowse deep-research run), each with its typical credit range on the default model and speed. Show the same tasks as monthly counts per plan (Free, Pro at 3x, Max at 40x), show yearly dollar amounts when the switch is on Yearly, and add a short FAQ: what a credit is, what happens when they run out, whether bringing your own subscription uses credits, and how add-on credits carry over. Link the table from the in-app Plan & Usage screen.

    Usage-based plans convert when buyers can map the unit to their own work. A concrete table turns “3x” and “40x” into “about N research briefs a month”, removes the main reason to hesitate at the paywall, and gives search and AI assistants a first-party answer to “how much does Aside cost”.

    Ready to use

    CopyWhat 500 credits gets you (draft layout)

    Task | Typical credits | Free (500) | Pro (3x) | Max (40x)
    Sourced research brief | from usage data | per month | per month | per month
    CRM field check with source URLs | from usage data | ... | ... | ...
    Weekly invoice download routine | from usage data | ... | ... | ...
    Ultrabrowse deep research | Pro and Max only | n/a | ... | ...
    
    FAQ: What is a credit? What happens when I run out? Does using my own subscription use credits? Do add-on credits expire? What does yearly cost?

    Risks and how to handle them

    • Credit costs change as models and the harness get cheaper

      Show ranges, date the table and update it with each harness release.

    • Heavy users anchor on the high end

      Show the typical figure first and the range second.

    Evidence

    • Your websitePricing page: Free “$0/forever” with “500 credits per month” and “Up to 3 routines”; Pro “$20/month” with “3x Free usage”, Ultrabrowse, unlimited routines, “Channels for remote control” and “Cloud handoff” marked “Coming Soon”; Max “$200/month” with “40x Free usage”; Enterprise by “Contact sales” for teams on shared company accounts; a Monthly/Yearly switch labelled “-20% off” with no yearly amounts shown; no FAQ.Source
    • Help centerSubscription help article: does not define a credit or say how many a task uses (“Credit rates differ by model and speed”); users check charges under Settings > Plan & Usage; Free credits renew each UTC month; purchased add-on credits carry over and do not expire.Source
    • Competitor researchA rival browser, Tabbit, publishes pages about Aside (“Aside Browser · official link, pricing & differences”, a review roundup and “Aside vs Chrome”) that restate Aside's plans and end with “Download Tabbit free”.Source
    • Your websiteBlog: eight entries, the newest a July 10, 2026 link to 9to5Mac coverage; the five use-case posts carry copyable prompts but no customer stories (the sales post uses a hypothetical account).Source
  • 02

    aside.com has 17 URLs and none of them is a comparison, review or FAQ page. A rival browser, Tabbit, publishes “Aside Browser · official link, pricing & differences”, a review roundup and “Aside vs Chrome”, all ending in “Download Tabbit free”. An early review still says Aside has no public pricing page. People searching for Aside by name are getting their answers from other companies.

    High impactLow effort First result in 1 month GTM

    First action

    Check Search Console for queries containing “aside” plus vs, review, pricing or safe.

    How to measure it

    Branded comparison clicks

    Now
    –
    Target
    +40%
    By
    90 days

    Steps

    1. 1

      List the queries

      Search Console queries containing “aside” plus vs, review, pricing or safe.

      GTM 1 day

    2. 2

      Check rival facts

      Each rival's current platforms and pricing, taken from their own sites.

      GTM 1 day

    3. 3

      Write four pages

      Three comparisons and one “is it worth it” page.

      GTM 1 week

    4. 4

      Add structured data

      FAQ markup, footer links and llms.txt entries.

      MTS, Product 1 day

    5. 5

      Review claims

      Founder review of every comparative claim.

      CEO 1 day

    6. 6

      Refresh

      Quarterly, and with each pricing change.

      GTM quarterly

    First result
    1 month
    Cost
    Staff time only
    Owner
    GTM

    Smallest test

    Publish “Is Aside worth it?” first and watch branded clicks for three weeks.

    Why it works

    Publish four first-party pages: Aside vs Comet, Aside vs Dia, Aside vs Chrome and “Is Aside worth it? Pricing, safety and limits”. Each gives a fair table (where the agent runs, logged-in sites, memory, password handling, platforms, price), links to the benchmark post and the credits table, and is dated and kept current. Add FAQ structured data and link the pages from the footer and the llms.txt.

    Searches that include the brand name are the highest-intent traffic a young product gets. Owning them stops buyers being redirected to a rival and gives AI assistants an accurate source.

    Risks and how to handle them

    • Comparison claims go stale in a fast market

      Date every page and link to rivals' own pages.

    Builds on

    Evidence

    • Competitor researchA rival browser, Tabbit, publishes pages about Aside (“Aside Browser · official link, pricing & differences”, a review roundup and “Aside vs Chrome”) that restate Aside's plans and end with “Download Tabbit free”.Source
    • Your websiteSitemap (17 URLs, checked November 9, 2026): home, download, developers, pricing, three feature pages (browser agent, memory, password manager), blog, manifesto, two policies, one benchmark post and five use-case posts (developers, HR, operations, researchers, sales) dated June 22, 2026. No comparison, switching, FAQ, customer-story or Windows pages.Source
    • Reviewseesel review (June 2026): “genuinely promising but unproven”; said there was no public pricing page (since added); early Hacker News users mention “some little bugs”; benchmark claims are self-reported; the approval step for sensitive actions is not spelled out in the privacy policy.Source
    • Competitor researchRivals' own homepages: Comet, “The browser that works for you”, on Mac, Windows, iOS and Android; Dia, “More than a browser. A better workday.”, free, with Dia for Work adding SSO and admin tools; BrowserOS, a free open-source browser that agents control with your existing logins, on macOS, Windows and Linux.Source
  • 03

    Aside's onboarding already imports browsing history, cookies and bookmarks, and its agent works best in sites the user is already signed into. Yet the download page never mentions importing, no page names the browsers people switch from, and the help center has only a Safari-specific instruction. Press listed Aside as an option for people leaving a discontinued AI browser, which shows how much switching drives this category.

    Medium impactLow effort First result in 2 weeks GTM

    First action

    Add the import line to the download page.

    How to measure it

    Import completion rate

    Now
    –
    Target
    +15 points
    By
    60 days

    Steps

    1. 1

      Measure import

      Current import completion by source browser.

      MTS, Product 1 day

    2. 2

      Update download page

      One line and a link about import.

      GTM 1 hour

    3. 3

      Write guides

      One per source browser, with screenshots.

      GTM 3 days

    4. 4

      Add first tasks

      Three tasks to try after import on each guide.

      GTM 1 day

    5. 5

      Fix friction

      Address the import problems listed in troubleshooting.

      MTS, Product 2 weeks

    6. 6

      Promote

      Use /switch in posts and paid tests.

      GTM ongoing

    First result
    2 weeks
    Cost
    Staff time only
    Owner
    GTM

    Smallest test

    Add the download-page line and compare import completion for two weeks.

    Why it works

    Publish /switch with short guides for moving from Chrome, Arc, Safari and other AI browsers: what imports (history, cookies, bookmarks, passwords from six managers), what stays signed in, and the first three tasks to try once you're signed in. Add one line and a link to the download page (“Bring your logins, history and bookmarks in two minutes”) and a matching help article per browser.

    Switching cost is the main barrier for a new browser. Showing that sessions come across, so the agent works in your accounts on day one, is Aside's own pitch made concrete.

    Risks and how to handle them

    • Import fails for some profiles

      Name known limits and link troubleshooting.

    Evidence

    • Help centerGet started: “Open Aside”, “Sign in or create an account”, “Import browser data” (browsing history, cookies and bookmarks; Safari by ZIP export), “Set up Password Manager”, “Start a task”, with no suggested first tasks. The download page does not mention importing and points to Discord, X and the docs.Source
    • Press9to5Mac (July 10, 2026): with OpenAI discontinuing its ChatGPT Atlas browser, it lists the alternatives people use (Chrome, Comet, Dia, Opera Neon) and Aside for macOS, “a newcomer” that “looks like a great option to test”.Source
    • Your websitePassword manager page: “The first password manager built for agents”; agents sign in by autofill without seeing passwords; scoped access and an audit log; free; one-click import from 1Password, Apple Passwords, Bitwarden, Dashlane, LastPass and Chrome; a comparison table against 1Password and Chrome Password Manager.Source
    • Help centerHelp center index (docs.aside.com): 19 entries including routines, Ultrabrowse, devices, agent permissions, troubleshooting and two changelogs; no referral programme and no guide for a specific browser you might be switching from.Source
  • 04

    Onboarding ends at “Start a task” with no suggested tasks, while the five use-case posts already contain tested, copyable prompts (call prep, CRM checks, invoice downloads, source-backed research, CI triage). New users who face an empty prompt box after a viral install often never reach the moment the agent finishes real work.

    High impactMedium effort First result in 3 weeks Founding Designer

    First action

    Pull the six sales prompts and the best two from each other post into a list.

    How to measure it

    First-session task completion

    Now
    –
    Target
    +25%
    By
    60 days

    Steps

    1. 1

      Collect prompts

      From the five use-case posts.

      GTM 1 day

    2. 2

      Test them

      Run each on common sites and keep the reliable ones.

      MTS, Agent Harness 3 days

    3. 3

      Design cards

      Role question plus 9-12 cards.

      Founding Designer 3 days

    4. 4

      Build

      Onboarding end screen and empty task state.

      MTS, Product 1 week

    5. 5

      Instrument

      Card clicks, completions and day-7 return.

      MTS, Product 1 day

    6. 6

      Rotate

      Replace weak cards monthly.

      GTM monthly

    First result
    3 weeks
    Cost
    Staff time only
    Owner
    Founding Designer

    Smallest test

    Show three cards to half of new users for two weeks.

    Why it works

    Add a first-task gallery to the end of onboarding and the empty task screen: 9-12 cards grouped by role, each a one-click prompt that fills in the sites the user just imported. Ask one question (“What do you do most in your browser?”) to sort the cards. Track which cards complete and promote the best ones.

    Activation in agent products depends on a first success. Prompts already written for real jobs get users to that success quickly and teach them what to ask.

    Risks and how to handle them

    • A failed first task hurts trust more than none

      Only ship prompts with a high tested completion rate.

    Evidence

    • Help centerGet started: “Open Aside”, “Sign in or create an account”, “Import browser data” (browsing history, cookies and bookmarks; Safari by ZIP export), “Set up Password Manager”, “Start a task”, with no suggested first tasks. The download page does not mention importing and points to Discord, X and the docs.Source
    • Your websiteBlog: eight entries, the newest a July 10, 2026 link to 9to5Mac coverage; the five use-case posts carry copyable prompts but no customer stories (the sales post uses a hypothetical account).Source
    • HiringMTS, Founding Designer: owns onboarding, browser chrome, task interfaces and settings, and designs progress, review and recovery states (“Chat alone is not the answer”) so help feels “calm, clear, and trustworthy”.Source
  • 05

    Aside's safety model is a real differentiator: agents sign in by autofill without seeing passwords, access is scoped, every credential use is logged and sensitive actions wait for confirmation. But a reviewer still found the approval step unclear and called the product “promising but unproven”. For a browser that asks for your logins, trust is the conversion barrier.

    Medium impactLow effort First result in 3 weeks GTM

    First action

    Record the clip on a test account.

    How to measure it

    Homepage download conversion

    Now
    –
    Target
    +10%
    By
    60 days

    Steps

    1. 1

      Script

      One task, one approval, one log entry.

      GTM 1 day

    2. 2

      Record

      Test account and test merchant only.

      GTM 1 day

    3. 3

      Write the list

      Actions that always wait for the user.

      CPO 1 day

    4. 4

      Place it

      Homepage and browser agent page.

      Founding Designer 1 day

    5. 5

      A/B test

      With and without the clip.

      MTS, Product 3 weeks

    6. 6

      Reuse

      In ads and comparison pages.

      GTM ongoing

    First result
    3 weeks
    Cost
    $0
    Owner
    GTM

    Smallest test

    Pin the clip on X and the homepage for two weeks.

    Why it works

    Add a 30-second clip to the homepage and browser-agent page showing a task pausing for approval before a payment or a sent message, then the audit-log entry. Add a short “What Aside will never do without you” list, link the help article on agent permissions, and use the same clip in the comparison pages and paid tests.

    Showing the approval step happen answers the main fear faster than a policy page does, and it is the strongest contrast with agents that act unsupervised.

    Risks and how to handle them

    • Overpromising on safety

      List only behaviour that is enforced in the product today.

    Evidence

    • Your websitePassword manager page: “The first password manager built for agents”; agents sign in by autofill without seeing passwords; scoped access and an audit log; free; one-click import from 1Password, Apple Passwords, Bitwarden, Dashlane, LastPass and Chrome; a comparison table against 1Password and Chrome Password Manager.Source
    • Reviewseesel review (June 2026): “genuinely promising but unproven”; said there was no public pricing page (since added); early Hacker News users mention “some little bugs”; benchmark claims are self-reported; the approval step for sensitive actions is not spelled out in the privacy policy.Source
    • Your websiteHomepage (checked November 9, 2026): H1 “Aside — The most intelligent AI assistant, but it's a browser”; “a browser rebuilt for people and agents” that “uses websites and accounts directly”; runs on the device with an existing AI subscription or your own API key; macOS 15+ and Windows 10 (22H2)+ or 11; “Backed by Y Combinator”; no customer names, testimonials or user numbers.Source
  • 06

    Aside shipped Windows in September, less than three months after the macOS launch, taking its agent to “where Chrome already lives”. The early press that drove the launch described “Aside for macOS”, and the 17-URL sitemap has no Windows page, so searches for an AI browser for Windows have no dedicated page to land on.

    Medium impactLow effort First result in 1 month GTM

    First action

    Draft /windows from the existing requirements and release notes.

    How to measure it

    Weekly Windows installs

    Now
    –
    Target
    +30%
    By
    90 days

    Steps

    1. 1

      Draft

      Requirements, import and first tasks.

      GTM 2 days

    2. 2

      Design

      Windows screenshots throughout.

      Founding Designer 2 days

    3. 3

      Publish

      Add to sitemap and llms.txt.

      MTS, Product 1 day

    4. 4

      Press updates

      Ask earlier coverage to add Windows.

      CEO 1 week

    5. 5

      Paid test

      Windows-targeted ads to /windows.

      GTM 2 weeks

    6. 6

      Measure

      Windows install share and activation.

      GTM monthly

    First result
    1 month
    Cost
    Small ad test budget
    Owner
    GTM

    Smallest test

    Publish the page and track Windows-intent search clicks for a month.

    Why it works

    Publish /windows: what works on Windows, requirements (Windows 10 22H2+ or 11, x64 and ARM64), import from the browsers people use on Windows, three Windows-relevant first tasks and a Windows download button. Point Windows-specific ads and posts to it, and ask the publications that covered the Mac launch to update their pieces.

    Most desktop users are on Windows, and a dedicated page gives search, AI assistants and press a single accurate source for the launch.

    Risks and how to handle them

    • Windows activation lags behind Mac

      Check first-task completion by OS before scaling ads.

    Evidence

    • PressRuntimeWire (September 14, 2026): Aside released its Windows app less than three months after the macOS launch, taking its browser agent to “where Chrome already lives”.Source
    • Press9to5Mac (July 10, 2026): with OpenAI discontinuing its ChatGPT Atlas browser, it lists the alternatives people use (Chrome, Comet, Dia, Opera Neon) and Aside for macOS, “a newcomer” that “looks like a great option to test”.Source
    • Your websiteSitemap (17 URLs, checked November 9, 2026): home, download, developers, pricing, three feature pages (browser agent, memory, password manager), blog, manifesto, two policies, one benchmark post and five use-case posts (developers, HR, operations, researchers, sales) dated June 22, 2026. No comparison, switching, FAQ, customer-story or Windows pages.Source
  • 07

    The Pro plan lists “Cloud handoff” as “Coming Soon” and promises “Channels for remote control”, and an iPhone app for starting and approving tasks is being built. Right now a visitor who wants those features has nowhere to say so, so the demand that could pre-sell Pro and size the mobile launch is not being recorded.

    Medium impactLow effort First result in 1 week GTM

    First action

    Add the waitlist link next to “Coming Soon”.

    How to measure it

    Waitlist-to-Pro rate

    Now
    –
    Target
    15%
    By
    90 days

    Steps

    1. 1

      Add links

      On Cloud handoff and a new iPhone app line.

      MTS, Product 1 day

    2. 2

      One-click join

      For signed-in users, in the app.

      MTS, Product 1 day

    3. 3

      Early-access rule

      Pro waitlisters get access first.

      CPO 1 day

    4. 4

      Announce

      Post and Discord message.

      GTM 1 day

    5. 5

      Share counts

      Waitlist size with the iOS and harness engineers.

      GTM monthly

    6. 6

      Launch email

      To the waitlist on release.

      GTM at launch

    First result
    1 week
    Cost
    $0
    Owner
    GTM

    Smallest test

    Add the Cloud handoff waitlist link only and count sign-ups for a week.

    Why it works

    Make “Coming Soon” a link that opens a one-field waitlist (signed-in users join with one click) for Cloud handoff and the iPhone app. Show the waitlist count on the pricing page once it is meaningful, offer waitlisters early access when they are on Pro, and email them at launch.

    Waitlists turn feature curiosity into a launch audience and an upgrade reason, and they tell the team which bet users want first.

    Risks and how to handle them

    • Delays frustrate waitlisters

      Send monthly progress notes, never dates you cannot keep.

    Evidence

    • Your websitePricing page: Free “$0/forever” with “500 credits per month” and “Up to 3 routines”; Pro “$20/month” with “3x Free usage”, Ultrabrowse, unlimited routines, “Channels for remote control” and “Cloud handoff” marked “Coming Soon”; Max “$200/month” with “40x Free usage”; Enterprise by “Contact sales” for teams on shared company accounts; a Monthly/Yearly switch labelled “-20% off” with no yearly amounts shown; no FAQ.Source
    • HiringMTS, Mobile Engineer (iOS): a new iPhone app to start browser-agent tasks, follow progress, approve actions and “use the result from anywhere”, with notifications, Live Activities and App Intents; the post says the launch reached “20K+ users and 3M+ views within two weeks”.Source
  • 08

    Aside's Y Combinator profile describes the AI browser, but its Launches section still shows “Aside: Live answers for enterprise tech sales”, an earlier product that listened to sales calls. The YC profile is where candidates for the five open roles, investors and other YC founders (often the first users of a new tool) check what a company does.

    Low impactLow effort First result in 1 week CEO

    First action

    Draft the browser launch post.

    How to measure it

    YC-sourced activations

    Now
    –
    Target
    300
    By
    60 days

    Steps

    1. 1

      Draft launch

      Numbers, demo and Windows.

      CEO 1 day

    2. 2

      Retire old launch

      Mark it as an earlier product.

      CEO 1 hour

    3. 3

      Founder offer

      Bonus credits for YC founders.

      GTM 1 day

    4. 4

      Track

      A tagged link for the YC channel.

      GTM 1 hour

    5. 5

      Follow up

      Interview active YC users.

      CPO 2 weeks

    6. 6

      Turn into stories

      Ask the best for quotes.

      GTM 1 month

    First result
    1 week
    Cost
    Credit giveaway only
    Owner
    CEO

    Smallest test

    Update the launch section today.

    Why it works

    Post a YC launch for the browser (the June launch numbers, the Windows release, the benchmark results, a short demo) and retire or mark the earlier launch as past. Offer YC founders an extended trial or bonus credits through the launch, since founders who use Aside every day are likely to become paying users and to recommend it.

    It removes a confusing first impression for candidates and opens a high-trust channel of founders who try new tools early.

    Risks and how to handle them

    • Low response

      Include a demo clip and a concrete task founders do weekly.

    Evidence

    • Y CombinatorYC company profile: Fall 2025, team of 8, San Francisco; describes the AI browser, but its Launches section still shows “Aside: Live answers for enterprise tech sales”, an earlier product that listened to sales calls.Source
    • HiringGTM role (San Francisco, $100K–$180K): the first GTM hire; the June 2026 launch brought “20K+ users within two weeks”; “This is not a narrow performance marketing role”; owns acquisition, activation and retention: launches with demos, landing pages and videos; founder-led posts, user stories, short-form video and UGC; paid tests on Meta, TikTok and Google; referrals, SEO, partnerships and community.Source

This quarter 9

The core plays for the next three months.

  • 09

    Aside reports more than 50k users growing 25% a week, yet the homepage and feature pages show no testimonials, and the use-case posts use a hypothetical account. The GTM role names “user stories” as part of the job. A reviewer's verdict, “promising but unproven”, is exactly what real stories answer.

    High impactMedium effort First result in 1 month GTM

    First action

    Find the 30 users with the most routines run in the last month.

    How to measure it

    Published user stories

    Now
    –
    Target
    10
    By
    60 days

    Steps

    1. 1

      Find users

      Top routine users, with their consent.

      MTS, Product 2 days

    2. 2

      Invite

      Credits as a thank-you.

      GTM 1 week

    3. 3

      Interview

      Task, time saved and screen recording.

      GTM 3 weeks

    4. 4

      Publish

      /stories plus quotes on key pages.

      GTM 1 week

    5. 5

      Cut clips

      Short video per story.

      GTM 2 weeks

    6. 6

      Measure

      Conversion on pages with quotes.

      GTM monthly

    First result
    1 month
    Cost
    Credits and staff time
    Owner
    GTM

    Smallest test

    Publish three stories first.

    Why it works

    Recruit ten active users from Discord and usage data across the five use cases. For each, record the task, the routine, the minutes saved per week and a short screen recording. Publish them as /stories, add three quotes to the homepage and pricing page, and cut each into a short clip.

    Proof from people in similar jobs is the strongest persuasion for a new category, and it turns the use-case posts from instructions into evidence.

    Risks and how to handle them

    • Users' work data appears in recordings

      Record on sample data and get written approval.

    Evidence

    • HiringMTS, Product and MTS, Agent Harness: “More than 50k” users, “growing 25% week over week”, a team of 8; the product engineer turns community feedback into fixes (“fix every issue that comes up”); the harness engineer cuts token use, cost and latency and builds benchmarks to catch regressions (“No benchmaxxing”).Source
    • Your websiteBrowser agent page: “The SOTA browser agent”; tasks such as research, updating a spreadsheet, follow-ups and signing in to email, dashboards and internal tools; Ultrabrowse for deep research; use cases for operations, hiring, sales, developers and researchers; no testimonials.Source
    • Your websiteBlog: eight entries, the newest a July 10, 2026 link to 9to5Mac coverage; the five use-case posts carry copyable prompts but no customer stories (the sales post uses a hypothetical account).Source
    • HiringGTM role (San Francisco, $100K–$180K): the first GTM hire; the June 2026 launch brought “20K+ users within two weeks”; “This is not a narrow performance marketing role”; owns acquisition, activation and retention: launches with demos, landing pages and videos; founder-led posts, user stories, short-form video and UGC; paid tests on Meta, TikTok and Google; referrals, SEO, partnerships and community.Source
  • 10

    Aside's password manager is free, imports from six managers in one click and lets agents sign in without seeing passwords, with scoped access and an audit log. It already has its own comparison table against 1Password and Chrome. Yet it is sold only as a feature of the browser, while people setting up any agent are searching for a safe way to give it logins.

    Medium impactMedium effort First result in 6 weeks GTM

    First action

    Check search volume for agent password and credential queries.

    How to measure it

    Vault-first installs

    Now
    –
    Target
    2,000 a month
    By
    1 quarter

    Steps

    1. 1

      Size intent

      Search and community questions about agent logins.

      GTM 2 days

    2. 2

      Landing page

      Vault-first page with the comparison table.

      GTM 1 week

    3. 3

      Vault-first onboarding

      Import first for this entry point.

      Founding Designer 2 weeks

    4. 4

      Write the guide

      Safe logins for agents.

      GTM 1 week

    5. 5

      Paid test

      Search ads on agent-credential intent.

      GTM 3 weeks

    6. 6

      Measure

      Vault-to-agent activation.

      MTS, Product monthly

    First result
    6 weeks
    Cost
    Ad test budget
    Owner
    GTM

    Smallest test

    Run one ad group to the existing password-manager page for two weeks.

    Why it works

    Position the password manager as its own product: a landing page for “password manager for AI agents”, a guide for people running other agents who need logins handled safely, and paid tests on that search intent. Make import the first onboarding step for this audience, then introduce the browser agent once their vault is in Aside.

    Moving a vault is a heavy commitment that makes users stay. It reaches people worried about agent security before they are ready for an agent browser.

    Risks and how to handle them

    • Comparison claims about other managers

      Keep the table factual and dated.

    Evidence

    • Your websitePassword manager page: “The first password manager built for agents”; agents sign in by autofill without seeing passwords; scoped access and an audit log; free; one-click import from 1Password, Apple Passwords, Bitwarden, Dashlane, LastPass and Chrome; a comparison table against 1Password and Chrome Password Manager.Source
    • Competitor researchRivals' own homepages: Comet, “The browser that works for you”, on Mac, Windows, iOS and Android; Dia, “More than a browser. A better workday.”, free, with Dia for Work adding SSO and admin tools; BrowserOS, a free open-source browser that agents control with your existing logins, on macOS, Windows and Linux.Source
  • 11

    Aside already ships a local MCP server (aside mcp), a CLI and a REPL, and a use-case post covers CI triage and staging checks. But the developer page has no plan guidance and no call to sign up, and a free open-source rival pitches itself directly as the browser for coding agents. Developers who connect Aside to their own agents are a distribution channel that is not being used.

    Medium impactMedium effort First result in 1 month MTS

    First action

    List the directories and submit the MCP server to the top three.

    How to measure it

    Developer-surface weekly actives

    Now
    –
    Target
    +50%
    By
    1 quarter

    Steps

    1. 1

      List directories

      MCP registries and agent-tool lists.

      MTS, Agent Harness 1 day

    2. 2

      Submit

      With setup snippets.

      MTS, Agent Harness 1 week

    3. 3

      Write recipes

      Three developer workflows.

      MTS, Agent Harness 2 weeks

    4. 4

      Plan note

      Plan guidance on /developers.

      GTM 1 day

    5. 5

      Track

      CLI- and MCP-first installs.

      MTS, Product 1 week

    6. 6

      Iterate

      Fix the top setup issue each month.

      MTS, Agent Harness monthly

    First result
    1 month
    Cost
    Staff time only
    Owner
    MTS, Agent Harness

    Smallest test

    Submit to one directory and publish one recipe.

    Why it works

    Submit Aside to the main MCP server directories and agent-tool lists, add copy-paste setup snippets for the most-used coding agents and clients, and publish three recipes (visual regression check, logged-in dashboard scrape, release evidence). Add a plan note to /developers (which plan suits CLI usage) and track installs that start from the CLI.

    Developers find tools through directories and recipes, and an MCP connection makes Aside part of their daily workflow.

    Risks and how to handle them

    • Engineering time pulled from the harness

      Cap at one day a week.

    Evidence

    • Your websiteDeveloper docs: a CLI for natural-language tasks, a local MCP server (aside mcp), a REPL for deterministic browser steps and a public download API, plus an llms.txt for agents; no pricing or plan guidance on the page.Source
    • Competitor researchRivals' own homepages: Comet, “The browser that works for you”, on Mac, Windows, iOS and Android; Dia, “More than a browser. A better workday.”, free, with Dia for Work adding SSO and admin tools; BrowserOS, a free open-source browser that agents control with your existing logins, on macOS, Windows and Linux.Source
    • Your websiteSitemap (17 URLs, checked November 9, 2026): home, download, developers, pricing, three feature pages (browser agent, memory, password manager), blog, manifesto, two policies, one benchmark post and five use-case posts (developers, HR, operations, researchers, sales) dated June 22, 2026. No comparison, switching, FAQ, customer-story or Windows pages.Source
  • 12

    Aside grew to 20K+ users in two weeks largely by word of mouth and now reports 25% weekly growth, and the GTM role lists referrals as a channel to build. But the help center has no referral programme. Credits are the natural reward: they cost Aside only compute, and purchased credits never expire.

    High impactMedium effort First result in 1 month GTM

    First action

    Decide the credit amount from the median cost of a first task.

    How to measure it

    Referred share of activations

    Now
    –
    Target
    20%
    By
    1 quarter

    Steps

    1. 1

      Set the reward

      Credits on first completed task.

      CPO 1 day

    2. 2

      Build links

      Referral codes and attribution.

      MTS, Product 1 week

    3. 3

      Success prompt

      After a routine finishes.

      Founding Designer 3 days

    4. 4

      Fraud limits

      Caps per account and device.

      MTS, Product 2 days

    5. 5

      Announce

      Discord, X and email.

      GTM 1 day

    6. 6

      Tune

      Reward size versus referred-user retention.

      GTM monthly

    First result
    1 month
    Cost
    Compute cost of bonus credits
    Owner
    GTM

    Smallest test

    Give links to Discord members for two weeks.

    Why it works

    Launch “give credits, get credits”: each user gets a link; when a referred friend completes a first task, both get bonus credits. Add a share prompt after a successful routine (“This saved you N minutes, send Aside to a colleague”) and a referral counter in Plan & Usage.

    Rewarding with the product's own usage unit is cheap and attracts people who will use it, and asking right after a success is when users are most willing to share.

    Risks and how to handle them

    • Credit farming

      Reward only after a real completed task, with caps.

    Builds on

    Evidence

    • HiringGTM role (San Francisco, $100K–$180K): the first GTM hire; the June 2026 launch brought “20K+ users within two weeks”; “This is not a narrow performance marketing role”; owns acquisition, activation and retention: launches with demos, landing pages and videos; founder-led posts, user stories, short-form video and UGC; paid tests on Meta, TikTok and Google; referrals, SEO, partnerships and community.Source
    • Help centerHelp center index (docs.aside.com): 19 entries including routines, Ultrabrowse, devices, agent permissions, troubleshooting and two changelogs; no referral programme and no guide for a specific browser you might be switching from.Source
    • Help centerSubscription help article: does not define a credit or say how many a task uses (“Credit rates differ by model and speed”); users check charges under Settings > Plan & Usage; Free credits renew each UTC month; purchased add-on credits carry over and do not expire.Source
  • 13

    The launch reached 3M+ views and 20K+ users in two weeks, so short demos of Aside finishing real work spread well. The GTM role asks for founder-led posts, short-form video and UGC, plus paid tests on Meta, TikTok and Google. But the blog has had nothing new since July 10, so there is no steady stream.

    High impactMedium effort First result in 1 month GTM

    First action

    Record this week's clip from the sales use-case post.

    How to measure it

    Paid cost per activation

    Now
    –
    Target
    -25%
    By
    1 quarter

    Steps

    1. 1

      Pick the task

      One real, visual task a week.

      GTM weekly

    2. 2

      Record

      Founder voice, real screen.

      CEO weekly

    3. 3

      Post

      X, LinkedIn, TikTok and Reels.

      GTM weekly

    4. 4

      Rank

      By watch-through and installs.

      GTM weekly

    5. 5

      Promote

      Top two as ads.

      GTM weekly

    6. 6

      Report

      Cost per activation by clip.

      GTM monthly

    First result
    1 month
    Cost
    Ad test budget
    Owner
    GTM

    Smallest test

    Four clips in four weeks, one boosted.

    Why it works

    Each week, record a 45-second clip of Aside finishing one real task from the first-task gallery or a user story, posted by a founder on X and LinkedIn and cut for TikTok and Reels. Each Friday, the two clips with the best watch-through become ad variants pointing to the matching use-case page.

    Organic posts test hooks for free, so paid budget goes only to proven creative, and a steady stream of clips keeps the launch momentum going.

    Risks and how to handle them

    • Founder time

      Batch-record four clips in one session.

    Builds on

    Evidence

    • HiringMTS, Mobile Engineer (iOS): a new iPhone app to start browser-agent tasks, follow progress, approve actions and “use the result from anywhere”, with notifications, Live Activities and App Intents; the post says the launch reached “20K+ users and 3M+ views within two weeks”.Source
    • HiringGTM role (San Francisco, $100K–$180K): the first GTM hire; the June 2026 launch brought “20K+ users within two weeks”; “This is not a narrow performance marketing role”; owns acquisition, activation and retention: launches with demos, landing pages and videos; founder-led posts, user stories, short-form video and UGC; paid tests on Meta, TikTok and Google; referrals, SEO, partnerships and community.Source
    • Your websiteBlog: eight entries, the newest a July 10, 2026 link to 9to5Mac coverage; the five use-case posts carry copyable prompts but no customer stories (the sales post uses a hypothetical account).Source
  • 14

    Free allows “Up to 3 routines” and Pro offers unlimited routines. A user creating a fourth routine has already proven Aside saves them time every week, which makes them the best upgrade prospect. Today upgrades sit behind Settings > Plan & Usage, and credits renew monthly, so a Free user can simply wait.

    High impactMedium effort First result in 1 month MTS

    First action

    Count Free users who already have three routines.

    How to measure it

    Monthly Free-to-paid rate

    Now
    –
    Target
    +30%
    By
    1 quarter

    Steps

    1. 1

      Size moments

      Users at the routine limit and at 80% of credits.

      MTS, Product 2 days

    2. 2

      Design prompts

      Own usage, time saved and the yearly price.

      Founding Designer 1 week

    3. 3

      Build

      Three in-app triggers.

      MTS, Product 2 weeks

    4. 4

      Email backup

      Same message by email.

      GTM 3 days

    5. 5

      Test

      Against a holdout group.

      MTS, Product 4 weeks

    6. 6

      Keep winners

      Retire weak triggers.

      CPO monthly

    First result
    1 month
    Cost
    Staff time only
    Owner
    MTS, Product

    Smallest test

    Show the fourth-routine prompt only.

    Why it works

    Add three upgrade moments: when a user creates a fourth routine (“Keep all four running with Pro”), at 80% of monthly credits (showing the tasks they ran and the time saved), and after a third successful Ultrabrowse-eligible research task. Each shows the user's own usage and the yearly saving.

    Upgrade offers convert best when they arrive at a moment of proven value and show the user's own numbers.

    Risks and how to handle them

    • Feels pushy

      One prompt per trigger per month.

    Builds on

    Evidence

    • Your websitePricing page: Free “$0/forever” with “500 credits per month” and “Up to 3 routines”; Pro “$20/month” with “3x Free usage”, Ultrabrowse, unlimited routines, “Channels for remote control” and “Cloud handoff” marked “Coming Soon”; Max “$200/month” with “40x Free usage”; Enterprise by “Contact sales” for teams on shared company accounts; a Monthly/Yearly switch labelled “-20% off” with no yearly amounts shown; no FAQ.Source
    • Help centerSubscription help article: does not define a credit or say how many a task uses (“Credit rates differ by model and speed”); users check charges under Settings > Plan & Usage; Free credits renew each UTC month; purchased add-on credits carry over and do not expire.Source
  • 15

    Routines are what bring users back each week (scheduled pipeline reviews, invoice downloads, renewals, dashboard checks all appear in the use-case posts), but each user has to write them from scratch. The help center explains how routines work, but offers no ready-made templates.

    High impactMedium effort First result in 1 month MTS

    First action

    Turn the recurring tasks in the five use-case posts into ten templates.

    How to measure it

    Day-14 routine adoption

    Now
    –
    Target
    +40%
    By
    1 quarter

    Steps

    1. 1

      Draft templates

      Ten from the use-case posts.

      GTM 3 days

    2. 2

      Harden them

      Reliability on common sites.

      MTS, Agent Harness 1 week

    3. 3

      Gallery

      Browse and one-click install.

      MTS, Product 2 weeks

    4. 4

      Offer after first task

      Two relevant templates.

      Founding Designer 3 days

    5. 5

      User templates

      Publish and share.

      MTS, Product 3 weeks

    6. 6

      Measure

      Day-14 and day-30 retention.

      MTS, Product monthly

    First result
    1 month
    Cost
    Staff time only
    Owner
    MTS, Product

    Smallest test

    Five templates offered to half of new users.

    Why it works

    Ship 20 routine templates grouped by role, each with a schedule, the sites it needs and a sample output. Offer two templates right after a user's first successful task, and let users publish their own templates to a gallery, credited with their name if they choose.

    Scheduled work brings users back without a reminder, and templates turn one success into a weekly habit.

    Risks and how to handle them

    • Templates break when sites change

      Run a nightly check on each template.

    Builds on

    Evidence

    • Your websiteBlog: eight entries, the newest a July 10, 2026 link to 9to5Mac coverage; the five use-case posts carry copyable prompts but no customer stories (the sales post uses a hypothetical account).Source
    • Help centerHelp center index (docs.aside.com): 19 entries including routines, Ultrabrowse, devices, agent permissions, troubleshooting and two changelogs; no referral programme and no guide for a specific browser you might be switching from.Source
    • Your websitePricing page: Free “$0/forever” with “500 credits per month” and “Up to 3 routines”; Pro “$20/month” with “3x Free usage”, Ultrabrowse, unlimited routines, “Channels for remote control” and “Cloud handoff” marked “Coming Soon”; Max “$200/month” with “40x Free usage”; Enterprise by “Contact sales” for teams on shared company accounts; a Monthly/Yearly switch labelled “-20% off” with no yearly amounts shown; no FAQ.Source
  • 16

    Aside ships fast (two changelogs on docs.aside.com, a product engineer tasked to “fix every issue that comes up”, a harness team cutting cost and latency), but the blog has had nothing new since July 10. Users who tried Aside at the June launch and hit “some little bugs” have no reason to think it has improved.

    Medium impactLow effort First result in 1 month GTM

    First action

    Count users active in June who have not opened Aside in 30 days.

    How to measure it

    Monthly reactivations

    Now
    –
    Target
    5% of dormant users
    By
    90 days

    Steps

    1. 1

      Segment

      Dormant users by last task type.

      MTS, Product 2 days

    2. 2

      Pick improvements

      Three visible fixes from the changelog.

      CPO 1 day

    3. 3

      Write

      Email, in-app note and blog post.

      GTM 2 days

    4. 4

      Deep links

      “Try this again” opens the task.

      MTS, Product 3 days

    5. 5

      Send

      Monthly.

      GTM monthly

    6. 6

      Measure

      Reactivation and unsubscribe rates.

      GTM monthly

    First result
    1 month
    Cost
    Email tool cost
    Owner
    GTM

    Smallest test

    One email to June sign-ups who are now dormant.

    Why it works

    Send a monthly email and in-app note: three tasks that now work better, one new routine template and one user story, with a “try this again” button that opens the task. Send a version to users inactive for 30+ days that names the task they last tried, if it has since improved, and mirror the email as a blog post.

    In a product improving this fast, telling lapsed users about specific fixes is the cheapest way to win them back, and it keeps the blog active.

    Risks and how to handle them

    • Unsubscribes

      Keep it short, monthly and specific.

    Evidence

    • Help centerHelp center index (docs.aside.com): 19 entries including routines, Ultrabrowse, devices, agent permissions, troubleshooting and two changelogs; no referral programme and no guide for a specific browser you might be switching from.Source
    • HiringMTS, Product and MTS, Agent Harness: “More than 50k” users, “growing 25% week over week”, a team of 8; the product engineer turns community feedback into fixes (“fix every issue that comes up”); the harness engineer cuts token use, cost and latency and builds benchmarks to catch regressions (“No benchmaxxing”).Source
    • Reviewseesel review (June 2026): “genuinely promising but unproven”; said there was no public pricing page (since added); early Hacker News users mention “some little bugs”; benchmark claims are self-reported; the approval step for sensitive actions is not spelled out in the privacy policy.Source
    • Your websiteBlog: eight entries, the newest a July 10, 2026 link to 9to5Mac coverage; the five use-case posts carry copyable prompts but no customer stories (the sales post uses a hypothetical account).Source
  • 17

    Aside's strongest selling copy sits in five blog posts dated June 22 (operations, hiring teams, sales, developers, researchers), each with concrete tasks and copyable prompts. As blog posts they have no demo video, proof or a download button near the top, and paid tests on Google are part of the GTM plan.

    Medium impactMedium effort First result in 6 weeks GTM

    First action

    Build /for/sales from the existing sales post.

    How to measure it

    Search cost per activation

    Now
    –
    Target
    Under $15
    By
    1 quarter

    Steps

    1. 1

      Template

      One landing layout.

      Founding Designer 3 days

    2. 2

      First page

      /for/sales.

      GTM 3 days

    3. 3

      Demos

      One clip per role.

      GTM 2 weeks

    4. 4

      Other pages

      Four more roles.

      GTM 2 weeks

    5. 5

      Ad tests

      Job-task keywords per page.

      GTM 4 weeks

    6. 6

      Keep winners

      Scale the best two roles.

      GTM monthly

    First result
    6 weeks
    Cost
    Ad test budget
    Owner
    GTM

    Smallest test

    One page and one ad group for three weeks.

    Why it works

    Create /for/sales, /for/operations, /for/recruiting, /for/research and /for/developers: a headline about the job, a 30-second demo, three one-click prompts that open in Aside, a quote from a user story and a download button. Run small Google tests on job-task searches (e.g. “automate CRM updates”, “download invoices automatically”) to each.

    Searches for a specific job task are high intent and cheaper than generic “AI browser” terms, and each page matches the visitor's exact job.

    Risks and how to handle them

    • Low search volume per task

      Group tasks by role and broaden only the winners.

    Builds on

    Evidence

    • Your websiteSitemap (17 URLs, checked November 9, 2026): home, download, developers, pricing, three feature pages (browser agent, memory, password manager), blog, manifesto, two policies, one benchmark post and five use-case posts (developers, HR, operations, researchers, sales) dated June 22, 2026. No comparison, switching, FAQ, customer-story or Windows pages.Source
    • HiringGTM role (San Francisco, $100K–$180K): the first GTM hire; the June 2026 launch brought “20K+ users within two weeks”; “This is not a narrow performance marketing role”; owns acquisition, activation and retention: launches with demos, landing pages and videos; founder-led posts, user stories, short-form video and UGC; paid tests on Meta, TikTok and Google; referrals, SEO, partnerships and community.Source
    • Your websiteBrowser agent page: “The SOTA browser agent”; tasks such as research, updating a spreadsheet, follow-ups and signing in to email, dashboards and internal tools; Ultrabrowse for deep research; use cases for operations, hiring, sales, developers and researchers; no testimonials.Source

Big bets 3

Larger plays that can change the trajectory.

  • 18

    Aside's headline claim is #1 on three browser-agent benchmarks (99.0% on Online-Mind2Web, 93.0% on BU-Bench-V1), but several competitor scores come from Aside's own runs with run conditions not described, and reviewers call the results self-reported. The harness role already builds evaluation environments and rejects “benchmaxxing”.

    High impactHigh effort First result in 1 quarter MTS

    First action

    Package the most recent Online-Mind2Web run logs.

    How to measure it

    Third-party reproductions

    Now
    –
    Target
    2
    By
    2 quarters

    Steps

    1. 1

      Package runs

      Logs, settings and task results.

      MTS, Agent Harness 2 weeks

    2. 2

      Run script

      One command to reproduce.

      MTS, Agent Harness 2 weeks

    3. 3

      Add cost

      Credits and seconds per task.

      MTS, Agent Harness 1 week

    4. 4

      Invite reviewers

      Benchmark authors and maintainers.

      CTO 1 month

    5. 5

      Scoreboard page

      Updated each release.

      GTM 1 week

    6. 6

      Announce

      With the reproductions.

      CEO 1 day

    First result
    1 quarter
    Cost
    Compute for public runs
    Owner
    MTS, Agent Harness

    Smallest test

    Publish the logs for one benchmark.

    Why it works

    Publish the evaluation setup: task lists, harness version, model, settings, run logs and per-task pass/fail, plus a script anyone can run. Invite an independent academic or open-source maintainer to reproduce the result, and update a public scoreboard with each harness release, including cost and time per task.

    A result others can reproduce turns a disputed claim into lasting credibility, and publishing cost per task backs up the pricing story.

    Risks and how to handle them

    • Reproduced scores come in lower

      Publish honestly; a reproduced 95% is worth more than a disputed 99%.

    Evidence

    • Your websiteBenchmark post (June 25, 2026): #1 on Online-Mind2Web (99.0%), BU-Bench-V1 (93.0%) and Odyssey; several competitor scores come from Aside's own runs, and run conditions are not described in detail.Source
    • Reviewseesel review (June 2026): “genuinely promising but unproven”; said there was no public pricing page (since added); early Hacker News users mention “some little bugs”; benchmark claims are self-reported; the approval step for sensitive actions is not spelled out in the privacy policy.Source
    • HiringMTS, Product and MTS, Agent Harness: “More than 50k” users, “growing 25% week over week”, a team of 8; the product engineer turns community feedback into fixes (“fix every issue that comes up”); the harness engineer cuts token use, cost and latency and builds benchmarks to catch regressions (“No benchmaxxing”).Source
  • 19

    Sensitive actions wait for the user's approval, and Ultrabrowse and routines run for minutes or hours. If the user has left their desk, a long task simply stops at the approval step. The iOS hire is building an app to start tasks, follow progress and approve actions, with Live Activities and notifications. Comet already runs on iOS and Android.

    High impactHigh effort First result in 3 months MTS

    First action

    Measure how many tasks today sit waiting for approval for over 10 minutes.

    How to measure it

    Long-task completion

    Now
    –
    Target
    +20 points
    By
    2 quarters after launch

    Steps

    1. 1

      Measure stalls

      Tasks waiting on approval.

      MTS, Product 2 days

    2. 2

      Scope v1

      Approvals, progress and results.

      MTS, Mobile Engineer (iOS) 1 week

    3. 3

      Build

      Sync, notifications and Live Activities.

      MTS, Mobile Engineer (iOS) 2 months

    4. 4

      Beta

      To the waitlist from idea 7.

      MTS, Mobile Engineer (iOS) 1 month

    5. 5

      Launch

      As a Pro feature.

      GTM 1 week

    6. 6

      Measure

      Completion and Pro upgrades.

      MTS, Product monthly

    First result
    3 months
    Cost
    One engineer for two quarters
    Owner
    MTS, Mobile Engineer (iOS)

    Smallest test

    Push-notification approvals through an existing channel before the app.

    Why it works

    Launch the iPhone app around one promise: “Your tasks don't stall when you leave your desk.” Version 1 handles approvals, progress through Live Activities and results; starting tasks comes next. Make it a Pro feature alongside Channels, and measure how much long-task completion rises for users with the app.

    Unblocking approvals directly raises task completion, the core value metric, and a phone app adds a daily touchpoint that keeps users coming back.

    Risks and how to handle them

    • Approving on a small screen is risky

      Show the full action and destination, with Face ID for payments.

    Builds on

    Evidence

    • HiringMTS, Mobile Engineer (iOS): a new iPhone app to start browser-agent tasks, follow progress, approve actions and “use the result from anywhere”, with notifications, Live Activities and App Intents; the post says the launch reached “20K+ users and 3M+ views within two weeks”.Source
    • Your websitePassword manager page: “The first password manager built for agents”; agents sign in by autofill without seeing passwords; scoped access and an audit log; free; one-click import from 1Password, Apple Passwords, Bitwarden, Dashlane, LastPass and Chrome; a comparison table against 1Password and Chrome Password Manager.Source
    • Competitor researchRivals' own homepages: Comet, “The browser that works for you”, on Mac, Windows, iOS and Android; Dia, “More than a browser. A better workday.”, free, with Dia for Work adding SSO and admin tools; BrowserOS, a free open-source browser that agents control with your existing logins, on macOS, Windows and Linux.Source
  • 20

    Pricing jumps from individual plans ($20 and $200) straight to a sales-led Enterprise tier, whose features (shared profiles, vaults, team logins without exposing secrets, seat and billing management) are exactly what a five-person ops or sales team wants. Dia already sells Dia for Work with SSO and admin tools. Small teams that will not book a sales call have no plan to buy.

    High impactHigh effort First result in 1 quarter CEO

    First action

    Count Enterprise enquiries from companies with under 20 people.

    How to measure it

    Team MRR

    Now
    –
    Target
    15% of total MRR
    By
    2 quarters

    Steps

    1. 1

      Size demand

      Small-company Enterprise enquiries.

      CEO 1 week

    2. 2

      Define the plan

      Seats, pooled credits, shared vault.

      CPO 2 weeks

    3. 3

      Build admin

      Invites, roles and billing.

      MTS, Product 6 weeks

    4. 4

      Design

      Team setup flow.

      Founding Designer 3 weeks

    5. 5

      Launch

      On pricing and to Pro users.

      GTM 1 week

    6. 6

      Review

      Seat growth and Enterprise upgrades.

      CEO monthly

    First result
    1 quarter
    Cost
    Engineering for one quarter
    Owner
    CEO

    Smallest test

    Offer a manual Team invoice to ten small-company enquiries.

    Why it works

    Launch Team at a per-seat price with pooled credits, a shared vault, shared routines and basic admin, bought by card. Keep Enterprise for SSO, cookies and history shared across the company, cron-triggered runs and rollout support. Put Team on /pricing and invite Pro users who share routines or logins with colleagues.

    Per-seat team plans expand revenue as usage spreads inside a company, and shared vaults and routines are stickier than individual use.

    Risks and how to handle them

    • Team cannibalises Enterprise

      Keep SSO, cron runs and rollout support on Enterprise.

    Builds on

    Evidence

    • Your websitePricing page: Free “$0/forever” with “500 credits per month” and “Up to 3 routines”; Pro “$20/month” with “3x Free usage”, Ultrabrowse, unlimited routines, “Channels for remote control” and “Cloud handoff” marked “Coming Soon”; Max “$200/month” with “40x Free usage”; Enterprise by “Contact sales” for teams on shared company accounts; a Monthly/Yearly switch labelled “-20% off” with no yearly amounts shown; no FAQ.Source
    • Competitor researchRivals' own homepages: Comet, “The browser that works for you”, on Mac, Windows, iOS and Android; Dia, “More than a browser. A better workday.”, free, with Dia for Work adding SSO and admin tools; BrowserOS, a free open-source browser that agents control with your existing logins, on macOS, Windows and Linux.Source

Wildcards 2

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

  • 21

    Aside's download page sends new users to Discord, and every job post says the team talks to users through the community. With a team of 8 and 50k+ users, the most active community members are already doing unpaid support, bug-finding and teaching. A formal programme can make them a growth channel.

    Medium impactMedium effort First result in 1 month GTM

    First action

    List the 30 members with the most helpful answers in Discord.

    How to measure it

    Community-made templates

    Now
    –
    Target
    50
    By
    1 quarter

    Steps

    1. 1

      Pick members

      Most helpful in Discord.

      GTM 2 days

    2. 2

      Define perks

      Early builds, credits, founder calls.

      CEO 1 day

    3. 3

      Invite

      Personal messages.

      GTM 1 week

    4. 4

      Monthly call

      Roadmap and feedback.

      CPO monthly

    5. 5

      Showcase

      Feature their templates and clips.

      GTM monthly

    6. 6

      Review

      Output and retention of the group.

      GTM quarterly

    First result
    1 month
    Cost
    Credits and founder time
    Owner
    GTM

    Smallest test

    Invite ten members for one month.

    Why it works

    Invite the 30 most helpful Discord members to an “Aside Builders” group: early builds, a monthly call with the founders, bonus credits, and credit on routine templates they publish. Ask each to publish one template and one short clip a month, and to answer new-user questions.

    Power users create the templates, tutorials and social proof a small team cannot, and recognition keeps them loyal.

    Risks and how to handle them

    • Early builds leak bugs publicly

      Ask members not to share before release, and keep the group small.

    Builds on

    Evidence

    • Help centerGet started: “Open Aside”, “Sign in or create an account”, “Import browser data” (browsing history, cookies and bookmarks; Safari by ZIP export), “Set up Password Manager”, “Start a task”, with no suggested first tasks. The download page does not mention importing and points to Discord, X and the docs.Source
    • HiringMTS, Product and MTS, Agent Harness: “More than 50k” users, “growing 25% week over week”, a team of 8; the product engineer turns community feedback into fixes (“fix every issue that comes up”); the harness engineer cuts token use, cost and latency and builds benchmarks to catch regressions (“No benchmaxxing”).Source
    • HiringGTM role (San Francisco, $100K–$180K): the first GTM hire; the June 2026 launch brought “20K+ users within two weeks”; “This is not a narrow performance marketing role”; owns acquisition, activation and retention: launches with demos, landing pages and videos; founder-led posts, user stories, short-form video and UGC; paid tests on Meta, TikTok and Google; referrals, SEO, partnerships and community.Source
  • 22

    Aside's manifesto argues that the browser is “the operating system for AI”, and its harness team already runs evaluations across live websites. Nobody publishes a plain-language, regularly updated answer to the question buyers actually ask: can an agent do my weekly task in the tools I use?

    Medium impactMedium effort First result in 2 months CTO

    First action

    Choose the 100 tasks from the use-case posts and user stories.

    How to measure it

    Referring domains to the index

    Now
    –
    Target
    50
    By
    2 quarters

    Steps

    1. 1

      Choose tasks

      100 everyday tasks.

      GTM 1 week

    2. 2

      Run monthly

      Using the existing evaluation setup.

      MTS, Agent Harness monthly

    3. 3

      Grade simply

      Reliable, assisted or not yet.

      CTO monthly

    4. 4

      Publish

      Index page with trend charts.

      GTM 1 week

    5. 5

      Pitch

      Tech press and newsletters.

      CEO monthly

    6. 6

      Link templates

      Each task to a routine.

      MTS, Product 2 weeks

    First result
    2 months
    Cost
    Compute and staff time
    Owner
    CTO

    Smallest test

    Publish 25 tasks once and measure links.

    Why it works

    Publish a monthly “Agent-ready web” index: 100 everyday tasks across common categories (CRM updates, invoice downloads, scheduling, research), each marked as done reliably, done with help, or not yet, with the trend over time. Present results by task and category, without listing site weaknesses. Each task links to a routine template.

    A regularly updated public index earns press, links and AI-assistant citations, sets Aside as the category's reference, and turns each task into a sign-up.

    Risks and how to handle them

    • Seen as self-serving

      Publish the method and include tasks Aside cannot yet do.

    Builds on

    Evidence

    • Your websiteBenchmark post (June 25, 2026): #1 on Online-Mind2Web (99.0%), BU-Bench-V1 (93.0%) and Odyssey; several competitor scores come from Aside's own runs, and run conditions are not described in detail.Source
    • HiringMTS, Product and MTS, Agent Harness: “More than 50k” users, “growing 25% week over week”, a team of 8; the product engineer turns community feedback into fixes (“fix every issue that comes up”); the harness engineer cuts token use, cost and latency and builds benchmarks to catch regressions (“No benchmaxxing”).Source
    • Your websiteBrowser agent page: “The SOTA browser agent”; tasks such as research, updating a spreadsheet, follow-ups and signing in to email, dashboards and internal tools; Ultrabrowse for deep research; use cases for operations, hiring, sales, developers and researchers; no testimonials.Source
More example playbooks and reviews

More growth playbooks

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

  • Imagine AI growth strategyA data-driven agent that runs LinkedIn for B2B companies. It drafts posts and graphics in each person's voice, comments, sends connection…23 growth ideas
  • Genius Sports growth strategyOfficial sports data, technology and media: live data and odds for sportsbooks, performance and officiating tools for leagues, and sports…21 growth ideas
  • Yazio growth strategyA calorie and nutrition tracking app with AI photo tracking, barcode scanning, intermittent fasting timers, recipes and fitness tracker…21 growth ideas
  • Liftoff growth strategyAn AI-powered performance marketing platform for mobile apps: user acquisition (Accelerate), publisher monetization (Monetize, Vungle…21 growth ideas
  • Brillio growth strategyA digital and AI services firm that builds agentic AI, data, cloud, customer experience and engineering solutions, positioned as "The…20 growth ideas
  • Tinuiti growth strategyThe largest independent full-funnel media agency in the US: commerce and Amazon, retail media, search, social, TV, audio and display…21 growth ideas
  • RareGlo Organic Shea Products growth strategyA small-batch maker of vegan shea skin and hair butters, butter bars, cleansing bars, curl care, fragrance oils and mists in Gaithersburg…20 growth ideas
  • IHS | Industrial Heat Sources growth strategyAn authorized distributor of industrial hot air tools, heaters, blowers and plastic welding equipment from 19 brands, the maker of its own…24 growth ideas

See every playbook and website review

Independent research. Programmable wrote this playbook from public sources, as they stood on November 10, 2026, and cites them throughout. Aside did not commission, review or endorse it, and Programmable is not affiliated with Aside. 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 Aside runs its business.

Something wrong, or do you represent Aside? 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

    Credits explained on pricing, answering searches about Aside on aside.com, switching guides, a first-task gallery, safety answers, a Windows page, waitlists for coming-soon features and the YC profile.

  2. Days 31 to 609 ideas

    User stories, the password manager as a way in, developer distribution, referrals, founder clips into paid ads, upgrade moments, routine templates, a monthly what's-new email and use-case landing pages.

  3. Days 61 to 905 ideas

    Commit to the big bets (reproducible benchmarks, phone approvals, a team plan) and test the community and public-index wildcards.

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 an agent that works in the sites they are already signed into, with no integrations. The most cited problem: repetitive browser work across dashboards, CRMs and admin tools.

  • Local memory and data kept on their own device
  • Not knowing what usage-based AI will cost
  • In their words: “browser agent”

Market

Competitive landscape

5 competitors compared: Comet, Dia, BrowserOS, Tabbit and 1 more.

  • Gap to exploit: Agent browsing with clear, predictable usage costs
  • Comet: Consumer and search focus rather than work across logged-in tools
  • Dia: Reads context more than it completes multi-step tasks

Hiring signals

Aside, a team of 8, is hiring for the next stage after a viral launch.

  • GTM: First GTM hire owning acquisition, activation and retention. Ideas 1, 2, 9, 12, 13, 17.
  • MTS - Product: 50k+ users, 25% weekly growth; fix what the community reports. Ideas 14, 16.
  • MTS - Agent Harness: Cost, latency and real evaluations, “No benchmaxxing”. Ideas 11, 18, 22.

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