ProgrammableMarketing/capy.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 · 9
  • Big bets · 3
  • Wildcards · 3
  • How we can help
Generated Oct 26, 2026 2 views
  1. Programmable
  2. Playbooks
  3. Capy
Capy logo

growth playbook · Oct 26, 2026

capy.ai (opens in a new tab)

Capy growth playbook

Diagnosis

Capy is a small, Y Combinator-backed San Francisco team with a fast release pace, 70,000+ engineers claimed, and one excellent customer story (Goblins: ~60 to 300+ merged PRs a week).

Its new growth hire is asked to get developers to find Capy, try it and bring their team, with content, events, outreach automation and a funnel measured to the first merged PR.

The gaps are proof and paths: 11 logos without stories, articles stalled since July, use cases buried under anchors, an open source showcase that is promised but missing, and a homepage that fails Core Web Vitals for real visitors. This playbook fills the proof first, instruments activation around the first merged PR, and then builds events, signal outreach, benchmarks and an automation library.

Growth Score

7/10

Biggest Opportunity

Capy already has the numbers that win developers: merged PRs and reviews per customer. Publishing them for every logo, every team and its own engineering turns a crowded market's claims into proof.

Biggest Risk

AI coding is crowded and well funded (Devin, Codex cloud, Copilot coding agent, Cursor and others), and model providers keep adding agent features; without visible proof and fast activation, small teams struggle to be heard.

01

Start here

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

  1. 1

    Turn the 11 logos without a story into Goblins-style case studies, with numbers Capy already has

    Ask the three most active logo customers for a 20-minute call and permission to use their PR numbers.

    High impact First result in 3 weeks
  2. 2

    Instrument "time to first merged PR" and design the $1 trial around it

    Query the median hours from signup to first merged PR for last month's trials.

    High impact First result in 3 weeks
  3. 3

    Fix the homepage on phones: real visitors fail Core Web Vitals

    Replace the largest code demo with a video on mobile and re-run PageSpeed.

    Medium impact First result in 1 month

02

Your idea portfolio

21 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.

  • 6 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 · 1Big bets · 9Fill-ins · 7Reconsider · 4
Effort →

Showing 21 of 21 ideas

This week 6

Quick wins with a clear first action.

  • 01

    The homepage says "Loved by 70,000+ engineers" and shows 12 customer logos, but the customers page has exactly one story. That one story, Goblins, is the strongest page on the site: ~60 to 300+ merged PRs a week, 98% of merged PRs built or reviewed by Capy, 14,000+ reviews, and a video. Supermemory, RunPod, Firecrawl, Exa, Million, Autumn and the rest are on the homepage with no proof behind them, in a market the engineering post itself calls crowded.

    High impactMedium effort First result in 3 weeks Member of Growth Staff

    First action

    Ask the three most active logo customers for a 20-minute call and permission to use their PR numbers.

    How to measure it

    Published stories

    Now
    –
    Target
    8
    By
    90 days

    Steps

    1. 1

      Rank the logos

      By usage and willingness; start with the three most active teams.

      Member of Growth Staff 1 day

    2. 2

      Pull the numbers

      Merged PRs before and after, share built or reviewed by Capy, reviews run.

      Member of Technical Staff 2 days

    3. 3

      Interview

      Pain, switch moment, what changed in the team's week.

      Member of Growth Staff 1 week

    4. 4

      Film

      A short screen-and-face video like the Goblins one.

      Member of Growth Staff 1 week

    5. 5

      Publish weekly

      Same layout as Goblins, with a team-size and use-case filter.

      Member of Growth Staff weekly

    6. 6

      Link the logos

      Each homepage logo links to its story with the headline number.

      Member of Technical Staff 1 day

    7. 7

      Reuse

      Clips for X, Discord and outbound sequences.

      Member of Growth Staff ongoing

    8. 8

      Measure

      Signup and paid conversion for sessions that view a story.

      Member of Growth Staff monthly

    First result
    3 weeks
    Cost
    Staff time only
    Owner
    Member of Growth Staff

    Smallest test

    Publish one more story with PR numbers in the next two weeks and compare its traffic to Goblins.

    Why it works

    Use the Goblins template for every logo: before and after merged PRs, share built or reviewed by Capy, review count, the pain that triggered the switch, and a short video. Pull the PR numbers from Capy's own data (with each customer's permission), so every story has hard figures instead of quotes. Ship one a week, add a filter by team size and use case, and put the headline number on each homepage logo.

    Developer buyers trust measured results from teams like theirs, and the growth role already owns customer calls and video. Each story also answers the "does this hold on a real codebase" question that comparison pages cannot.

    Ready to use

    ChecklistCustomer story kit

    [ ] Permission to use PR metrics
    [ ] Merged PRs per week, before and after
    [ ] Share of merged PRs built or reviewed by Capy
    [ ] Reviews run
    [ ] The pain that triggered the switch
    [ ] Three quotes from engineers
    [ ] 2-3 minute video
    [ ] Industry, team size, features used
    [ ] Link from the homepage logo

    Risks and how to handle them

    • Customers decline to share numbers

      Offer ranges or percentages, and credits as thanks.

    • Stories look the same

      Lead each with a different use case: review, bug triage, migrations, E2E testing.

    Evidence

    • Your websiteHomepage: H1 "The best AI software engineer"; "Orchestrate a fleet of cloud coding agents in all of the apps your team already lives in."; "Loved by 70,000+ engineers at" followed by 12 logos (Y Combinator, Supermemory, RunPod, Revyl, Million, Loop, Goblins, Firecrawl, Finta, Exa, Autumn, Aside AI). It shows Terminal-Bench results, bring-your-own-subscription, automations, a review agent and a Slack integration.Source
    • Your websiteCustomers page (checked October 23, 2026): the filter reads "All [1]"; the only story is Goblins. The Goblins story reports merged pull requests going from "~60" to "300+ a week" with the same small team, "1,363 (98%)" of 1,392 merged PRs built or reviewed by Capy between September 10 and October 10, 2026, and "more than 14,000 reviews", with a 2:42 video.Source
    • HiringMember of Growth Staff (San Francisco, posted October 6, 2026; $200K-$250K plus 0.5%-1% equity): "Growth at Capy is about getting developers to find Capy, try it, and bring their team." The role builds "Outreach and lifecycle automation, enrichment and qualification pipelines, landing pages, onboarding flows, demos", talks to customers "a lot", makes "Articles, videos, and demos about Capy ... launch videos, side-by-side comparisons, and benchmarks included", and runs "Hackathons, meetups, dinners, and workshops, often with other startups and developer communities".Source
    • HiringMember of Technical Staff (posted October 6, 2026): "AI coding is a crowded market now, and developers expect first-minute magic and a quality bar that holds across every surface." The role spans the React web app, Electron desktop app, agent loop, backend and Kubernetes infrastructure, and "Talk to users". Capy says "We build Capy with Capy: our own PRs, reviews, and automations run through it."Source
  • 02

    The growth post asks for exactly this: how fast developers "reach a merged PR". The $1 trial lasts 7 days, so a developer who has not merged anything by day 2 is unlikely to keep paying $20 a month.

    High impactMedium effort First result in 3 weeks Member of Growth Staff

    First action

    Query the median hours from signup to first merged PR for last month's trials.

    How to measure it

    Trials merging a PR in 48 hours

    Now
    –
    Target
    60%
    By
    60 days

    Steps

    1. 1

      Define the event

      First merged PR from a Capy thread.

      Member of Growth Staff 1 day

    2. 2

      Baseline

      Median time and share by entry point.

      Member of Growth Staff 2 days

    3. 3

      Starter tasks

      Three safe tasks generated from the connected repo.

      Member of Technical Staff 1 week

    4. 4

      Nudges

      Email or Slack at 24 and 72 hours with a suggested task.

      Member of Growth Staff 3 days

    5. 5

      Dashboard

      Daily trial cohort chart.

      Member of Growth Staff 2 days

    6. 6

      Test

      Starter tasks on for half of new trials.

      Member of Growth Staff 3 weeks

    7. 7

      Tie to paid

      Day-8 conversion by activation time.

      Member of Growth Staff monthly

    First result
    3 weeks
    Cost
    Engineering time
    Owner
    Member of Growth Staff

    Smallest test

    Hand-pick a starter task for the next 30 trials and compare their day-8 conversion.

    Why it works

    Define the activation event as the first merged PR from a Capy thread. Track hours from signup to that event, by entry point (web, desktop, Slack, Linear). Offer three ready starter tasks on the first screen (fix a lint error, add a missing test, update a stale README) that usually merge on the first try, and nudge by email or Slack at 24 and 72 hours if nothing has merged.

    One clear activation metric focuses every onboarding experiment, and small, safe first tasks give the first-minute magic the engineering post asks for.

    Risks and how to handle them

    • Trivial first tasks feel like a toy

      Pick tasks from the user's real repo and backlog.

    Evidence

    • HiringThe same post: "Own the funnel. How developers find Capy, what they see first, how fast they reach a merged PR, and why they bring their team. You instrument every step and improve it." and "Run experiments weekly. Ship, measure, keep what works, and kill what doesn't." The mandate is "revenue leverage".Source
    • Your websitePricing: Lite $20/month ($16 annual) with "$20 credits", Pro $100/month with $105 credits, Max $200 to $1,000 in $100 steps with 10% bonus credits, and Enterprise custom; "unlimited members on every one", "no per-seat price"; "Try Capy Lite for $1 for 7 days" with a verified card, then $20/month. Credits "cover AI model costs and VM compute time" and "don't roll over".Source
    • HiringMember of Technical Staff (posted October 6, 2026): "AI coding is a crowded market now, and developers expect first-minute magic and a quality bar that holds across every surface." The role spans the React web app, Electron desktop app, agent loop, backend and Kubernetes infrastructure, and "Talk to users". Capy says "We build Capy with Capy: our own PRs, reviews, and automations run through it."Source
    • Your websiteHomepage demo of an automated run shows a per-task cost ("Find critical bugs $0.49"), and the use-cases page promises "25+ agents per repo, coordinated across branches", "Answers review feedback without pings" and "Rebases, CI failures, and merge conflicts handled".Source
  • 03

    Real Chrome users wait 3.2 s for the homepage's main content and see 288 ms input delay, failing both Core Web Vitals. The lab shows 11.3 s LCP on mobile, 7,438 DOM elements from the inline code and chat demos, and 251 KB of unused JavaScript. Developers often first see Capy on a phone from a link on X or Discord.

    Medium impactMedium effort First result in 1 month Member of Technical Staff

    First action

    Replace the largest code demo with a video on mobile and re-run PageSpeed.

    How to measure it

    Homepage field LCP

    Now
    –
    Target
    under 2.5 s
    By
    60 days

    Steps

    1. 1

      Profile

      Which demo blocks the main thread longest.

      Member of Technical Staff 1 day

    2. 2

      Swap on mobile

      Video or static frames for code and chat demos.

      Member of Technical Staff 3 days

    3. 3

      Split JavaScript

      Load demo code on interaction.

      Member of Technical Staff 2 days

    4. 4

      Cache

      Long lifetimes for static assets.

      Member of Technical Staff 2 hours

    5. 5

      Accessibility

      Contrast and aria-hidden focus fixes.

      Member of Technical Staff 1 day

    6. 6

      Verify

      PageSpeed and Chrome UX Report after 28 days.

      Member of Growth Staff monthly

    First result
    1 month
    Cost
    Engineering time
    Owner
    Member of Technical Staff

    Smallest test

    Ship the mobile video swap for one demo and compare mobile signup rate for a week.

    Why it works

    Render the interactive demos as lightweight video or static images on mobile and load the live versions only on interaction; split the demo JavaScript; set long cache lifetimes for static assets; and fix the five accessibility issues Lighthouse lists. Re-test until real-user LCP is under 2.5 s.

    A slow first page undercuts a product sold on speed, and a lighter page lifts both signups and search visibility.

    Risks and how to handle them

    • Demos lose their wow

      Keep the live demos on desktop and behind a tap on mobile.

    Evidence

    • Google PageSpeedHomepage on mobile: performance 45/100 (desktop 62), largest contentful paint 11.3 s in the lab, 589 ms total blocking time, 7,438 DOM elements, 872 KB of JavaScript (251 KB unused), 1.8 MB page weight. Real Chrome users on the page: LCP 3.2 s and INP 288 ms, failing Core Web Vitals on both; origin-wide LCP is 2.4 s and 43% faster than before. Lighthouse lists 5 accessibility issues including colour contrast and focusable elements inside aria-hidden regions.Source
    • HiringMember of Technical Staff (posted October 6, 2026): "AI coding is a crowded market now, and developers expect first-minute magic and a quality bar that holds across every surface." The role spans the React web app, Electron desktop app, agent loop, backend and Kubernetes infrastructure, and "Talk to users". Capy says "We build Capy with Capy: our own PRs, reviews, and automations run through it."Source
    • Your websiteHomepage: H1 "The best AI software engineer"; "Orchestrate a fleet of cloud coding agents in all of the apps your team already lives in."; "Loved by 70,000+ engineers at" followed by 12 logos (Y Combinator, Supermemory, RunPod, Revyl, Million, Loop, Goblins, Firecrawl, Finta, Exa, Autumn, Aside AI). It shows Terminal-Bench results, bring-your-own-subscription, automations, a review agent and a Slack integration.Source
  • 04

    The Open Source Program offers maintainers "Community Recognition: Get featured in our open source showcase", but there is no showcase page anywhere on the site. The promise costs nothing to keep and is a reason for maintainers to apply and to talk about it.

    Medium impactLow effort First result in 2 weeks Member of Growth Staff

    First action

    List the projects already in the program and ask five maintainers for a quote.

    How to measure it

    Program applications

    Now
    –
    Target
    +50%
    By
    90 days

    Steps

    1. 1

      List projects

      Current program members.

      Member of Growth Staff 1 day

    2. 2

      Collect quotes

      Five maintainers.

      Member of Growth Staff 1 week

    3. 3

      Pull stats

      Merged PRs and reviews per project.

      Member of Technical Staff 1 day

    4. 4

      Build the page

      Cards with stats and repo links.

      Member of Growth Staff 2 days

    5. 5

      Announce

      X and Discord posts tagging maintainers.

      Member of Growth Staff 1 day

    6. 6

      Refresh

      Add projects monthly.

      Member of Growth Staff monthly

    First result
    2 weeks
    Cost
    $0
    Owner
    Member of Growth Staff

    Smallest test

    Launch the page with five projects.

    Why it works

    Publish /open-source/showcase listing approved projects with what Capy did for each (merged PRs, reviews, migrations), a short maintainer quote and a link to the repo. Announce each new project on X and Discord and tag the maintainers.

    Maintainers share recognition with their communities, which reaches developers in their own feeds, and public repos make Capy's work visible and verifiable.

    Risks and how to handle them

    • Small projects look weak

      Sort by activity and show the work, not the stars.

    Evidence

    • Your websiteOpen Source Program page: free Capy for open source projects with an OSI-approved license, "No call needed", "Unlimited Contributors", and "Community Recognition: Get featured in our open source showcase and community highlights." No showcase page is linked or listed in the sitemap.Source
    • Your websiteSitemap (77 URLs on capy.ai, checked October 23, 2026): 24 articles, the newest dated July 9, 2026, almost all comparisons and lists (devin-vs-capy, codex-cloud-vs-capy, cursor-alternative, coderabbit-vs-capy, best-ai-coding-agents-2026 and others); 9 blog posts, the newest October 8; one customer story; one use-cases page whose sections (PR review, bug triage, documentation, migrations) are anchors rather than pages; no events, showcase or templates pages.Source
    • HiringMember of Growth Staff (San Francisco, posted October 6, 2026; $200K-$250K plus 0.5%-1% equity): "Growth at Capy is about getting developers to find Capy, try it, and bring their team." The role builds "Outreach and lifecycle automation, enrichment and qualification pipelines, landing pages, onboarding flows, demos", talks to customers "a lot", makes "Articles, videos, and demos about Capy ... launch videos, side-by-side comparisons, and benchmarks included", and runs "Hackathons, meetups, dinners, and workshops, often with other startups and developer communities".Source
  • 05

    Capy's 24 articles are mostly comparisons and "best of 2026" lists, and the newest is dated July 9, 2026. The use-cases page puts PR review, bug triage, documentation and migrations under anchors on one URL, so none can rank for searches like "AI code migration agent" or "automated bug triage".

    High impactMedium effort First result in 1 month Member of Growth Staff

    First action

    Draft the code migration page from one real Capy run.

    How to measure it

    Organic signups

    Now
    –
    Target
    +30%
    By
    90 days

    Steps

    1. 1

      List use cases

      Six pages from the nav items.

      Member of Growth Staff 1 day

    2. 2

      Pick example runs

      One real run per use case.

      Member of Technical Staff 1 week

    3. 3

      Write pages

      Example, time, cost, recipe.

      Member of Growth Staff 2 weeks

    4. 4

      Refresh lists

      Best-of articles updated with dates.

      Member of Growth Staff 1 week

    5. 5

      New comparisons

      Review-focused rivals.

      Member of Growth Staff 2 weeks

    6. 6

      Internal links

      From the homepage and nav.

      Member of Technical Staff 1 day

    7. 7

      Measure

      Signups by landing page.

      Member of Growth Staff monthly

    First result
    1 month
    Cost
    Staff time only
    Owner
    Member of Growth Staff

    Smallest test

    Publish the migration page and track its search impressions for a month.

    Why it works

    Split each use case into its own page with a real example run, the time and cost it took, and the automation recipe. Update the 2026 lists with what has changed since July, and add comparisons Capy does not have yet for review-focused rivals. Publish two pieces a week from the growth role, using Capy agents for the first draft and engineers for the facts.

    Search and AI assistants answer "which coding agent for X" from pages that show the work; stale lists and anchor sections lose those queries.

    Risks and how to handle them

    • Comparison claims go stale

      Date every comparison and recheck quarterly.

    Evidence

    • Your websiteSitemap (77 URLs on capy.ai, checked October 23, 2026): 24 articles, the newest dated July 9, 2026, almost all comparisons and lists (devin-vs-capy, codex-cloud-vs-capy, cursor-alternative, coderabbit-vs-capy, best-ai-coding-agents-2026 and others); 9 blog posts, the newest October 8; one customer story; one use-cases page whose sections (PR review, bug triage, documentation, migrations) are anchors rather than pages; no events, showcase or templates pages.Source
    • HiringMember of Growth Staff (San Francisco, posted October 6, 2026; $200K-$250K plus 0.5%-1% equity): "Growth at Capy is about getting developers to find Capy, try it, and bring their team." The role builds "Outreach and lifecycle automation, enrichment and qualification pipelines, landing pages, onboarding flows, demos", talks to customers "a lot", makes "Articles, videos, and demos about Capy ... launch videos, side-by-side comparisons, and benchmarks included", and runs "Hackathons, meetups, dinners, and workshops, often with other startups and developer communities".Source
    • Your websiteHomepage demo of an automated run shows a per-task cost ("Find critical bugs $0.49"), and the use-cases page promises "25+ agents per repo, coordinated across branches", "Answers review feedback without pings" and "Rebases, CI failures, and merge conflicts handled".Source
    • Competitor researchRivals Capy itself compares against in its articles: Devin, Codex cloud, GitHub Copilot coding agent, Cursor, Conductor, Amp, Aider, Roo Code and CodeRabbit for review. Reported funding for Capy in public databases is small and inconsistent; no priced round has been announced.Source
  • 06

    Capy has no seat fees and the growth post says the job is getting developers to "bring their team". Slack is where teammates see Capy work, yet a teammate who sees @Capy finish a fix in a thread has no one-click way in.

    High impactLow effort First result in 2 weeks Member of Growth Staff

    First action

    Count how many Slack channels with Capy have non-members in them.

    How to measure it

    Members per org

    Now
    –
    Target
    +30%
    By
    90 days

    Steps

    1. 1

      Measure

      Non-members in Capy-enabled channels.

      Member of Growth Staff 1 day

    2. 2

      Design the link

      Quiet, one per thread.

      Member of Technical Staff 2 days

    3. 3

      Deep link

      Join lands in the same thread.

      Member of Technical Staff 3 days

    4. 4

      Admin setting

      Let admins turn it off.

      Member of Technical Staff 1 day

    5. 5

      Launch

      All workspaces.

      Member of Growth Staff 1 day

    6. 6

      Track

      Joins and teammates' first threads.

      Member of Growth Staff weekly

    First result
    2 weeks
    Cost
    Engineering time
    Owner
    Member of Growth Staff

    Smallest test

    Test the link in five friendly workspaces.

    Why it works

    When Capy replies in a Slack thread, add a quiet "Join this workspace on Capy" link for people in the channel who are not members, landing straight in the thread. Show the inviting user how many teammates joined and what they shipped.

    Seeing a teammate's PR land is the strongest possible demo, and with no seat price there is no reason to hold teammates back.

    Risks and how to handle them

    • Feels spammy

      One link per thread and an admin off switch.

    Evidence

    • HiringThe same post: "Own the funnel. How developers find Capy, what they see first, how fast they reach a merged PR, and why they bring their team. You instrument every step and improve it." and "Run experiments weekly. Ship, measure, keep what works, and kill what doesn't." The mandate is "revenue leverage".Source
    • Your websitePricing: Lite $20/month ($16 annual) with "$20 credits", Pro $100/month with $105 credits, Max $200 to $1,000 in $100 steps with 10% bonus credits, and Enterprise custom; "unlimited members on every one", "no per-seat price"; "Try Capy Lite for $1 for 7 days" with a verified card, then $20/month. Credits "cover AI model costs and VM compute time" and "don't roll over".Source
    • Your websiteIntegrations: GitHub, Slack ("Tag @Capy in any channel to turn a thread into a running task"), Linear, Vercel, Tailscale and MCP. Enterprise offers up to 64 Large cloud machines, SSO, custom environments, VM snapshots, audit logging, SOC 2 Type II and "a dedicated onboarding manager"; "Most teams are fully onboarded within a week."Source

This quarter 9

The core plays for the next three months.

  • 07

    The growth role runs hackathons and meetups "often with other startups and developer communities", and Capy already has startup customers that sell to developers (Firecrawl, Exa, Supermemory, RunPod). Co-hosting shares the audience and the cost.

    High impactMedium effort First result in 1 quarter Member of Growth Staff

    First action

    Ask two logo customers to co-host the first event.

    How to measure it

    Event-to-active rate

    Now
    –
    Target
    40%
    By
    2 quarters

    Steps

    1. 1

      Pick co-hosts

      Two customers with developer audiences.

      Member of Growth Staff 1 week

    2. 2

      Book venue

      Co-host office or partner space.

      Member of Growth Staff 1 week

    3. 3

      Set the format

      Ship a merged PR; credits as prizes.

      Member of Growth Staff 2 days

    4. 4

      Promote

      Co-host lists, X, Discord, local groups.

      Member of Growth Staff 2 weeks

    5. 5

      Run

      Evening event with demos.

      Member of Growth Staff monthly

    6. 6

      Follow up

      Credits and starter task within 24 hours.

      Member of Growth Staff 1 day

    7. 7

      Measure

      Signups, merges, paid conversion.

      Member of Growth Staff monthly

    First result
    1 quarter
    Cost
    $2,000-$5,000 per event
    Owner
    Member of Growth Staff

    Smallest test

    Run one small event with one co-host.

    Why it works

    Run one evening build event a month in San Francisco with two logo customers as co-hosts: their APIs plus Capy, $1,000-credit prizes, and a rule that entries ship as merged PRs. Follow up every attendee with credits and a starter task within 24 hours, and record demos for content.

    Developers who ship something real with Capy in one evening become users, and co-hosts bring their own lists.

    Risks and how to handle them

    • Attendees come for food, not product

      Require a working repo connection to enter.

    Evidence

    • HiringMember of Growth Staff (San Francisco, posted October 6, 2026; $200K-$250K plus 0.5%-1% equity): "Growth at Capy is about getting developers to find Capy, try it, and bring their team." The role builds "Outreach and lifecycle automation, enrichment and qualification pipelines, landing pages, onboarding flows, demos", talks to customers "a lot", makes "Articles, videos, and demos about Capy ... launch videos, side-by-side comparisons, and benchmarks included", and runs "Hackathons, meetups, dinners, and workshops, often with other startups and developer communities".Source
    • Your websiteHomepage: H1 "The best AI software engineer"; "Orchestrate a fleet of cloud coding agents in all of the apps your team already lives in."; "Loved by 70,000+ engineers at" followed by 12 logos (Y Combinator, Supermemory, RunPod, Revyl, Million, Loop, Goblins, Firecrawl, Finta, Exa, Autumn, Aside AI). It shows Terminal-Bench results, bring-your-own-subscription, automations, a review agent and a Slack integration.Source
    • Your websitePartner deals: YC companies get "$1,000 in credits" redeemed on Bookface, and Solo Founders members get "$1,000 in Capy credits" through its perks library; both apply to Capy Pro. The YC page says "Trusted by your favorite YC companies".Source
  • 08

    The growth post asks for outreach, enrichment and qualification pipelines, and says much of it "runs on Capy agents". Capy's pitch is "Ship more with the team you already have", so the best moment to reach a startup is when it posts engineering jobs.

    High impactMedium effort First result in 1 month Member of Growth Staff

    First action

    Pull last week's seed-stage engineering job posts and hand-qualify 50.

    How to measure it

    Booked calls

    Now
    –
    Target
    20 a month
    By
    1 quarter

    Steps

    1. 1

      Define the signal

      New engineering posts at 5-50 person teams.

      Member of Growth Staff 2 days

    2. 2

      Build the pipeline

      Capy automation on a daily schedule.

      Member of Growth Staff 1 week

    3. 3

      Qualify

      Stack and size checks.

      Member of Growth Staff 3 days

    4. 4

      Draft notes

      Proof plus a relevant starter task.

      Member of Growth Staff 3 days

    5. 5

      Review and send

      A person approves each note.

      Member of Growth Staff daily

    6. 6

      Measure

      Replies, calls, trials, paid.

      Member of Growth Staff weekly

    7. 7

      Write it up

      How the pipeline runs on Capy.

      Member of Growth Staff 1 week

    First result
    1 month
    Cost
    Data and email tooling
    Owner
    Member of Growth Staff

    Smallest test

    Send 50 hand-reviewed notes and count replies.

    Why it works

    Have a Capy automation pull new engineering job posts from public job boards each day, qualify companies by team size and stack (GitHub plus Linear or Slack), and draft a short note to the CTO with the Goblins numbers and a starter task for their kind of codebase. A person reviews and sends.

    The trigger matches the pain, the pipeline itself is a Capy demo, and the growth role can write it up as content.

    Risks and how to handle them

    • Cold outreach annoys developers

      Short, specific, one follow-up only.

    Evidence

    • HiringMember of Growth Staff (San Francisco, posted October 6, 2026; $200K-$250K plus 0.5%-1% equity): "Growth at Capy is about getting developers to find Capy, try it, and bring their team." The role builds "Outreach and lifecycle automation, enrichment and qualification pipelines, landing pages, onboarding flows, demos", talks to customers "a lot", makes "Articles, videos, and demos about Capy ... launch videos, side-by-side comparisons, and benchmarks included", and runs "Hackathons, meetups, dinners, and workshops, often with other startups and developer communities".Source
    • Your websitePartner deals: YC companies get "$1,000 in credits" redeemed on Bookface, and Solo Founders members get "$1,000 in Capy credits" through its perks library; both apply to Capy Pro. The YC page says "Trusted by your favorite YC companies".Source
    • Your websiteCustomers page (checked October 23, 2026): the filter reads "All [1]"; the only story is Goblins. The Goblins story reports merged pull requests going from "~60" to "300+ a week" with the same small team, "1,363 (98%)" of 1,392 merged PRs built or reviewed by Capy between September 10 and October 10, 2026, and "more than 14,000 reviews", with a 2:42 video.Source
  • 09

    Capy's $1,000 credit offer already runs for YC on Bookface and for Solo Founders members. Most accelerators and startup programs run perk libraries that founders browse when they set up their stack.

    Medium impactLow effort First result in 2 months Member of Growth Staff

    First action

    Shortlist ten programs with perk libraries.

    How to measure it

    Partner redemptions

    Now
    –
    Target
    200 a month
    By
    2 quarters

    Steps

    1. 1

      List programs

      Ten perk libraries.

      Member of Growth Staff 1 week

    2. 2

      Contact

      Program managers.

      Member of Growth Staff 2 weeks

    3. 3

      Clone the page

      One landing page per program.

      Member of Growth Staff 1 day each

    4. 4

      Track

      Redemption codes per program.

      Member of Technical Staff 2 days

    5. 5

      Onboard

      Starter tasks for redeemers.

      Member of Growth Staff ongoing

    6. 6

      Review

      Drop programs that do not convert.

      Member of Growth Staff quarterly

    First result
    2 months
    Cost
    Credits only
    Owner
    Member of Growth Staff

    Smallest test

    Sign two programs this quarter.

    Why it works

    List the ten largest accelerator and startup-program perk libraries, offer the same $1,000 Pro credit deal with a dedicated landing page each, and track redemptions and paid conversion per program.

    Perks reach founders at the moment they choose tools, and the offer and pages already exist as a template.

    Risks and how to handle them

    • Credit burners never pay

      Watch paid conversion by program and adjust.

    Evidence

    • Your websitePartner deals: YC companies get "$1,000 in credits" redeemed on Bookface, and Solo Founders members get "$1,000 in Capy credits" through its perks library; both apply to Capy Pro. The YC page says "Trusted by your favorite YC companies".Source
    • Your websitePricing: Lite $20/month ($16 annual) with "$20 credits", Pro $100/month with $105 credits, Max $200 to $1,000 in $100 steps with 10% bonus credits, and Enterprise custom; "unlimited members on every one", "no per-seat price"; "Try Capy Lite for $1 for 7 days" with a verified card, then $20/month. Credits "cover AI model costs and VM compute time" and "don't roll over".Source
  • 10

    Capy shows Terminal-Bench results on its homepage and has published model evaluations as captain. The growth role is asked for "side-by-side comparisons, and benchmarks". Each new model release sends developers searching for how it performs in real agent work.

    High impactMedium effort First result in 2 months Member of Growth Staff

    First action

    Pick 30 representative tasks and publish the method.

    How to measure it

    Monthly leaderboard visitors

    Now
    –
    Target
    20,000
    By
    2 quarters

    Steps

    1. 1

      Pick tasks

      30 tasks across use cases.

      Member of Technical Staff 1 week

    2. 2

      Automate runs

      One automation per model release.

      Member of Technical Staff 2 weeks

    3. 3

      Build the page

      Table, cost and method.

      Member of Growth Staff 1 week

    4. 4

      Publish

      With a short video.

      Member of Growth Staff 1 day

    5. 5

      Update

      Within days of each release.

      Member of Growth Staff per release

    6. 6

      Measure

      Visits, links, signups.

      Member of Growth Staff monthly

    First result
    2 months
    Cost
    Compute credits
    Owner
    Member of Growth Staff

    Smallest test

    Publish the first table with five models.

    Why it works

    Publish a page that ranks every model Capy supports on the same set of real tasks in the Capy harness (success rate, time, cost per task), updated within days of each release, with the method and tasks shown. Post the update to X and Discord each time.

    A neutral, refreshed benchmark becomes a page developers bookmark, and it shows Capy's model-agnostic pitch in numbers.

    Risks and how to handle them

    • Accusations of bias

      Publish tasks and raw results.

    Evidence

    • Your websiteHomepage: H1 "The best AI software engineer"; "Orchestrate a fleet of cloud coding agents in all of the apps your team already lives in."; "Loved by 70,000+ engineers at" followed by 12 logos (Y Combinator, Supermemory, RunPod, Revyl, Million, Loop, Goblins, Firecrawl, Finta, Exa, Autumn, Aside AI). It shows Terminal-Bench results, bring-your-own-subscription, automations, a review agent and a Slack integration.Source
    • HiringMember of Growth Staff (San Francisco, posted October 6, 2026; $200K-$250K plus 0.5%-1% equity): "Growth at Capy is about getting developers to find Capy, try it, and bring their team." The role builds "Outreach and lifecycle automation, enrichment and qualification pipelines, landing pages, onboarding flows, demos", talks to customers "a lot", makes "Articles, videos, and demos about Capy ... launch videos, side-by-side comparisons, and benchmarks included", and runs "Hackathons, meetups, dinners, and workshops, often with other startups and developer communities".Source
    • Your websiteChangelog: Capy 0.5.0 on October 22, 2026 added running Claude models on a user's own Claude Pro or Max subscription, OIDC cloud credentials for Google Cloud, AWS, Azure and Vault, automations in private threads and subagents on paired devices; 15 releases are listed and 12 site updates in the last 90 days.Source
    • Your websiteHomepage demo of an automated run shows a per-task cost ("Find critical bugs $0.49"), and the use-cases page promises "25+ agents per repo, coordinated across branches", "Answers review feedback without pings" and "Rebases, CI failures, and merge conflicts handled".Source
  • 11

    Capy ships often: 15 releases, 12 site updates in the last 90 days, and 0.5.0 just added cloud credentials, private-thread automations and subagents on paired devices. Changelog text rarely travels; short videos of a feature working do.

    Medium impactLow effort First result in 1 month Member of Growth Staff

    First action

    Record the 0.5.0 clip this week.

    How to measure it

    Signups from release clips

    Now
    –
    Target
    100 a month
    By
    1 quarter

    Steps

    1. 1

      Pick the feature

      One headline per release.

      Member of Growth Staff per release

    2. 2

      Record

      60 seconds on a real repo.

      Member of Growth Staff 2 hours

    3. 3

      Post

      X and Discord.

      Member of Growth Staff 1 hour

    4. 4

      Embed

      At the top of the changelog entry.

      Member of Technical Staff 30 minutes

    5. 5

      Track

      Views and signups.

      Member of Growth Staff weekly

    6. 6

      Collect

      A best-of reel each quarter.

      Member of Growth Staff quarterly

    First result
    1 month
    Cost
    Staff time only
    Owner
    Member of Growth Staff

    Smallest test

    Post the 0.5.0 clip.

    Why it works

    For each release, record one 60-second clip of the headline feature on a real repo, post it to X with the changelog link, pin it in Discord, and embed it at the top of the changelog entry.

    Developers share things they can see working, and a steady release cadence is proof of momentum.

    Risks and how to handle them

    • Too many clips

      One per release only.

    Evidence

    • Your websiteChangelog: Capy 0.5.0 on October 22, 2026 added running Claude models on a user's own Claude Pro or Max subscription, OIDC cloud credentials for Google Cloud, AWS, Azure and Vault, automations in private threads and subagents on paired devices; 15 releases are listed and 12 site updates in the last 90 days.Source
    • HiringMember of Growth Staff (San Francisco, posted October 6, 2026; $200K-$250K plus 0.5%-1% equity): "Growth at Capy is about getting developers to find Capy, try it, and bring their team." The role builds "Outreach and lifecycle automation, enrichment and qualification pipelines, landing pages, onboarding flows, demos", talks to customers "a lot", makes "Articles, videos, and demos about Capy ... launch videos, side-by-side comparisons, and benchmarks included", and runs "Hackathons, meetups, dinners, and workshops, often with other startups and developer communities".Source
    • Your websiteAbout page: "Capy is built by Scrapybara, Inc." in San Francisco, "Backed by Y Combinator and tier-1 investors"; "One agent is a copilot. Twenty-five agents are a workforce." Community links go to X (@capydotai) and a Discord server.Source
  • 12

    The enterprise page lists SSO, audit logging, 64 machines and onboarding "within a week", but no enterprise proof: the only story is a small team. Larger buyers want a measured trial before a contract.

    High impactMedium effort First result in 1 quarter Co founder

    First action

    Write the sample pilot report from Goblins' numbers.

    How to measure it

    Pilot to contract rate

    Now
    –
    Target
    50%
    By
    2 quarters

    Steps

    1. 1

      Define the pilot

      Scope, length, success measures.

      Co-founder 1 week

    2. 2

      Sample report

      Goblins format.

      Member of Growth Staff 3 days

    3. 3

      Security pack

      SOC 2 report and trust center links.

      Co-founder 2 days

    4. 4

      Page update

      Pilot offer on /enterprise.

      Member of Growth Staff 1 day

    5. 5

      Run pilots

      Onboarding manager per pilot.

      Co-founder ongoing

    6. 6

      Readout

      Report and proposal.

      Co-founder per pilot

    First result
    1 quarter
    Cost
    Staff time and credits
    Owner
    Co founder

    Smallest test

    Offer the pilot to the next three enterprise inquiries.

    Why it works

    Offer a fixed two-week pilot: one team, one repo, an onboarding manager, and a report at the end in the Goblins format (merged PRs, share by Capy, review count, hours saved). Put the pilot on the enterprise page with a sample report.

    A defined pilot with a clear readout shortens security and budget discussions, and each one can become a story.

    Risks and how to handle them

    • Pilots without a buyer

      Require a named budget owner to start.

    Evidence

    • Your websiteIntegrations: GitHub, Slack ("Tag @Capy in any channel to turn a thread into a running task"), Linear, Vercel, Tailscale and MCP. Enterprise offers up to 64 Large cloud machines, SSO, custom environments, VM snapshots, audit logging, SOC 2 Type II and "a dedicated onboarding manager"; "Most teams are fully onboarded within a week."Source
    • Your websiteCustomers page (checked October 23, 2026): the filter reads "All [1]"; the only story is Goblins. The Goblins story reports merged pull requests going from "~60" to "300+ a week" with the same small team, "1,363 (98%)" of 1,392 merged PRs built or reviewed by Capy between September 10 and October 10, 2026, and "more than 14,000 reviews", with a 2:42 video.Source
    • HiringMember of Technical Staff (posted October 6, 2026): "AI coding is a crowded market now, and developers expect first-minute magic and a quality bar that holds across every surface." The role spans the React web app, Electron desktop app, agent loop, backend and Kubernetes infrastructure, and "Talk to users". Capy says "We build Capy with Capy: our own PRs, reviews, and automations run through it."Source
  • 13

    Capy can count what it did for Goblins down to the PR. Every paying team could get the same numbers each month, and credits that don't roll over make value visible use important.

    Medium impactMedium effort First result in 1 quarter Member of Growth Staff

    First action

    Generate the report by hand for ten customers.

    How to measure it

    Net revenue retention

    Now
    –
    Target
    +5 points
    By
    2 quarters

    Steps

    1. 1

      Define the numbers

      Six metrics per org.

      Member of Growth Staff 1 week

    2. 2

      Build the query

      Monthly job.

      Member of Technical Staff 1 week

    3. 3

      Design the email

      One page, one suggestion.

      Member of Growth Staff 3 days

    4. 4

      Send

      All paying orgs.

      Member of Growth Staff monthly

    5. 5

      Follow up

      Calls with low-usage orgs.

      Member of Growth Staff monthly

    6. 6

      Measure

      Retention by report open.

      Member of Growth Staff quarterly

    First result
    1 quarter
    Cost
    Engineering time
    Owner
    Member of Growth Staff

    Smallest test

    Send ten hand-made reports.

    Why it works

    Email each org admin a monthly report: merged PRs, reviews, bugs fixed, automations run, credits used and the cost per merged PR, with one suggestion for an automation they are not using yet.

    Teams that see value renew and expand; the report also gives champions something to forward to their manager.

    Risks and how to handle them

    • Low numbers discourage

      Pair them with a concrete next step.

    Evidence

    • Your websiteCustomers page (checked October 23, 2026): the filter reads "All [1]"; the only story is Goblins. The Goblins story reports merged pull requests going from "~60" to "300+ a week" with the same small team, "1,363 (98%)" of 1,392 merged PRs built or reviewed by Capy between September 10 and October 10, 2026, and "more than 14,000 reviews", with a 2:42 video.Source
    • Your websitePricing: Lite $20/month ($16 annual) with "$20 credits", Pro $100/month with $105 credits, Max $200 to $1,000 in $100 steps with 10% bonus credits, and Enterprise custom; "unlimited members on every one", "no per-seat price"; "Try Capy Lite for $1 for 7 days" with a verified card, then $20/month. Credits "cover AI model costs and VM compute time" and "don't roll over".Source
    • Your websiteHomepage demo of an automated run shows a per-task cost ("Find critical bugs $0.49"), and the use-cases page promises "25+ agents per repo, coordinated across branches", "Answers review feedback without pings" and "Rebases, CI failures, and merge conflicts handled".Source
  • 14

    Pricing is in credits that cover model and machine time, while the homepage shows a single run costing $0.49. Developers comparing plans cannot tell how many PRs $20 or $100 buys.

    Medium impactLow effort First result in 1 month Member of Growth Staff

    First action

    Pull median cost for five task types.

    How to measure it

    Pricing-to-paid conversion

    Now
    –
    Target
    +15%
    By
    1 quarter

    Steps

    1. 1

      Pull costs

      Five task types.

      Member of Technical Staff 2 days

    2. 2

      Write the table

      Medians and ranges.

      Member of Growth Staff 1 day

    3. 3

      Build calculator

      Tasks in, plan out.

      Member of Technical Staff 3 days

    4. 4

      Publish

      On /pricing.

      Member of Growth Staff 1 day

    5. 5

      Test

      Against the current page.

      Member of Growth Staff 3 weeks

    6. 6

      Refresh

      Quarterly as models change.

      Member of Growth Staff quarterly

    First result
    1 month
    Cost
    Staff time only
    Owner
    Member of Growth Staff

    Smallest test

    Add the table without the calculator first.

    Why it works

    Publish median cost per task type (bug fix, review, migration file, docs update) from real runs, and add a small calculator to the pricing page: tasks per week in, recommended plan out.

    Usage pricing converts better when buyers can picture the bill, and it supports upgrades to Pro and Max.

    Risks and how to handle them

    • Costs change with models

      Show ranges and update dates.

    Evidence

    • Your websitePricing: Lite $20/month ($16 annual) with "$20 credits", Pro $100/month with $105 credits, Max $200 to $1,000 in $100 steps with 10% bonus credits, and Enterprise custom; "unlimited members on every one", "no per-seat price"; "Try Capy Lite for $1 for 7 days" with a verified card, then $20/month. Credits "cover AI model costs and VM compute time" and "don't roll over".Source
    • Your websiteHomepage demo of an automated run shows a per-task cost ("Find critical bugs $0.49"), and the use-cases page promises "25+ agents per repo, coordinated across branches", "Answers review feedback without pings" and "Rebases, CI failures, and merge conflicts handled".Source
  • 15

    Capy lets teams bring the coding subscriptions they already pay for, and 0.5.0 extended this to personal Claude plans. Many developers think of agent tools as "another subscription"; Capy's answer is that it can run the plans they have, in parallel.

    Medium impactLow effort First result in 1 month Member of Growth Staff

    First action

    Write the page from the 0.5.0 notes and the pricing FAQ.

    How to measure it

    Subscription-connected signups

    Now
    –
    Target
    +25%
    By
    1 quarter

    Steps

    1. 1

      Write the page

      How it works and costs.

      Member of Growth Staff 2 days

    2. 2

      Add a setup video

      Two minutes.

      Member of Growth Staff 1 day

    3. 3

      Link from pricing

      And the homepage section.

      Member of Technical Staff 1 hour

    4. 4

      Promote

      X and Discord.

      Member of Growth Staff 1 day

    5. 5

      Track

      Connected subscriptions at signup.

      Member of Growth Staff weekly

    6. 6

      Update

      As providers change terms.

      Member of Growth Staff monthly

    First result
    1 month
    Cost
    $0
    Owner
    Member of Growth Staff

    Smallest test

    Publish the page and run one X post.

    Why it works

    Build one page, "Run the plan you already pay for, in parallel", explaining how bring-your-own-subscription works, what Capy adds (cloud machines, review, automations, Slack), and what it costs on top. Target searches about running coding plans in parallel and in the cloud.

    Removing the "second bill" objection lowers the bar to try, and it is a message few rivals can match.

    Risks and how to handle them

    • Provider terms change

      Check each provider's current terms before publishing.

    Evidence

    • Your websiteChangelog: Capy 0.5.0 on October 22, 2026 added running Claude models on a user's own Claude Pro or Max subscription, OIDC cloud credentials for Google Cloud, AWS, Azure and Vault, automations in private threads and subagents on paired devices; 15 releases are listed and 12 site updates in the last 90 days.Source
    • Your websitePricing: Lite $20/month ($16 annual) with "$20 credits", Pro $100/month with $105 credits, Max $200 to $1,000 in $100 steps with 10% bonus credits, and Enterprise custom; "unlimited members on every one", "no per-seat price"; "Try Capy Lite for $1 for 7 days" with a verified card, then $20/month. Credits "cover AI model costs and VM compute time" and "don't roll over".Source
    • Your websiteHomepage: H1 "The best AI software engineer"; "Orchestrate a fleet of cloud coding agents in all of the apps your team already lives in."; "Loved by 70,000+ engineers at" followed by 12 logos (Y Combinator, Supermemory, RunPod, Revyl, Million, Loop, Goblins, Firecrawl, Finta, Exa, Autumn, Aside AI). It shows Terminal-Bench results, bring-your-own-subscription, automations, a review agent and a Slack integration.Source

Big bets 3

Larger plays that can change the trajectory.

  • 16

    Capy says "We build Capy with Capy: our own PRs, reviews, and automations run through it." That is the most credible case study it has, and it is not on the site.

    High impactMedium effort First result in 1 quarter Co founder

    First action

    Pull last month's internal PR numbers.

    How to measure it

    Monthly build-log readers

    Now
    –
    Target
    15,000
    By
    3 quarters

    Steps

    1. 1

      Pull numbers

      Internal PRs and reviews.

      Member of Technical Staff 2 days

    2. 2

      Write the first post

      Numbers, automations, one miss.

      Co-founder 1 week

    3. 3

      Stats page

      Updated monthly.

      Member of Growth Staff 1 week

    4. 4

      Release templates

      One internal automation a month.

      Member of Technical Staff monthly

    5. 5

      Promote

      X, Discord, newsletter.

      Member of Growth Staff monthly

    6. 6

      Measure

      Readers and signups.

      Member of Growth Staff monthly

    First result
    1 quarter
    Cost
    Staff time only
    Owner
    Co founder

    Smallest test

    Publish the first month's numbers.

    Why it works

    Publish a monthly post and a public stats page: PRs merged by the team, the share built or reviewed by Capy, the automations that run the company, and one thing that went wrong and how it was fixed. Release one internal automation as a template each month.

    Developers trust a team that uses its own product in public, and the honest misses make it believable.

    Risks and how to handle them

    • Shares internal details

      Review each post for anything private before it goes out.

    Evidence

    • HiringMember of Technical Staff (posted October 6, 2026): "AI coding is a crowded market now, and developers expect first-minute magic and a quality bar that holds across every surface." The role spans the React web app, Electron desktop app, agent loop, backend and Kubernetes infrastructure, and "Talk to users". Capy says "We build Capy with Capy: our own PRs, reviews, and automations run through it."Source
    • Your websiteCustomers page (checked October 23, 2026): the filter reads "All [1]"; the only story is Goblins. The Goblins story reports merged pull requests going from "~60" to "300+ a week" with the same small team, "1,363 (98%)" of 1,392 merged PRs built or reviewed by Capy between September 10 and October 10, 2026, and "more than 14,000 reviews", with a 2:42 video.Source
    • HiringMember of Growth Staff (San Francisco, posted October 6, 2026; $200K-$250K plus 0.5%-1% equity): "Growth at Capy is about getting developers to find Capy, try it, and bring their team." The role builds "Outreach and lifecycle automation, enrichment and qualification pipelines, landing pages, onboarding flows, demos", talks to customers "a lot", makes "Articles, videos, and demos about Capy ... launch videos, side-by-side comparisons, and benchmarks included", and runs "Hackathons, meetups, dinners, and workshops, often with other startups and developer communities".Source
  • 17

    Automations are a core feature (Sentry alert to PR, Linear issue to work, daily security checks), and Goblins calls its daily merged-code check its most valuable automation. There is no public library of recipes; the templates path is not public.

    High impactHigh effort First result in 2 quarters Member of Technical Staff

    First action

    Write ten recipes from the most common automations.

    How to measure it

    Library installs

    Now
    –
    Target
    1,000 a month
    By
    3 quarters

    Steps

    1. 1

      Pick recipes

      Ten most common.

      Member of Growth Staff 1 week

    2. 2

      Build pages

      Recipe template with install.

      Member of Technical Staff 1 month

    3. 3

      Install flow

      One click into a project.

      Member of Technical Staff 1 month

    4. 4

      Seed 30

      Team and customers.

      Member of Growth Staff 1 month

    5. 5

      Submissions

      User recipes with credit.

      Member of Technical Staff 1 quarter

    6. 6

      Measure

      Installs and retention.

      Member of Growth Staff monthly

    First result
    2 quarters
    Cost
    Engineering time
    Owner
    Member of Technical Staff

    Smallest test

    Publish ten recipe pages without one-click install.

    Why it works

    Build an indexable library: each recipe has a page with what it does, a sample run, the cost, and an "Install in Capy" button. Seed it with 30 recipes from customers and the team, and let users submit their own with credit.

    Recipes rank in search, show the product working, and turn a blank automations screen into a starting point.

    Risks and how to handle them

    • Low-quality submissions

      Review before listing.

    Evidence

    • Your websiteIntegrations: GitHub, Slack ("Tag @Capy in any channel to turn a thread into a running task"), Linear, Vercel, Tailscale and MCP. Enterprise offers up to 64 Large cloud machines, SSO, custom environments, VM snapshots, audit logging, SOC 2 Type II and "a dedicated onboarding manager"; "Most teams are fully onboarded within a week."Source
    • Your websiteCustomers page (checked October 23, 2026): the filter reads "All [1]"; the only story is Goblins. The Goblins story reports merged pull requests going from "~60" to "300+ a week" with the same small team, "1,363 (98%)" of 1,392 merged PRs built or reviewed by Capy between September 10 and October 10, 2026, and "more than 14,000 reviews", with a 2:42 video.Source
    • Your websiteSitemap (77 URLs on capy.ai, checked October 23, 2026): 24 articles, the newest dated July 9, 2026, almost all comparisons and lists (devin-vs-capy, codex-cloud-vs-capy, cursor-alternative, coderabbit-vs-capy, best-ai-coding-agents-2026 and others); 9 blog posts, the newest October 8; one customer story; one use-cases page whose sections (PR review, bug triage, documentation, migrations) are anchors rather than pages; no events, showcase or templates pages.Source
    • Your websiteHomepage demo of an automated run shows a per-task cost ("Find critical bugs $0.49"), and the use-cases page promises "25+ agents per repo, coordinated across branches", "Answers review feedback without pings" and "Rebases, CI failures, and merge conflicts handled".Source
  • 18

    Capy's promise is "Twenty-five agents are a workforce". Development agencies sell engineering capacity and have many client repos; unlimited members and Max plans suit them.

    Medium impactHigh effort First result in 3 quarters Co founder

    First action

    Interview five agencies about how they would use Capy.

    How to measure it

    Partner-sourced orgs

    Now
    –
    Target
    50
    By
    4 quarters

    Steps

    1. 1

      Interview

      Five agencies.

      Co-founder 1 month

    2. 2

      Design the tier

      Billing, separation, margin.

      Co-founder 1 month

    3. 3

      Build separation

      Client projects.

      Member of Technical Staff 1 quarter

    4. 4

      Sign five

      Pilot agencies.

      Co-founder 1 quarter

    5. 5

      Directory

      Certified partners page.

      Member of Growth Staff 1 month

    6. 6

      Review

      Revenue and retention.

      Co-founder quarterly

    First result
    3 quarters
    Cost
    Product and partner time
    Owner
    Co founder

    Smallest test

    Sign one agency on a custom deal.

    Why it works

    Offer agencies a partner tier: client-project separation, consolidated billing, a margin on referred client accounts, and a directory listing of certified partners. Start with five agencies and publish their delivery numbers.

    Agencies bring many codebases at once, and their clients often keep using the tool after the project.

    Risks and how to handle them

    • Agencies churn between projects

      Bill per client org so accounts survive projects.

    Evidence

    • Your websiteAbout page: "Capy is built by Scrapybara, Inc." in San Francisco, "Backed by Y Combinator and tier-1 investors"; "One agent is a copilot. Twenty-five agents are a workforce." Community links go to X (@capydotai) and a Discord server.Source
    • Your websitePricing: Lite $20/month ($16 annual) with "$20 credits", Pro $100/month with $105 credits, Max $200 to $1,000 in $100 steps with 10% bonus credits, and Enterprise custom; "unlimited members on every one", "no per-seat price"; "Try Capy Lite for $1 for 7 days" with a verified card, then $20/month. Credits "cover AI model costs and VM compute time" and "don't roll over".Source
    • Your websiteIntegrations: GitHub, Slack ("Tag @Capy in any channel to turn a thread into a running task"), Linear, Vercel, Tailscale and MCP. Enterprise offers up to 64 Large cloud machines, SSO, custom environments, VM snapshots, audit logging, SOC 2 Type II and "a dedicated onboarding manager"; "Most teams are fully onboarded within a week."Source

Wildcards 3

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

  • 19

    Capy opens and drives PRs to merge, including for open source projects in its program. Every merged PR in a public repo is seen by maintainers and contributors who might want the same help.

    Medium impactLow effort First result in 1 month Member of Technical Staff

    First action

    Count Capy PRs merged in public repos last month.

    How to measure it

    Footer clicks

    Now
    –
    Target
    2,000 a month
    By
    2 quarters

    Steps

    1. 1

      Count

      Public repo PRs.

      Member of Technical Staff 1 day

    2. 2

      Design the line

      One short sentence.

      Member of Growth Staff 1 day

    3. 3

      Setting

      Per org, on by default for OSS.

      Member of Technical Staff 2 days

    4. 4

      Landing page

      Open source program.

      Member of Growth Staff 1 day

    5. 5

      Launch

      With a note to maintainers.

      Member of Growth Staff 1 day

    6. 6

      Measure

      Clicks and applications.

      Member of Growth Staff monthly

    First result
    1 month
    Cost
    $0
    Owner
    Member of Technical Staff

    Smallest test

    Turn it on for five consenting projects.

    Why it works

    Add an optional one-line footer to Capy PR descriptions in public repos ("Opened and tested by Capy") linking to a page about the open source program. Default it on for program members, with an easy off switch.

    It puts real, reviewed work in front of the exact audience, at zero cost per impression.

    Risks and how to handle them

    • Maintainers object

      Off switch and no footer outside public repos.

    Evidence

    • Your websiteOpen Source Program page: free Capy for open source projects with an OSI-approved license, "No call needed", "Unlimited Contributors", and "Community Recognition: Get featured in our open source showcase and community highlights." No showcase page is linked or listed in the sitemap.Source
    • Your websiteHomepage demo of an automated run shows a per-task cost ("Find critical bugs $0.49"), and the use-cases page promises "25+ agents per repo, coordinated across branches", "Answers review feedback without pings" and "Rebases, CI failures, and merge conflicts handled".Source
    • HiringMember of Technical Staff (posted October 6, 2026): "AI coding is a crowded market now, and developers expect first-minute magic and a quality bar that holds across every surface." The role spans the React web app, Electron desktop app, agent loop, backend and Kubernetes infrastructure, and "Talk to users". Capy says "We build Capy with Capy: our own PRs, reviews, and automations run through it."Source
  • 20

    Capy has a Discord and a team that spends "most of your day ... writing specs, steering agents, reviewing their PRs". Watching an expert direct 25 agents is something users want to learn.

    Medium impactLow effort First result in 1 month Member of Growth Staff

    First action

    Schedule the first session and ask Discord for a task.

    How to measure it

    Active members

    Now
    –
    Target
    +40%
    By
    1 quarter

    Steps

    1. 1

      Schedule

      Same time weekly.

      Member of Growth Staff 1 day

    2. 2

      Collect tasks

      From Discord users.

      Member of Growth Staff weekly

    3. 3

      Run

      One hour live.

      Member of Technical Staff weekly

    4. 4

      Record

      Post recording and notes.

      Member of Growth Staff weekly

    5. 5

      Clip

      Short videos for X.

      Member of Growth Staff weekly

    6. 6

      Measure

      Attendance and usage after.

      Member of Growth Staff monthly

    First result
    1 month
    Cost
    Staff time only
    Owner
    Member of Growth Staff

    Smallest test

    Run one session.

    Why it works

    Run a weekly one-hour live session: a team member takes a user's real backlog item and directs Capy through it on screen, answering questions. Post the recording and a written summary.

    Live problem-solving teaches the orchestration skill that drives usage, and recordings become content.

    Risks and how to handle them

    • Low attendance early

      Promote recordings, not just the live slot.

    Evidence

    • Your websiteAbout page: "Capy is built by Scrapybara, Inc." in San Francisco, "Backed by Y Combinator and tier-1 investors"; "One agent is a copilot. Twenty-five agents are a workforce." Community links go to X (@capydotai) and a Discord server.Source
    • HiringMember of Technical Staff (posted October 6, 2026): "AI coding is a crowded market now, and developers expect first-minute magic and a quality bar that holds across every surface." The role spans the React web app, Electron desktop app, agent loop, backend and Kubernetes infrastructure, and "Talk to users". Capy says "We build Capy with Capy: our own PRs, reviews, and automations run through it."Source
    • HiringMember of Growth Staff (San Francisco, posted October 6, 2026; $200K-$250K plus 0.5%-1% equity): "Growth at Capy is about getting developers to find Capy, try it, and bring their team." The role builds "Outreach and lifecycle automation, enrichment and qualification pipelines, landing pages, onboarding flows, demos", talks to customers "a lot", makes "Articles, videos, and demos about Capy ... launch videos, side-by-side comparisons, and benchmarks included", and runs "Hackathons, meetups, dinners, and workshops, often with other startups and developer communities".Source
  • 21

    Capy argues that "The age of the solo coder is over" and that developers must learn to orchestrate fleets. Few people know how to scope tasks for 25 agents; the company that teaches it owns the category language.

    Medium impactMedium effort First result in 2 months Member of Growth Staff

    First action

    Outline the five lessons.

    How to measure it

    Completions

    Now
    –
    Target
    1,000
    By
    2 quarters

    Steps

    1. 1

      Outline

      Five lessons.

      Member of Growth Staff 1 week

    2. 2

      Sample repo

      Exercises.

      Member of Technical Staff 2 weeks

    3. 3

      Record

      Short videos.

      Member of Growth Staff 1 month

    4. 4

      Certificate

      Shareable image.

      Member of Growth Staff 1 week

    5. 5

      Promote

      Bootcamps, universities, events.

      Member of Growth Staff ongoing

    6. 6

      Measure

      Completions and activation.

      Member of Growth Staff monthly

    First result
    2 months
    Cost
    Production time
    Owner
    Member of Growth Staff

    Smallest test

    Publish lesson one.

    Why it works

    Make a free five-lesson course (scoping tasks, defining done, review, automations, cost control) with exercises on a sample repo in Capy, and a shareable certificate. Promote it to bootcamps, universities and the developer communities at events.

    Education creates skilled users and a reason to share, and positions Capy as the reference for the new way of working.

    Risks and how to handle them

    • Course becomes outdated

      Update a lesson each quarter.

    Evidence

    • Your websiteAbout page: "Capy is built by Scrapybara, Inc." in San Francisco, "Backed by Y Combinator and tier-1 investors"; "One agent is a copilot. Twenty-five agents are a workforce." Community links go to X (@capydotai) and a Discord server.Source
    • HiringMember of Growth Staff (San Francisco, posted October 6, 2026; $200K-$250K plus 0.5%-1% equity): "Growth at Capy is about getting developers to find Capy, try it, and bring their team." The role builds "Outreach and lifecycle automation, enrichment and qualification pipelines, landing pages, onboarding flows, demos", talks to customers "a lot", makes "Articles, videos, and demos about Capy ... launch videos, side-by-side comparisons, and benchmarks included", and runs "Hackathons, meetups, dinners, and workshops, often with other startups and developer communities".Source
    • Your websiteHomepage demo of an automated run shows a per-task cost ("Find critical bugs $0.49"), and the use-cases page promises "25+ agents per repo, coordinated across branches", "Answers review feedback without pings" and "Rebases, CI failures, and merge conflicts handled".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 #8 (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)
ProgrammableTalk

Fill the events in this plan

These ideas are events. Talk gets registrants for them and tells you who's coming and what their company does before the day.

Plan an event (opens programmable.talk 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.

  • Avi Medical growth strategyRuns a network of tech-enabled GP practices in German cities, with its own app for booking, video consultations, messages, results and…21 growth ideas
  • Prokeep growth strategyAn AI-powered Order Engine for wholesale distributors: one inbox for text, email, web chat, fax and voice, Order Automation that turns…21 growth ideas
  • Semble growth strategyAI software for building-system integrators: takeoffs and device counts from PDF and DWG plan sets, device layout with the integrator's…22 growth ideas
  • Wilson Elser growth strategyA national defense litigation and insurance coverage law firm (AmLaw 100), handling liability, coverage, professional liability…21 growth ideas
  • Fashion Nova growth strategyA Los Angeles trend-to-market fashion brand selling affordable, size-inclusive clothing online for women, plus, men and kids, plus beauty…22 growth ideas
  • M-KOPA growth strategyA fintech that finances smartphones, electric motorbikes and other assets for Every Day Earners on daily, weekly or monthly payments, with…21 growth ideas
  • Dupuis Design growth strategyAn interior architecture and design studio founded by Peggy Dupuis, with a by-appointment art gallery and a (currently empty) online shop…20 growth ideas
  • Curio Digital growth strategyA B2B web agency that designs, builds and manages conversion-focused websites, mostly on Webflow, and sells them as one system (GrowthOS)…23 growth ideas

See every playbook and website review

Independent research. Programmable wrote this playbook from public sources, as they stood on October 26, 2026, and cites them throughout. Capy did not commission, review or endorse it, and Programmable is not affiliated with Capy. 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 Capy runs its business.

Something wrong, or do you represent Capy? 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

    Customer stories, first-merged-PR activation, mobile speed, the open source showcase, fresh articles and use-case pages, and Slack team invites.

  2. Days 31 to 6011 ideas

    Co-hosted hackathons, signal outreach on Capy agents, more perk libraries, the harness leaderboard, release clips, enterprise pilots, monthly value reports, cost per task and the bring-your-plan page; PR footers and Discord office hours.

  3. Days 61 to 904 ideas

    Capy on Capy build log, the public automation library, agency partners and the orchestration course.

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 many agents running at once, each on its own machine. The most cited problem: laptops limit how many agents can run.

  • PRs driven to merge through CI and review comments
  • Review becomes the bottleneck
  • In their words: “merged PRs”

Market

Competitive landscape

5 competitors compared: Devin, Codex cloud, GitHub Copilot coding agent, Cursor and 1 more.

  • Gap to exploit: Measured, public proof of merged PRs per team
  • Devin: Same autonomy claim, so proof decides
  • Codex cloud: Tied to one model family

Hiring signals

Capy is hiring its first dedicated growth person and more full-stack engineers.

  • Member of Growth Staff: Find, try, bring the team; content, events, outreach automation, funnel to first merged PR. Ideas 1, 2, 4-11.
  • Member of Technical Staff: First-minute magic in a crowded market; Capy built with Capy. Ideas 3, 16, 17.

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