ProgrammableMarketing/doplexlabs.vercel.app (opens in a new tab)
PDF

Views

  • Plan
  • Roadmap
  • Research

On this page

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

growth playbook · Nov 4, 2026

doplexlabs.vercel.app (opens in a new tab)

Doplex Labs growth playbook

Diagnosis

Doplex Labs ships fast and well: two products in two months, a published white paper with open data, a fast site and funny Shorts that reached about 3,500 views. Growth is the gap.

Rotty has 8 users because the videos give viewers no clickable or easy-to-say path to install, links are not tagged, the lab site's share card is broken and everything lives on three vercel.app addresses. This playbook fixes the path from attention to install first, then builds share loops (weekly card, game, Slop Index), developer distribution for Belay (directories, GitHub, the preprint) and a paid Teams plan.

Growth Score

6/10

Biggest Opportunity

The jokes already find the right audience. Give every view a one-step path to Rotty and measure it.

Biggest Risk

A crowded field of free slop filters and coding-agent tools, and no revenue yet: growth without a paid plan can stall the company.

01

Start here

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

  1. 1

    Turn Rotty's 3,500 YouTube views into installs: a link people can actually click or say

    Pick the short address and add it to the end card of the next Short.

    High impact First result in 1 week
  2. 2

    Tag every route to the Chrome Web Store, so you know which channel brings users

    Tag the lab homepage's Rotty button.

    Medium impact First result in 1 week
  3. 3

    Ask happy Rotty users for a store rating at the right moment

    Pick the threshold and write the line.

    High impact First result in 2 weeks

02

Your idea portfolio

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

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

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

Effort vs impactEach number is an idea. Select one to open it.
Impact →
Quick wins · 4Big bets · 8Fill-ins · 8Reconsider · 2
Effort →

Showing 22 of 22 ideas

This week 6

Quick wins with a clear first action.

  • 01

    Doplex's funny Shorts work: the bot-family jokes reached 1,002 and 877 views. But Rotty has 8 users. The Shorts point people to "the pinned comment", "comment woof" or "search Rotty", and YouTube has not made links in Shorts descriptions or comments clickable since 2023. The address to type, getrotty.vercel.app, is hard to say or remember. Almost none of the attention reaches the install button.

    High impactLow effort First result in 1 week Growth lead

    First action

    Pick the short address and add it to the end card of the next Short.

    How to measure it

    Installs tagged utm_source=youtube

    Now
    –
    Target
    100
    By
    30 days

    Steps

    1. 1

      Pick the address

      A short domain or path that is easy to say, pointing to the Rotty store page.

      Founder 1 day

    2. 2

      Tag the links

      utm_source=youtube, utm_campaign=<video>, for every video.

      Growth lead 1 day

    3. 3

      Fix the long video

      Tagged store link as the first line of the "Introducing Rotty" description.

      Video creator 1 hour

    4. 4

      Link the Shorts

      Set the related video on all 11 Shorts to the Rotty intro.

      Video creator 2 hours

    5. 5

      Add an end card

      Address on screen and spoken in the last two seconds.

      Video creator ongoing

    6. 6

      Profile links

      Tagged store link in the channel links near Subscribe.

      Growth lead 1 hour

    7. 7

      Read the report

      Installs by campaign in the store dashboard each week.

      Growth lead weekly

    First result
    1 week
    Cost
    An hour of editing per video, plus a domain
    Owner
    Growth lead

    Smallest test

    Add the spoken address and related-video link to the next Short and compare installs with the last one.

    Why it works

    Give every video one path to install. Put a short, easy-to-say address on screen in the last two seconds of each Short, and say it out loud. Link each Short to the long "Introducing Rotty" video with YouTube's related-video setting, and put a tagged Chrome Web Store link at the top of that video's description. Add the same tagged link to the channel's profile links. Use utm_source=youtube with one campaign per video, the same way getrotty already tags its own buttons, so the store's install report shows which joke sold the extension.

    People who laughed at a bot joke are the exact audience for a bot filter. Right now they have to do three steps to find it. One tap or one remembered word turns views into users. That gives a growth hire a measured funnel to improve from week one.

    Ready to use

    ChecklistYouTube to install checklist

    [ ] Short address that redirects to the Rotty store page
    [ ] utm_source=youtube&utm_campaign=<video> on every link
    [ ] Tagged link first in the long video description
    [ ] Related video set on every Short
    [ ] Address on screen and spoken at the end
    [ ] Tagged link in channel profile links
    [ ] Weekly installs by campaign

    Risks and how to handle them

    • End cards hurt watch time

      Keep it to two seconds after the punchline.

    • Store reports lag

      Also count clicks on the short address.

    Evidence

    • YouTubeYouTube @DoplexLabs: 6 subscribers, 13 videos, about 3,500 views. Top videos are jokes: "My dad has been arguing with a bot on Facebook since 9am" (1,002 views), "The bots' group chat right now" (877), "the bots will flirt with anything." (464); the two "Introducing Rotty" videos have 125 and 62. Descriptions say "Download link in the pinned comment, or search 'Rotty'", "comment 'woof' and i'll reply with the link" or just "Released on Chrome Store"; one promises "Facebook is next." Latest upload October 8, 2026.Source
    • News coverageSince August 31, 2023 YouTube has not made links in Shorts descriptions or Shorts comments clickable, to fight spam; links in long-form video descriptions still work.Source
    • Chrome Web StoreRotty listing: 8 users, 2 ratings (5.0), version 0.1.0, updated October 8, 2026, in Social Networking. "Hide AI slop, filler, engagement bait and spam on Reddit, X and LinkedIn. Free with your ChatGPT plan, a free key, or on-device."Source
    • Your websitegetrotty.vercel.app: a full-URL share image, a sitemap of 4 pages, and every "Add to Chrome" button tagged utm_source=getrotty with a campaign per placement (nav, hero, faq, closing, footer); a feedback page; contact is a gmail address; Product Hunt and Spacerr badges; "+ more sites soon". It links to Belay in the footer.Source
  • 02

    The lab homepage gives its share image as a relative path, "brand/og-image.png", with no og:url, no twitter:image and no canonical link. X, LinkedIn and Slack expect a full URL, so a shared link to the lab may show no picture. There is also no robots.txt or sitemap. Rotty's own site already does all of this correctly.

    Medium impactLow effort First result in 1 week Founding engineer

    First action

    Change og:image to the full URL and test it in a link preview tool.

    How to measure it

    Platforms showing the image

    Now
    –
    Target
    X, LinkedIn, Slack
    By
    1 week

    Steps

    1. 1

      Full URLs

      og:image, og:url, twitter:image.

      Founding engineer 1 hour

    2. 2

      Canonical

      One canonical URL for the homepage.

      Founding engineer 15 minutes

    3. 3

      Robots and sitemap

      For the lab site and Belay.

      Founding engineer 1 hour

    4. 4

      Structured data

      Organization with products and profiles.

      Founding engineer 1 hour

    5. 5

      Test previews

      X, LinkedIn and Slack.

      Growth lead 1 hour

    6. 6

      Submit

      Add both sitemaps to Google Search Console.

      Growth lead 30 minutes

    First result
    1 week
    Cost
    An afternoon
    Owner
    Founding engineer

    Smallest test

    Fix the og:image line today.

    Why it works

    Copy Rotty's head tags: a full https:// og:image, og:url, twitter:image and a canonical link. Add robots.txt and a sitemap. Do the same on getbelay.vercel.app, which has share tags but no sitemap. Add Organization structured data that lists Belay, Rotty and the social profiles.

    Doplex posts on X, TikTok, Instagram and YouTube. Every link shared there should show the brand image. Search engines and AI assistants also find a new site faster with a sitemap.

    Risks and how to handle them

    • Previews are cached

      Re-scrape with each platform's preview tool.

    Evidence

    • Your websiteLab homepage code: og:image is the relative path "brand/og-image.png" (the file exists, but the tag has no full URL), there is no og:url, no twitter:image and no canonical link, and /robots.txt and /sitemap.xml both return 404. The Internet Archive first saw the site on October 8, 2026.Source
    • Your websitegetrotty.vercel.app: a full-URL share image, a sitemap of 4 pages, and every "Add to Chrome" button tagged utm_source=getrotty with a campaign per placement (nav, hero, faq, closing, footer); a feedback page; contact is a gmail address; Product Hunt and Spacerr badges; "+ more sites soon". It links to Belay in the footer.Source
    • Your websitegetbelay.vercel.app: "Belay debriefs your Claude Code and Codex sessions, then carries the lesson into the next one"; a one-line install for macOS and Windows, "Free. Local. 30 sec install."; lists Claude Code, Codex, Cursor and Antigravity; source under the MIT License on GitHub; "Belay for Teams · Enterprise Coming soon" with a waitlist; a "Talk to the founders" form. It has full share tags and a canonical URL, but no robots.txt or sitemap.Source
    • Your websiteLab homepage (checked November 4, 2026): "Doplex Labs is an applied AI lab"; products Belay ("Alpha", "Debriefs your coding agent's sessions"), Rotty ("Live", "A guard dog for your feed") and an "Unannounced product" with a "Get early access" button; channels X, TikTok, Instagram and YouTube. The Rotty button links straight to the Chrome Web Store with no campaign tags, and early access and contact both open one chat box that sends a free-text message.Source
  • 03

    getrotty tags its own buttons by placement, but the lab homepage's Rotty button links to the store with no tags, and so do the YouTube descriptions and social bios. The extension itself has no analytics, by design. The store's source report is the only place to see where users come from, so untagged links leave growth blind.

    Medium impactLow effort First result in 1 week Growth lead

    First action

    Tag the lab homepage's Rotty button.

    How to measure it

    Installs with a source tag

    Now
    –
    Target
    80%
    By
    30 days

    Steps

    1. 1

      List channels

      Every place a store link appears.

      Growth lead 1 hour

    2. 2

      Build the table

      One tagged URL each.

      Growth lead 1 hour

    3. 3

      Replace links

      Site, bios, descriptions, launch pages.

      Growth lead 1 day

    4. 4

      Short links

      Readable redirects for bios and video.

      Founding engineer 2 hours

    5. 5

      Weekly read

      Installs and uninstalls by source.

      Growth lead weekly

    6. 6

      Share it

      A one-line weekly note to the founders.

      Growth lead weekly

    First result
    1 week
    Cost
    None
    Owner
    Growth lead

    Smallest test

    Tag the lab homepage link and the X bio today.

    Why it works

    Make one link table: one tagged store URL per channel (lab site, YouTube, X, TikTok, Instagram, Product Hunt, Spacerr, emails). Use only those links. Check the store dashboard every week and spend time on what brings installs.

    Rotty promises no tracking inside the extension. Tagged links let you measure growth without breaking that promise.

    Risks and how to handle them

    • Some installs come from store search

      Treat untagged installs as search and track store keywords.

    Evidence

    • Your websiteLab homepage (checked November 4, 2026): "Doplex Labs is an applied AI lab"; products Belay ("Alpha", "Debriefs your coding agent's sessions"), Rotty ("Live", "A guard dog for your feed") and an "Unannounced product" with a "Get early access" button; channels X, TikTok, Instagram and YouTube. The Rotty button links straight to the Chrome Web Store with no campaign tags, and early access and contact both open one chat box that sends a free-text message.Source
    • Your websitegetrotty.vercel.app: a full-URL share image, a sitemap of 4 pages, and every "Add to Chrome" button tagged utm_source=getrotty with a campaign per placement (nav, hero, faq, closing, footer); a feedback page; contact is a gmail address; Product Hunt and Spacerr badges; "+ more sites soon". It links to Belay in the footer.Source
    • Chrome Web StoreRotty's listing: "It never claims to detect whether a person or a machine wrote a post"; five kinds of junk, each with its own switch; the toolbar popup shows "how many were hidden this week"; models are your ChatGPT plan, a free Groq key, Rotty Local in the browser, or your own OpenAI, Anthropic or Google key with a 50-cent daily cap; "No servers", "no analytics or tracking in the extension".Source
    • YouTubeYouTube @DoplexLabs: 6 subscribers, 13 videos, about 3,500 views. Top videos are jokes: "My dad has been arguing with a bot on Facebook since 9am" (1,002 views), "The bots' group chat right now" (877), "the bots will flirt with anything." (464); the two "Introducing Rotty" videos have 125 and 62. Descriptions say "Download link in the pinned comment, or search 'Rotty'", "comment 'woof' and i'll reply with the link" or just "Released on Chrome Store"; one promises "Facebook is next." Latest upload October 8, 2026.Source
  • 04

    Rotty has 2 ratings. AI Slop Blocker, a direct rival, has 358 users and 9 ratings. The popup already shows how many posts Rotty hid this week, which is the perfect moment to ask for a rating.

    High impactLow effort First result in 2 weeks Founding engineer

    First action

    Pick the threshold and write the line.

    How to measure it

    Ratings

    Now
    –
    Target
    25
    By
    60 days

    Steps

    1. 1

      Set the trigger

      50 hidden, no "Not slop" today.

      Founding engineer 1 day

    2. 2

      Write the copy

      One line in Rotty's voice.

      Growth lead 1 hour

    3. 3

      Build it

      Popup line with store link, shown once.

      Founding engineer 2 days

    4. 4

      Ship 0.1.1

      Release note mentions it.

      Founding engineer 1 day

    5. 5

      Reply to reviews

      Every review, within a day.

      Founder ongoing

    6. 6

      Measure

      Ratings per week.

      Growth lead weekly

    First result
    2 weeks
    Cost
    Engineering time
    Owner
    Founding engineer

    Smallest test

    Ask for ratings in the pinned comment of the next Short and count new ratings that week.

    Why it works

    When a user has had at least 50 posts hidden and has not marked anything "Not slop" that day, show one line in the popup: "Rotty hid 50 posts for you. A rating helps others find it." Link to the store reviews page. Ask once and never again after a click or dismiss. Answer every review in the store.

    Ratings help ranking and trust in store search, where rivals already sit. Asking after a clear win gets kinder, more detailed reviews.

    Risks and how to handle them

    • Feels pushy

      Once only, after a win, easy to dismiss.

    Evidence

    • Chrome Web StoreRotty listing: 8 users, 2 ratings (5.0), version 0.1.0, updated October 8, 2026, in Social Networking. "Hide AI slop, filler, engagement bait and spam on Reddit, X and LinkedIn. Free with your ChatGPT plan, a free key, or on-device."Source
    • Competitor researchOther feed filters on the Chrome Web Store: AI Slop Blocker ("Hides AI-generated posts in your feed, including YouTube videos. Scoring runs on your device.") with 358 users and 9 ratings, updated October 31, 2026; FeedShield with 58 users. For coding agents, SpecStory: "Save the conversations behind your code with SpecStory. RunStory tests your software and sends defects back to your coding agent."Source
    • Chrome Web StoreRotty's listing: "It never claims to detect whether a person or a machine wrote a post"; five kinds of junk, each with its own switch; the toolbar popup shows "how many were hidden this week"; models are your ChatGPT plan, a free Groq key, Rotty Local in the browser, or your own OpenAI, Anthropic or Google key with a 50-cent daily cap; "No servers", "no analytics or tracking in the extension".Source
  • 05

    The lab homepage's "Get early access" button and contact both open a free-text chat box. Belay for Teams has a separate waitlist. There is no single list you can email on launch day.

    Medium impactLow effort First result in 2 weeks Growth lead

    First action

    Add the email field to the chat box.

    How to measure it

    List size

    Now
    –
    Target
    500
    By
    90 days

    Steps

    1. 1

      Pick a tool

      A simple newsletter tool with tags.

      Growth lead 1 day

    2. 2

      Add the field

      Email plus one choice.

      Founding engineer 1 day

    3. 3

      Merge lists

      Import the Teams waitlist.

      Growth lead 1 hour

    4. 4

      Welcome email

      What Doplex builds and why.

      Founder 1 day

    5. 5

      Monthly note

      Shipped, measured, next.

      Founder monthly

    6. 6

      Track

      Signups by source.

      Growth lead weekly

    First result
    2 weeks
    Cost
    Newsletter tool, low cost
    Owner
    Growth lead

    Smallest test

    Add the email field this week.

    Why it works

    Add an email field and one choice ("Rotty updates / Belay for Teams / the next product") to the chat box. Send all signups to one list with tags. Email it once a month with what shipped and what was learned, like the white paper.

    An owned list makes each launch start with people who asked for it, and does not depend on any social platform.

    Risks and how to handle them

    • Low volume at first

      Promote it in every video and launch post.

    Evidence

    • Your websiteLab homepage (checked November 4, 2026): "Doplex Labs is an applied AI lab"; products Belay ("Alpha", "Debriefs your coding agent's sessions"), Rotty ("Live", "A guard dog for your feed") and an "Unannounced product" with a "Get early access" button; channels X, TikTok, Instagram and YouTube. The Rotty button links straight to the Chrome Web Store with no campaign tags, and early access and contact both open one chat box that sends a free-text message.Source
    • Your websitegetbelay.vercel.app: "Belay debriefs your Claude Code and Codex sessions, then carries the lesson into the next one"; a one-line install for macOS and Windows, "Free. Local. 30 sec install."; lists Claude Code, Codex, Cursor and Antigravity; source under the MIT License on GitHub; "Belay for Teams · Enterprise Coming soon" with a waitlist; a "Talk to the founders" form. It has full share tags and a canonical URL, but no robots.txt or sitemap.Source
  • 06

    The lab homepage scores 99 on mobile and 100 for SEO. The one fix left is about 0.65 s of render-blocking Google Fonts (70 KB). Rotty's site already serves its own Geist file.

    Low impactLow effort First result in 1 day Founding engineer

    First action

    Copy the font setup from getrotty.

    How to measure it

    Largest paint on mobile

    Now
    –
    Target
    under 1.3 s
    By
    1 week

    Steps

    1. 1

      Download fonts

      Only the weights used.

      Founding engineer 1 hour

    2. 2

      Serve locally

      Same folder as the brand files.

      Founding engineer 1 hour

    3. 3

      Preload

      The main text weight.

      Founding engineer 15 minutes

    4. 4

      Remove links

      Google Fonts tags.

      Founding engineer 15 minutes

    5. 5

      Re-test

      PageSpeed on mobile.

      Founding engineer 15 minutes

    6. 6

      Repeat on Belay

      Same setup.

      Founding engineer 1 hour

    First result
    1 day
    Cost
    An hour
    Owner
    Founding engineer

    Smallest test

    Self-host the main Geist weight first.

    Why it works

    Serve Geist, Geist Mono and Instrument Serif from the site with font-display: swap and preload the main weight, as getrotty does.

    This is small, but it is a free win, and it removes the only third-party request on the page.

    Risks and how to handle them

    • Font licence

      All three are open fonts.

    Evidence

    • Google PageSpeedLab homepage: mobile performance 99, desktop 96, accessibility, best practices and SEO 100; mobile largest paint 1.7 s, 94 KB over 9 requests. The one fix listed is render-blocking requests (about 0.65 s), from 70 KB of Google Fonts.Source
    • Your websitegetrotty.vercel.app: a full-URL share image, a sitemap of 4 pages, and every "Add to Chrome" button tagged utm_source=getrotty with a campaign per placement (nav, hero, faq, closing, footer); a feedback page; contact is a gmail address; Product Hunt and Spacerr badges; "+ more sites soon". It links to Belay in the footer.Source

This quarter 10

The core plays for the next three months.

  • 07

    The popup already counts posts hidden this week, by reason. Users can't share that number, and each share would be a free ad on X and LinkedIn, the same feeds Rotty cleans.

    High impactMedium effort First result in 1 month Founding engineer

    First action

    Design the card in Rotty's style.

    How to measure it

    Installs tagged utm_source=card

    Now
    –
    Target
    200
    By
    1 quarter

    Steps

    1. 1

      Design

      One card, counts by kind.

      Video creator 1 week

    2. 2

      Build

      Make the image on the device.

      Founding engineer 1 week

    3. 3

      Share options

      X, LinkedIn, download.

      Founding engineer 3 days

    4. 4

      Tag the link

      utm_source=card.

      Growth lead 1 hour

    5. 5

      Seed it

      Founders post their own cards.

      Founder 1 week

    6. 6

      Measure

      Shares and installs.

      Growth lead weekly

    First result
    1 month
    Cost
    Engineering time
    Owner
    Founding engineer

    Smallest test

    Post a hand-made card from your own week and count clicks.

    Why it works

    Add a "Share my week" button to the popup that makes an image on the device: hidden posts by kind (filler, bait, plugs, spam), the site, and a tagged install link. No post text or usernames, so it keeps the privacy promise.

    The number is funny and a little shocking, and it shows the product at work. That is the kind of post people share without being asked.

    Risks and how to handle them

    • Leaks private data

      Counts only, never post text.

    Evidence

    • Chrome Web StoreRotty's listing: "It never claims to detect whether a person or a machine wrote a post"; five kinds of junk, each with its own switch; the toolbar popup shows "how many were hidden this week"; models are your ChatGPT plan, a free Groq key, Rotty Local in the browser, or your own OpenAI, Anthropic or Google key with a 50-cent daily cap; "No servers", "no analytics or tracking in the extension".Source
    • Chrome Web StoreRotty listing: 8 users, 2 ratings (5.0), version 0.1.0, updated October 8, 2026, in Social Networking. "Hide AI slop, filler, engagement bait and spam on Reddit, X and LinkedIn. Free with your ChatGPT plan, a free key, or on-device."Source
    • Your websitegetrotty.vercel.app: a full-URL share image, a sitemap of 4 pages, and every "Add to Chrome" button tagged utm_source=getrotty with a campaign per placement (nav, hero, faq, closing, footer); a feedback page; contact is a gmail address; Product Hunt and Spacerr badges; "+ more sites soon". It links to Belay in the footer.Source
  • 08

    Rivals already sit in the store: AI Slop Blocker (358 users) and FeedShield (58). Rotty's difference is clear: it judges substance and never claims to detect AI, it says why it hid each post, and you choose the model.

    Medium impactMedium effort First result in 1 month Growth lead

    First action

    Install each rival and list the facts.

    How to measure it

    Compare-page installs

    Now
    –
    Target
    50 a month
    By
    1 quarter

    Steps

    1. 1

      Test rivals

      Install and use each one.

      Growth lead 1 week

    2. 2

      Write the table

      Facts only, with dates.

      Growth lead 3 days

    3. 3

      Build the page

      On getrotty, in the sitemap.

      Founding engineer 2 days

    4. 4

      Link it

      Store description, FAQ.

      Growth lead 1 day

    5. 5

      Keep it current

      Recheck monthly.

      Growth lead monthly

    6. 6

      Measure

      Visits and tagged installs.

      Growth lead monthly

    First result
    1 month
    Cost
    Writing time
    Owner
    Growth lead

    Smallest test

    Write the table this week.

    Why it works

    Publish a /compare page on getrotty with a fair table: sites covered, what is hidden, how it decides, model choice, privacy and price. Link it from the store description and the FAQ.

    People who search for "AI slop blocker" are already shopping. A clear, fair page helps them choose, and is easy for AI assistants to quote.

    Risks and how to handle them

    • Seems unfair

      Credit rival strengths and date every fact.

    Evidence

    • Competitor researchOther feed filters on the Chrome Web Store: AI Slop Blocker ("Hides AI-generated posts in your feed, including YouTube videos. Scoring runs on your device.") with 358 users and 9 ratings, updated October 31, 2026; FeedShield with 58 users. For coding agents, SpecStory: "Save the conversations behind your code with SpecStory. RunStory tests your software and sends defects back to your coding agent."Source
    • Chrome Web StoreRotty's listing: "It never claims to detect whether a person or a machine wrote a post"; five kinds of junk, each with its own switch; the toolbar popup shows "how many were hidden this week"; models are your ChatGPT plan, a free Groq key, Rotty Local in the browser, or your own OpenAI, Anthropic or Google key with a 50-cent daily cap; "No servers", "no analytics or tracking in the extension".Source
    • Your websitegetrotty.vercel.app: a full-URL share image, a sitemap of 4 pages, and every "Add to Chrome" button tagged utm_source=getrotty with a campaign per placement (nav, hero, faq, closing, footer); a feedback page; contact is a gmail address; Product Hunt and Spacerr badges; "+ more sites soon". It links to Belay in the footer.Source
  • 09

    Belay's best moment is the debrief: "You named the files on your fourth message... Scoping about 24 min · roughly $6", with the exact message to send next time. The YouTube channel only covers Rotty so far.

    High impactMedium effort First result in 1 month Founder

    First action

    Record one debrief from your own sessions.

    How to measure it

    Install runs from video links

    Now
    –
    Target
    300
    By
    1 quarter

    Steps

    1. 1

      Pick sessions

      Own repos, clear mistakes.

      Founder 1 week

    2. 2

      Record

      60 seconds each.

      Video creator 1 week

    3. 3

      Write the hook

      The cost first.

      Growth lead 2 days

    4. 4

      Post

      X and YouTube, weekly.

      Video creator weekly

    5. 5

      Tag links

      Install page with campaign.

      Growth lead 1 hour

    6. 6

      Measure

      Install page visits.

      Growth lead weekly

    First result
    1 month
    Cost
    Recording time
    Owner
    Founder

    Smallest test

    Post one debrief this week.

    Why it works

    Record 60-second screen captures of real debriefs (with permission or on your own repos): the mistake, the cost, the better prompt, the approved rule. Post them on X and YouTube with the install line in the long-form description.

    Developers share things that save them time and money. A real debrief with a dollar figure is concrete proof that the product works.

    Risks and how to handle them

    • Private code on screen

      Use own open repos only.

    Evidence

    • Your websitegetbelay.vercel.app: "Belay debriefs your Claude Code and Codex sessions, then carries the lesson into the next one"; a one-line install for macOS and Windows, "Free. Local. 30 sec install."; lists Claude Code, Codex, Cursor and Antigravity; source under the MIT License on GitHub; "Belay for Teams · Enterprise Coming soon" with a waitlist; a "Talk to the founders" form. It has full share tags and a canonical URL, but no robots.txt or sitemap.Source
    • YouTubeYouTube @DoplexLabs: 6 subscribers, 13 videos, about 3,500 views. Top videos are jokes: "My dad has been arguing with a bot on Facebook since 9am" (1,002 views), "The bots' group chat right now" (877), "the bots will flirt with anything." (464); the two "Introducing Rotty" videos have 125 and 62. Descriptions say "Download link in the pinned comment, or search 'Rotty'", "comment 'woof' and i'll reply with the link" or just "Released on Chrome Store"; one promises "Facebook is next." Latest upload October 8, 2026.Source
    • Your websiteBelay white paper (version 1.1, September 15, 2026): agents met acceptance criteria on 11 of 16 tasks with a Belay lesson, 7 of 16 with a human-written instruction and 6 of 16 with no guidance; result CSVs are published; Belay exposes a "bounded stdio MCP server"; next studies are agent-specific delivery, explicit scope, focused verification and "Months of real repos, not one benchmark"; "full methodology and agent-level results will follow in a research preprint."Source
  • 10

    Belay runs a local MCP server and works with Claude Code, Codex, Cursor and Antigravity. Developers find tools like this in MCP directories, plugin marketplaces and curated GitHub lists.

    High impactLow effort First result in 1 month Founding engineer

    First action

    Make a list of ten directories and lists.

    How to measure it

    Listings

    Now
    –
    Target
    10
    By
    1 month

    Steps

    1. 1

      Find directories

      MCP and plugin catalogues.

      Growth lead 2 days

    2. 2

      Write one entry

      Name, line, install, privacy.

      Growth lead 1 day

    3. 3

      Package

      Whatever format each one needs.

      Founding engineer 1 week

    4. 4

      Submit

      Each directory and list.

      Growth lead 1 week

    5. 5

      Tag links

      One source per listing.

      Growth lead 1 hour

    6. 6

      Track

      Install page visits by source.

      Growth lead monthly

    First result
    1 month
    Cost
    Engineering time
    Owner
    Founding engineer

    Smallest test

    Submit to two directories this week.

    Why it works

    Submit Belay to the main MCP server directories and plugin marketplaces, and open pull requests to the curated lists for Claude Code and Codex tools. Use the same short description and a tagged install link.

    These are the exact places users go to add tools to their coding agents. Every listing is a lasting source of installs.

    Risks and how to handle them

    • Listings reviewed slowly

      Submit early and follow up.

    Evidence

    • Your websitegetbelay.vercel.app: "Belay debriefs your Claude Code and Codex sessions, then carries the lesson into the next one"; a one-line install for macOS and Windows, "Free. Local. 30 sec install."; lists Claude Code, Codex, Cursor and Antigravity; source under the MIT License on GitHub; "Belay for Teams · Enterprise Coming soon" with a waitlist; a "Talk to the founders" form. It has full share tags and a canonical URL, but no robots.txt or sitemap.Source
    • Your websiteBelay white paper (version 1.1, September 15, 2026): agents met acceptance criteria on 11 of 16 tasks with a Belay lesson, 7 of 16 with a human-written instruction and 6 of 16 with no guidance; result CSVs are published; Belay exposes a "bounded stdio MCP server"; next studies are agent-specific delivery, explicit scope, focused verification and "Months of real repos, not one benchmark"; "full methodology and agent-level results will follow in a research preprint."Source
    • GitHubGitHub organisation DoplexLabs: belay (1 star), belay-engine-os (0 stars, 6 forks), rotty-open-source (2 stars) and numbat (0 stars), all created in September and October 2026.Source
  • 11

    The repos have 0 to 2 stars, while belay-engine-os already has 6 forks: people are using the code before the repos sell it.

    Medium impactLow effort First result in 2 weeks Founding engineer

    First action

    Rewrite the belay README.

    How to measure it

    Stars across repos

    Now
    –
    Target
    200
    By
    1 quarter

    Steps

    1. 1

      READMEs

      What, GIF, install, link.

      Founding engineer 1 week

    2. 2

      Topics

      Relevant tags.

      Founding engineer 1 hour

    3. 3

      Releases

      Tagged versions with notes.

      Founding engineer 1 day

    4. 4

      Pin repos

      Organisation page.

      Founding engineer 15 minutes

    5. 5

      Star line

      In README and install output.

      Growth lead 1 hour

    6. 6

      Track

      Stars and forks monthly.

      Growth lead monthly

    First result
    2 weeks
    Cost
    A few days
    Owner
    Founding engineer

    Smallest test

    Rewrite one README this week.

    Why it works

    Give each repo a README that opens with what it does, a GIF, the one-line install and a link to the product site. Add topics (claude-code, codex, mcp, chrome-extension), a release page and a gentle "star if this helped" line. Pin the two product repos on the organisation page.

    GitHub is where developers check if a tool is real. Stars and topics also bring the repo into GitHub search and trending lists.

    Risks and how to handle them

    • Begging for stars

      Ask once, politely.

    Evidence

    • GitHubGitHub organisation DoplexLabs: belay (1 star), belay-engine-os (0 stars, 6 forks), rotty-open-source (2 stars) and numbat (0 stars), all created in September and October 2026.Source
    • Your websitegetbelay.vercel.app: "Belay debriefs your Claude Code and Codex sessions, then carries the lesson into the next one"; a one-line install for macOS and Windows, "Free. Local. 30 sec install."; lists Claude Code, Codex, Cursor and Antigravity; source under the MIT License on GitHub; "Belay for Teams · Enterprise Coming soon" with a waitlist; a "Talk to the founders" form. It has full share tags and a canonical URL, but no robots.txt or sitemap.Source
  • 12

    The white paper promises "a research preprint" with full method and agent results. The headline is strong: 11 of 16 tasks with a Belay lesson, against 7 with a human-written rule and 6 with none, and the CSVs are public.

    High impactMedium effort First result in 1 quarter Founder

    First action

    Set the preprint date.

    How to measure it

    Launch week visits

    Now
    –
    Target
    5,000
    By
    1 quarter

    Steps

    1. 1

      Finish the paper

      Method and agent results.

      Founder 1 month

    2. 2

      One chart

      The 11/7/6 result.

      Growth lead 2 days

    3. 3

      Thread

      Short, with the chart.

      Founder 1 day

    4. 4

      Show HN

      Same day, with data links.

      Founder 1 day

    5. 5

      Newsletters

      Pitch five.

      Growth lead 1 week

    6. 6

      Follow up

      Answer every comment.

      Founder 1 week

    First result
    1 quarter
    Cost
    Writing time
    Owner
    Founder

    Smallest test

    Draft the launch thread now.

    Why it works

    Publish the preprint, post a short thread with the one chart, and do a "Show HN" on the day with the open data. Offer the result to newsletters on AI coding.

    Evidence is Doplex's stated value ("We publish what we measure"). A preprint is a launch that people share because it is useful, not because it is an ad.

    Risks and how to handle them

    • Small sample questioned

      Say n=16 up front and invite replication.

    Evidence

    • Your websiteBelay white paper (version 1.1, September 15, 2026): agents met acceptance criteria on 11 of 16 tasks with a Belay lesson, 7 of 16 with a human-written instruction and 6 of 16 with no guidance; result CSVs are published; Belay exposes a "bounded stdio MCP server"; next studies are agent-specific delivery, explicit scope, focused verification and "Months of real repos, not one benchmark"; "full methodology and agent-level results will follow in a research preprint."Source
    • Your websiteLab homepage (checked November 4, 2026): "Doplex Labs is an applied AI lab"; products Belay ("Alpha", "Debriefs your coding agent's sessions"), Rotty ("Live", "A guard dog for your feed") and an "Unannounced product" with a "Get early access" button; channels X, TikTok, Instagram and YouTube. The Rotty button links straight to the Chrome Web Store with no campaign tags, and early access and contact both open one chat box that sends a free-text message.Source
  • 13

    Belay for Teams is "Coming soon" with a waitlist, promising "Fleet-wide patterns" and "One fix for everyone". There is no pricing yet.

    High impactMedium effort First result in 6 weeks Founder

    First action

    Email the waitlist asking who wants to be a design partner.

    How to measure it

    Teams in the programme

    Now
    –
    Target
    5
    By
    1 quarter

    Steps

    1. 1

      Pick teams

      From the waitlist.

      Founder 1 week

    2. 2

      Agree terms

      Six weeks, feedback, price talk.

      Founder 1 week

    3. 3

      Onboard

      Install on each machine.

      Founding engineer 1 week

    4. 4

      Meet

      Every two weeks.

      Founder 6 weeks

    5. 5

      Measure

      Repeated failures before and after.

      Founding engineer 6 weeks

    6. 6

      Price

      Agree a paid plan.

      Founder 1 week

    7. 7

      Case studies

      With permission.

      Growth lead 2 weeks

    First result
    6 weeks
    Cost
    Founder time
    Owner
    Founder

    Smallest test

    Email the waitlist this week.

    Why it works

    Pick five teams from the waitlist and "Talk to the founders" messages. Run Belay on their machines for six weeks, meet every two weeks, and agree a price before the end. Use their results as the first case studies.

    Teams pay; individuals often don't. Design partners shape the product and become the first revenue and proof.

    Risks and how to handle them

    • Security reviews slow it

      Lead with the local-only design.

    Evidence

    • Your websitegetbelay.vercel.app: "Belay debriefs your Claude Code and Codex sessions, then carries the lesson into the next one"; a one-line install for macOS and Windows, "Free. Local. 30 sec install."; lists Claude Code, Codex, Cursor and Antigravity; source under the MIT License on GitHub; "Belay for Teams · Enterprise Coming soon" with a waitlist; a "Talk to the founders" form. It has full share tags and a canonical URL, but no robots.txt or sitemap.Source
  • 14

    The best-watched Short (1,002 views) is about a dad arguing with a bot on Facebook, and its description says "Facebook is next." Rotty works only on Reddit, X and LinkedIn today, and its site says "+ more sites soon".

    High impactMedium effort First result in 1 quarter Founding engineer

    First action

    Scope Facebook support.

    How to measure it

    Installs in launch week

    Now
    –
    Target
    300
    By
    1 quarter

    Steps

    1. 1

      Scope

      Feed and comments.

      Founding engineer 1 week

    2. 2

      Build

      Same hidden-line style.

      Founding engineer 1 month

    3. 3

      Test

      With real feeds.

      Founding engineer 1 week

    4. 4

      Make the Short

      Sequel to the dad video.

      Video creator 1 week

    5. 5

      Update listing

      Store title and images.

      Growth lead 1 day

    6. 6

      Launch

      All channels and the list.

      Growth lead 1 day

    First result
    1 quarter
    Cost
    Engineering time
    Owner
    Founding engineer

    Smallest test

    Post a "Facebook is coming" Short to test interest.

    Why it works

    Build Facebook support, then launch it as an event: a new Short in the same style, an update to the store listing, a post on each social channel and an email to the list. Do the same for each new site.

    Each new site is a new reason to talk about Rotty, and Facebook comment sections are where many people meet bots. The audience is already asking for it.

    Risks and how to handle them

    • Facebook layout changes often

      Plan for quick fixes.

    Evidence

    • YouTubeYouTube @DoplexLabs: 6 subscribers, 13 videos, about 3,500 views. Top videos are jokes: "My dad has been arguing with a bot on Facebook since 9am" (1,002 views), "The bots' group chat right now" (877), "the bots will flirt with anything." (464); the two "Introducing Rotty" videos have 125 and 62. Descriptions say "Download link in the pinned comment, or search 'Rotty'", "comment 'woof' and i'll reply with the link" or just "Released on Chrome Store"; one promises "Facebook is next." Latest upload October 8, 2026.Source
    • Your websitegetrotty.vercel.app: a full-URL share image, a sitemap of 4 pages, and every "Add to Chrome" button tagged utm_source=getrotty with a campaign per placement (nav, hero, faq, closing, footer); a feedback page; contact is a gmail address; Product Hunt and Spacerr badges; "+ more sites soon". It links to Belay in the footer.Source
    • Chrome Web StoreRotty listing: 8 users, 2 ratings (5.0), version 0.1.0, updated October 8, 2026, in Social Networking. "Hide AI slop, filler, engagement bait and spam on Reddit, X and LinkedIn. Free with your ChatGPT plan, a free key, or on-device."Source
  • 15

    The lab, Belay and Rotty live on three separate vercel.app addresses, first seen in October 2026. Contact is a gmail address. Links and trust are split three ways, and none of the addresses is easy to say in a video.

    Medium impactMedium effort First result in 1 month Founder

    First action

    Choose and buy the domain.

    How to measure it

    Monthly search impressions

    Now
    –
    Target
    10,000
    By
    2 quarters

    Steps

    1. 1

      Pick the domain

      Short and easy to say.

      Founder 1 week

    2. 2

      Plan URLs

      Lab, /belay, /rotty.

      Founding engineer 1 day

    3. 3

      Redirects

      Permanent, page by page.

      Founding engineer 1 day

    4. 4

      Update links

      Store, GitHub, socials, install script.

      Growth lead 1 week

    5. 5

      Email

      Domain email for contact.

      Founder 1 day

    6. 6

      Search Console

      Add and submit sitemaps.

      Growth lead 1 day

    First result
    1 month
    Cost
    Domain cost
    Owner
    Founder

    Smallest test

    Buy the domain this week.

    Why it works

    Buy one short domain. Move the lab to the root and the products to /belay and /rotty (or short subdomains), with redirects from the old addresses. Move email to the domain. Keep getrotty and getbelay working as redirects.

    One domain builds one brand and search history. A real address and email also help with press, partners and team buyers.

    Risks and how to handle them

    • Breaking the install command

      Keep old install URLs working forever.

    Evidence

    • Your websiteLab homepage code: og:image is the relative path "brand/og-image.png" (the file exists, but the tag has no full URL), there is no og:url, no twitter:image and no canonical link, and /robots.txt and /sitemap.xml both return 404. The Internet Archive first saw the site on October 8, 2026.Source
    • Your websitegetrotty.vercel.app: a full-URL share image, a sitemap of 4 pages, and every "Add to Chrome" button tagged utm_source=getrotty with a campaign per placement (nav, hero, faq, closing, footer); a feedback page; contact is a gmail address; Product Hunt and Spacerr badges; "+ more sites soon". It links to Belay in the footer.Source
    • Your websitegetbelay.vercel.app: "Belay debriefs your Claude Code and Codex sessions, then carries the lesson into the next one"; a one-line install for macOS and Windows, "Free. Local. 30 sec install."; lists Claude Code, Codex, Cursor and Antigravity; source under the MIT License on GitHub; "Belay for Teams · Enterprise Coming soon" with a waitlist; a "Talk to the founders" form. It has full share tags and a canonical URL, but no robots.txt or sitemap.Source
  • 16

    Joke Shorts got 877 to 1,002 views; the two product intros got 125 and 62. Uploads stopped on October 8, 2026, the day Rotty shipped.

    High impactLow effort First result in 1 month Video creator

    First action

    Write ten scripts in the top three formats.

    How to measure it

    Weekly views

    Now
    –
    Target
    10,000
    By
    1 quarter

    Steps

    1. 1

      Review

      What the top three had in common.

      Video creator 1 day

    2. 2

      Script

      Ten ideas.

      Video creator 1 week

    3. 3

      Record

      Batch three a week.

      Video creator weekly

    4. 4

      Cross-post

      TikTok and Instagram.

      Video creator weekly

    5. 5

      End card

      Product as punchline.

      Video creator ongoing

    6. 6

      Review

      Views and installs by video.

      Growth lead weekly

    First result
    1 month
    Cost
    Video time
    Owner
    Video creator

    Smallest test

    Post one Short this week.

    Why it works

    Post three Shorts a week in the formats that worked (bot family, bot group chat, bots flirting). End each with Rotty hiding the bot in one second and the spoken address. Cut the same clip for TikTok and Instagram.

    The audience came for the jokes. Showing the product as the punchline sells it without a sales pitch.

    Risks and how to handle them

    • Burnout

      Batch recording once a week.

    Evidence

    • YouTubeYouTube @DoplexLabs: 6 subscribers, 13 videos, about 3,500 views. Top videos are jokes: "My dad has been arguing with a bot on Facebook since 9am" (1,002 views), "The bots' group chat right now" (877), "the bots will flirt with anything." (464); the two "Introducing Rotty" videos have 125 and 62. Descriptions say "Download link in the pinned comment, or search 'Rotty'", "comment 'woof' and i'll reply with the link" or just "Released on Chrome Store"; one promises "Facebook is next." Latest upload October 8, 2026.Source
    • Your websiteLab homepage (checked November 4, 2026): "Doplex Labs is an applied AI lab"; products Belay ("Alpha", "Debriefs your coding agent's sessions"), Rotty ("Live", "A guard dog for your feed") and an "Unannounced product" with a "Get early access" button; channels X, TikTok, Instagram and YouTube. The Rotty button links straight to the Chrome Web Store with no campaign tags, and early access and contact both open one chat box that sends a free-text message.Source

Big bets 3

Larger plays that can change the trajectory.

  • 17

    Rotty already sorts posts into kinds (filler, bait, plugs, spam). It collects no user data, by design, but Doplex can run the same judge on public threads itself.

    High impactHigh effort First result in 1 quarter Founder

    First action

    Run a first sample of 1,000 public replies.

    How to measure it

    Articles citing the index

    Now
    –
    Target
    10
    By
    2 quarters

    Steps

    1. 1

      Method

      Sampling and rules.

      Founder 2 weeks

    2. 2

      Sample

      Public threads only.

      Founding engineer 1 week

    3. 3

      Judge

      Rotty's classifier.

      Founding engineer 1 week

    4. 4

      Check

      Hand-label a subset.

      Founder 1 week

    5. 5

      Publish

      Charts and data.

      Growth lead 1 week

    6. 6

      Pitch

      Tech press.

      Growth lead 2 weeks

    7. 7

      Repeat

      Monthly.

      Growth lead monthly

    First result
    1 quarter
    Cost
    Research time
    Owner
    Founder

    Smallest test

    Publish one chart from 1,000 replies.

    Why it works

    Each month, sample public threads on Reddit, X and LinkedIn, run Rotty's judge, and publish the share of each kind by site and topic, with the method and data. Pitch it to tech press.

    A number journalists can quote ("41% of replies are filler") makes Doplex the source on the topic and brings people to Rotty.

    Risks and how to handle them

    • Platform terms

      Respect each site's rules for public data.

    • Claims of AI detection

      Measure substance, as Rotty does.

    Evidence

    • Chrome Web StoreRotty's listing: "It never claims to detect whether a person or a machine wrote a post"; five kinds of junk, each with its own switch; the toolbar popup shows "how many were hidden this week"; models are your ChatGPT plan, a free Groq key, Rotty Local in the browser, or your own OpenAI, Anthropic or Google key with a 50-cent daily cap; "No servers", "no analytics or tracking in the extension".Source
    • Your websiteBelay white paper (version 1.1, September 15, 2026): agents met acceptance criteria on 11 of 16 tasks with a Belay lesson, 7 of 16 with a human-written instruction and 6 of 16 with no guidance; result CSVs are published; Belay exposes a "bounded stdio MCP server"; next studies are agent-specific delivery, explicit scope, focused verification and "Months of real repos, not one benchmark"; "full methodology and agent-level results will follow in a research preprint."Source
    • Your websiteLab homepage (checked November 4, 2026): "Doplex Labs is an applied AI lab"; products Belay ("Alpha", "Debriefs your coding agent's sessions"), Rotty ("Live", "A guard dog for your feed") and an "Unannounced product" with a "Get early access" button; channels X, TikTok, Instagram and YouTube. The Rotty button links straight to the Chrome Web Store with no campaign tags, and early access and contact both open one chat box that sends a free-text message.Source
  • 18

    The white paper's next study is "Months of real repos, not one benchmark". Belay already finds loops, stalls and "said done while the typecheck still failed".

    High impactHigh effort First result in 2 quarters Founder

    First action

    Ask design partners to opt in.

    How to measure it

    Downloads

    Now
    –
    Target
    5,000
    By
    2 quarters

    Steps

    1. 1

      Consent

      Opt-in counts only.

      Founder 1 month

    2. 2

      Collect

      Failure types and fixes.

      Founding engineer 3 months

    3. 3

      Analyse

      By agent and kind.

      Founder 1 month

    4. 4

      Write

      Report and charts.

      Founder 1 month

    5. 5

      Launch

      Thread, HN, newsletters.

      Growth lead 1 week

    6. 6

      Repeat

      Every six months.

      Founder 6 months

    First result
    2 quarters
    Cost
    Research time
    Owner
    Founder

    Smallest test

    Publish the top three failures from your own sessions.

    Why it works

    With opt-in teams, publish a twice-yearly report on the most common repeated agent failures and the rules that fixed them, by agent. Share counts only, never code.

    Every developer using agents wants this list. It positions Belay as the expert on agent mistakes.

    Risks and how to handle them

    • Privacy fears

      Counts only, local by design.

    Evidence

    • Your websiteBelay white paper (version 1.1, September 15, 2026): agents met acceptance criteria on 11 of 16 tasks with a Belay lesson, 7 of 16 with a human-written instruction and 6 of 16 with no guidance; result CSVs are published; Belay exposes a "bounded stdio MCP server"; next studies are agent-specific delivery, explicit scope, focused verification and "Months of real repos, not one benchmark"; "full methodology and agent-level results will follow in a research preprint."Source
    • Your websitegetbelay.vercel.app: "Belay debriefs your Claude Code and Codex sessions, then carries the lesson into the next one"; a one-line install for macOS and Windows, "Free. Local. 30 sec install."; lists Claude Code, Codex, Cursor and Antigravity; source under the MIT License on GitHub; "Belay for Teams · Enterprise Coming soon" with a waitlist; a "Talk to the founders" form. It has full share tags and a canonical URL, but no robots.txt or sitemap.Source
  • 19

    Every Doplex product is free today: Belay is "Free. Local.", Rotty is free. Teams is the stated paid path.

    High impactHigh effort First result in 2 quarters Founder

    First action

    Draft three pricing options.

    How to measure it

    Paying teams

    Now
    –
    Target
    10
    By
    2 quarters

    Steps

    1. 1

      Options

      Seat price and tiers.

      Founder 1 week

    2. 2

      Test

      With design partners.

      Founder 1 month

    3. 3

      Security page

      For buyers.

      Founding engineer 2 weeks

    4. 4

      Pricing page

      On the site.

      Growth lead 1 week

    5. 5

      Trial

      Self-serve start.

      Founding engineer 1 month

    6. 6

      Measure

      Trials to paid.

      Growth lead monthly

    First result
    2 quarters
    Cost
    Product time
    Owner
    Founder

    Smallest test

    Ask three waitlist teams what they would pay.

    Why it works

    After the design partners, publish a Teams pricing page (per seat, with a free solo plan), a security page for buyers, and a trial. Keep solo Belay free and open source.

    A price makes the company sustainable and turns the waitlist into revenue.

    Risks and how to handle them

    • Open source is enough for teams

      Charge for fleet-wide features only.

    Builds on

    Evidence

    • Your websitegetbelay.vercel.app: "Belay debriefs your Claude Code and Codex sessions, then carries the lesson into the next one"; a one-line install for macOS and Windows, "Free. Local. 30 sec install."; lists Claude Code, Codex, Cursor and Antigravity; source under the MIT License on GitHub; "Belay for Teams · Enterprise Coming soon" with a waitlist; a "Talk to the founders" form. It has full share tags and a canonical URL, but no robots.txt or sitemap.Source
    • GitHubGitHub organisation DoplexLabs: belay (1 star), belay-engine-os (0 stars, 6 forks), rotty-open-source (2 stars) and numbat (0 stars), all created in September and October 2026.Source

Wildcards 3

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

  • 20

    Two Shorts asked viewers to spot the human among bot comments. People love to test themselves, and a game is easy to share.

    Medium impactLow effort First result in 1 month Founding engineer

    First action

    Pick 50 sample comments.

    How to measure it

    Installs from the game

    Now
    –
    Target
    200
    By
    1 quarter

    Steps

    1. 1

      Content

      50 comments, labelled.

      Video creator 1 week

    2. 2

      Build

      One page, scores, card.

      Founding engineer 1 week

    3. 3

      Tag links

      utm_source=game.

      Growth lead 1 hour

    4. 4

      Launch

      With a Short.

      Video creator 1 day

    5. 5

      Refresh

      New rounds monthly.

      Video creator monthly

    6. 6

      Measure

      Plays to installs.

      Growth lead weekly

    First result
    1 month
    Cost
    A week of build
    Owner
    Founding engineer

    Smallest test

    Post the game link in one Short.

    Why it works

    Build a one-page game on the Rotty site: ten rounds of comments, pick the human, get a score and a share card, and see Rotty hide the bots at the end.

    Games get shared far more than product pages, and this one shows exactly what Rotty does.

    Risks and how to handle them

    • Seen as AI detection

      Frame it as fun, not a claim.

    Evidence

    • YouTubeYouTube @DoplexLabs: 6 subscribers, 13 videos, about 3,500 views. Top videos are jokes: "My dad has been arguing with a bot on Facebook since 9am" (1,002 views), "The bots' group chat right now" (877), "the bots will flirt with anything." (464); the two "Introducing Rotty" videos have 125 and 62. Descriptions say "Download link in the pinned comment, or search 'Rotty'", "comment 'woof' and i'll reply with the link" or just "Released on Chrome Store"; one promises "Facebook is next." Latest upload October 8, 2026.Source
    • Your websitegetrotty.vercel.app: a full-URL share image, a sitemap of 4 pages, and every "Add to Chrome" button tagged utm_source=getrotty with a campaign per placement (nav, hero, faq, closing, footer); a feedback page; contact is a gmail address; Product Hunt and Spacerr badges; "+ more sites soon". It links to Belay in the footer.Source
  • 21

    The channel's best jokes are about bots in comment sections. Many creators complain about the same thing on their own channels.

    Medium impactLow effort First result in 1 month Growth lead

    First action

    List 30 creators.

    How to measure it

    Creators who mention Rotty

    Now
    –
    Target
    5
    By
    1 quarter

    Steps

    1. 1

      Find creators

      Videos about bot comments.

      Growth lead 1 week

    2. 2

      Record clips

      Rotty on their comments.

      Video creator 1 week

    3. 3

      Send notes

      Short, personal.

      Growth lead 1 week

    4. 4

      Tag links

      One per creator.

      Growth lead 1 hour

    5. 5

      Follow up

      Once.

      Growth lead 2 weeks

    6. 6

      Measure

      Installs by creator.

      Growth lead monthly

    First result
    1 month
    Cost
    Time only
    Owner
    Growth lead

    Smallest test

    Send five notes this week.

    Why it works

    Find 30 creators who have posted about bot comments. Send each a short note and a clip of Rotty cleaning their own comments. No payment, just a useful tool and a tagged link if they want to share it.

    Creators reach exactly the people who hate bot comments, and a tool that helps them is worth talking about.

    Risks and how to handle them

    • No replies

      Pick creators who already complained.

    Evidence

    • YouTubeYouTube @DoplexLabs: 6 subscribers, 13 videos, about 3,500 views. Top videos are jokes: "My dad has been arguing with a bot on Facebook since 9am" (1,002 views), "The bots' group chat right now" (877), "the bots will flirt with anything." (464); the two "Introducing Rotty" videos have 125 and 62. Descriptions say "Download link in the pinned comment, or search 'Rotty'", "comment 'woof' and i'll reply with the link" or just "Released on Chrome Store"; one promises "Facebook is next." Latest upload October 8, 2026.Source
    • Chrome Web StoreRotty's listing: "It never claims to detect whether a person or a machine wrote a post"; five kinds of junk, each with its own switch; the toolbar popup shows "how many were hidden this week"; models are your ChatGPT plan, a free Groq key, Rotty Local in the browser, or your own OpenAI, Anthropic or Google key with a 50-cent daily cap; "No servers", "no analytics or tracking in the extension".Source
  • 22

    Rotty names each kind it hides: generic filler, engagement bait, drive-by plugs, recycled formula posts, scams. The site has 4 pages, so none of this can be found in search.

    Medium impactLow effort First result in 1 month Growth lead

    First action

    Write the engagement bait page first.

    How to measure it

    Monthly visits to guide pages

    Now
    –
    Target
    2,000
    By
    2 quarters

    Steps

    1. 1

      List kinds

      From the extension.

      Growth lead 1 day

    2. 2

      Write pages

      One per kind.

      Growth lead 2 weeks

    3. 3

      Examples

      Made up, no names.

      Video creator 1 week

    4. 4

      Link from extension

      "Why was this hidden?"

      Founding engineer 2 days

    5. 5

      Sitemap

      Add pages.

      Founding engineer 1 hour

    6. 6

      Measure

      Visits and installs.

      Growth lead monthly

    First result
    1 month
    Cost
    Writing time
    Owner
    Growth lead

    Smallest test

    Publish one page this week.

    Why it works

    Write one page per kind, with real-looking examples (no names), why it spreads, and how Rotty handles it. Link them from the hidden-post line ("Why was this hidden?").

    People search for these patterns ("engagement bait examples", "LinkedIn broetry"). Each page is a way in, and the link from the extension explains the product too.

    Risks and how to handle them

    • Seen as mocking users

      Mock the pattern, never people.

    Evidence

    • Chrome Web StoreRotty's listing: "It never claims to detect whether a person or a machine wrote a post"; five kinds of junk, each with its own switch; the toolbar popup shows "how many were hidden this week"; models are your ChatGPT plan, a free Groq key, Rotty Local in the browser, or your own OpenAI, Anthropic or Google key with a 50-cent daily cap; "No servers", "no analytics or tracking in the extension".Source
    • Your websitegetrotty.vercel.app: a full-URL share image, a sitemap of 4 pages, and every "Add to Chrome" button tagged utm_source=getrotty with a campaign per placement (nav, hero, faq, closing, footer); a feedback page; contact is a gmail address; Product Hunt and Spacerr badges; "+ more sites soon". It links to Belay in the footer.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.

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.

  • Cymulate growth strategyAn exposure validation platform that continuously tests an organisation's security controls against real attack techniques (breach and…21 growth ideas
  • Runlayer growth strategyAn AI control plane and enablement platform: a governed MCP gateway and catalog of 18,000+ MCPs, shadow AI discovery, agent identity and…22 growth ideas
  • Frontal growth strategyA go-to-market agency for B2B tech companies. One team runs GTM engineering and outbound, account-based marketing, paid media, content…22 growth ideas
  • Hightouch growth strategyAn Agentic Marketing Platform built on a Composable CDP: Ad Studio, Lifecycle Studio, AI Decisioning and Content Assembly on top of…22 growth ideas
  • Leeroo growth strategyUniforge, an enterprise AI platform that connects to a company's systems (CRM, ERP, billing, ITSM, support, HR), learns them into one…23 growth ideas
  • Whalar growth strategyA creator and social agency: creator-led strategy, campaign management, paid amplification, measurement and creator events for global…21 growth ideas
  • Quintagroup growth strategyA Python and JavaScript software development company from Lviv, Ukraine, known for e-procurement and auction systems (the OpenProcurement…23 growth ideas
  • Ghar Soaps growth strategyAn Ayurveda-inspired skincare brand from Pune, best known for its Sandalwood & Saffron Magic Soap for tan. It also sells face washes…23 growth ideas

See every playbook and website review

Independent research. Programmable wrote this playbook from public sources, as they stood on November 4, 2026, and cites them throughout. Doplex Labs did not commission, review or endorse it, and Programmable is not affiliated with Doplex Labs. 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 Doplex Labs runs its business.

Something wrong, or do you represent Doplex Labs? 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

    YouTube to install path, share card and sitemap, tagged links, rating ask, early-access list and font fix.

  2. Days 31 to 6010 ideas

    Share card, compare page, Belay videos, directories, GitHub, preprint, design partners, Facebook launch, one domain and restarted Shorts.

  3. Days 61 to 906 ideas

    The Slop Index, the agent failure report and a paid Teams plan, plus the game, creators and field guide.

The research behind the plan

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

Customers

What customers say

Customers value A cleaner feed without blocking ads or opinions. The most cited problem: bot replies and filler in comment sections.

  • Seeing why each post was hidden
  • Coding agents repeating the same mistakes
  • In their words: “AI slop”

Market

Competitive landscape

3 competitors compared: AI Slop Blocker, FeedShield and SpecStory.

  • Gap to exploit: A filler filter that judges substance, not authorship
  • AI Slop Blocker: Claims to detect AI-generated posts
  • FeedShield: Small user base

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