ProgrammableMarketing/voyager.so (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 · 12
  • Big bets · 3
  • Wildcards · 2
  • How we can help
Generated Nov 10, 2026 6 views
  1. Programmable
  2. Playbooks
  3. Voyager
Voyager logo

growth playbook · Nov 10, 2026

voyager.so (opens in a new tab)

Voyager growth playbook

Diagnosis

Voyager launched on October 8, 2026 and already has a lot of search content: 79 tested model pages, 10 app pages, 8 comparison pages and 161 blog articles, plus a vork repo with 1.3k GitHub stars. Most of that traffic ends in the same place, a desktop download, and the homepage's hero button names only macOS.

Meanwhile a browser generator already exists at voyager.so/dashboard, and Voyager's own Higgsfield page names browser access as the rival's edge. The quickest gains are in the funnel: a browser trial on every model page, a lighter homepage that passes Core Web Vitals, every platform shown, claims that match the app pages, and full prices on the pricing page.

After that, the plan builds the loops a credits business needs: shared skills, a showcase, referral credits, cross-sell from Moda, and turning vork developers into Vork Cloud customers.

Growth Score

7/10

Biggest Opportunity

Turn high-intent model and price searches into a first generation in the browser, then into editable projects in the desktop app.

Biggest Risk

Browser-first rivals with far larger audiences (Higgsfield claims 40M+ creators, Krea over 30 million users) can win visitors who won't install a desktop app, before Voyager's editable-project advantage gets a chance.

01

Start here

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

  1. 1

    Add "Try it in your browser" to all 79 model pages, which now send searchers only to a desktop download

    Add a link format to the dashboard generator that pre-selects a model and a prompt, such as /dashboard?model=veo-3-1&prompt=<test prompt>.

    High impact First result in 3 weeks
  2. 2

    Make the homepage launch film load on click, so the 17.5 MB homepage passes Core Web Vitals

    Swap the launch film for a poster image plus a click-to-play loader.

    High impact First result in 1 week
  3. 3

    Show Windows and Linux in the homepage hero, not only "Download for macOS"

    Add a line under the hero button listing all three platforms.

    Medium impact First result in 1 week

02

Your idea portfolio

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

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

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

Effort vs impactEach number is an idea. Select one to open it.
Impact →
Quick wins · 2Big bets · 9Fill-ins · 10Reconsider · 2
Effort →

Showing 23 of 23 ideas

This week 6

Quick wins with a clear first action.

  • 01

    Voyager's tested model pages (for example "Veo 3.1: Pricing, Free Access and Real Clips (2026)" and "Kling 4.0: Pricing, 30-Second Clips and Real Tests (2026)") are written for people searching a model's price and access. Their only buttons are "Get Voyager", which goes to the download page, and "Sign up". Yet Voyager's own docs say anyone can make an image, clip or audio at voyager.so/dashboard "without installing anything", and Voyager's Higgsfield comparison lists "Browser access with nothing to install" as the rival's edge. The fastest route to a first result is hidden from the visitors most ready to generate.

    High impactMedium effort First result in 3 weeks Head of Growth

    First action

    Add a link format to the dashboard generator that pre-selects a model and a prompt, such as /dashboard?model=veo-3-1&prompt=<test prompt>.

    How to measure it

    Model page to first generation

    Now
    –
    Target
    +50% sign-ups from model pages, and 40% of those generate on day one
    By
    60 days

    Steps

    1. 1

      Baseline

      30 days of model-page visits, sign-ups, downloads and first generations in product analytics.

      Head of Growth 2 days

    2. 2

      Build the deep link

      The dashboard reads model and prompt from the URL and keeps them through sign-up.

      Frontend Engineer 3 days

    3. 3

      Add the button

      "Try <model> in your browser" as the main button on all 79 model pages, using each page's test prompt.

      Frontend Engineer 2 days

    4. 4

      Handle plan limits

      Free models run straight away; video models show the clip's credit cost and the Starter upgrade.

      Product Lead 2 days

    5. 5

      Offer the desktop app next

      After the first result: "Open this in Voyager to edit, cut and score it."

      Product Lead 1 week

    6. 6

      A/B test

      Half the model pages get the button for four weeks; compare sign-ups, first generations and paid starts.

      Head of Growth 4 weeks

    7. 7

      Roll out

      Extend to /ai-video, the comparison pages and the model posts on the blog.

      Content Lead 1 week

    First result
    3 weeks
    Cost
    About two weeks of engineering, plus free-tier generation credits
    Owner
    Head of Growth

    Smallest test

    Add the button to the 10 most-visited model pages for two weeks.

    Why it works

    Give every /resources/models page a main button, "Try <model> in your browser", that opens the dashboard generator with that model and the page's own test prompt already chosen (sign-up first for new visitors). Free image and voice models run straight away. Video models show the credit cost of the test clip and the Starter plan that unlocks it. Keep "Download Voyager" as the second button, and after the first generation offer the desktop app to turn the clip into an editable project. Then add the same button to /ai-video, the comparison pages and the model posts on the blog.

    The visitor has already picked a model. One click to run the exact prompt they just saw turns search intent into an account and a first result, with no install. That matters most for anyone on a phone, an Intel Mac or a work machine. It also closes the gap Voyager's own Higgsfield page admits, and the desktop app becomes the next step rather than the first.

    Ready to use

    ChecklistModel page browser-trial checklist

    [ ] Dashboard accepts ?model= and &prompt=
    [ ] Model and prompt survive sign-up
    [ ] Button text: Try <model> in your browser
    [ ] Free models run at once; video shows credit cost + Starter
    [ ] Download Voyager kept as the second button
    [ ] Post-generation: Open in Voyager to edit
    [ ] Event tracking: click, sign-up, first generation, paid start
    [ ] A/B split across model pages

    Risks and how to handle them

    • Free-tier abuse from anonymous traffic

      Keep the existing verified-email rule and the 5-hour credit window.

    • Browser users never install the desktop app

      Track desktop installs within 14 days, and make the editable-project hand-off the first thing shown after a generation.

    Evidence

    • Your websiteModels hub "AI image, video and voice models, tested": 79 tested model cards (38 image, 31 video, 10 voice) with real test outputs and prices; its only call to action is "Get Voyager" to the download page. Model pages such as "Veo 3.1: Pricing, Free Access and Real Clips (2026)" and "Kling 4.0: Pricing, 30-Second Clips and Real Tests (2026)" offer "Get Voyager" (to /download) and "Sign up", with no link to try the model in the browser.Source
    • Your websiteHelp docs, "Generate in your browser": "You can make a video clip, an image or a piece of audio on voyager.so/dashboard without installing anything." "Voyager Free includes some image and voice models; video needs a paid plan."Source
    • Your websiteComparison page "Voyager vs Higgsfield" (last verified 8 October 2026): lists Higgsfield's strengths as "Browser access with nothing to install" and preset-driven direction; shows Voyager cheaper per image on every matched plan for one model (4.2¢ against 10.6¢ at the $19 tier); says Voyager can "sign in with Higgsfield to spend Higgsfield credits"; the main call to action is "Download Voyager".Source
  • 02

    Real visitors fail Google's Core Web Vitals on the homepage (LCP 3.7 s and INP 275 ms at the 75th percentile). The page downloads about 17.5 MB, roughly 16 MB of it media from cdn.moda.app, and the largest element is the "Play the Voyager launch film" button. On a phone, the main content takes 12.1 s to appear in Google's lab test.

    High impactLow effort First result in 1 week Frontend Engineer

    First action

    Swap the launch film for a poster image plus a click-to-play loader.

    How to measure it

    Field LCP p75

    Now
    –
    Target
    Under 2.5 s, with page weight under 3 MB
    By
    45 days

    Steps

    1. 1

      Poster frame

      A WebP poster at hero size in place of the film.

      Designer 2 hours

    2. 2

      Click to load

      Fetch the video only on play.

      Frontend Engineer 1 day

    3. 3

      Lazy-load examples

      Media for the six workflow guides loads as it scrolls into view.

      Frontend Engineer 1 day

    4. 4

      Cache headers

      Long lifetimes on hashed static files (about 0.9 s saved).

      Frontend Engineer 2 hours

    5. 5

      Trim JavaScript

      Split off the roughly 247 KB of unused JavaScript (about 1.2 s).

      Frontend Engineer 3 days

    6. 6

      Measure

      PageSpeed after each change; field data after 28 days.

      Head of Growth 4 weeks

    First result
    1 week (lab), 4 weeks (field)
    Cost
    A few days of engineering
    Owner
    Frontend Engineer

    Smallest test

    Ship the poster-and-click change alone and re-run PageSpeed.

    Why it works

    Show a compressed poster frame for the launch film and fetch the video only when someone presses play. Lazy-load the workflow example media below the fold, set long cache lifetimes on hashed assets, and split unused JavaScript off the first load. Re-test in PageSpeed and watch the field data over 28 days.

    Most of the weight is media nobody has asked for yet. Moving it behind a click is the biggest single speed gain, and a fast homepage helps both search ranking and paid traffic.

    Risks and how to handle them

    • Fewer people watch the film

      Use a strong poster frame and track play rate.

    Evidence

    • Google PageSpeedHomepage, tested November 9, 2026: mobile performance 63/100 (desktop 77); real visitors fail Core Web Vitals on LCP (p75 3.7 s) and INP (p75 275 ms); in the lab, the main content appears after 12.1 s on mobile; the page downloads about 17.5 MB in 98 requests, about 16 MB of it media from cdn.moda.app (the CDN of Nullframe's sister product Moda); the largest element is the "Play the Voyager launch film" button; Google estimates unused JavaScript costs about 1.2 s and short cache lifetimes about 0.9 s.Source
  • 03

    The homepage hero button reads "Download for macOS ↓", and the homepage text never mentions Windows or Linux. Yet the download page offers all three, and the FAQ supports Windows 11 and Linux x64. A Windows editor reading the hero can reasonably assume Voyager isn't for them.

    Medium impactLow effort First result in 1 week Frontend Engineer

    First action

    Add a line under the hero button listing all three platforms.

    How to measure it

    Windows and Linux downloads

    Now
    –
    Target
    +40%
    By
    30 days

    Steps

    1. 1

      Add the platform line

      "macOS · Windows · Linux" under the button.

      Frontend Engineer 1 hour

    2. 2

      Detect the OS

      Pick the button label from the user agent.

      Frontend Engineer 1 day

    3. 3

      Handle the edge cases

      Intel Mac and phone visitors see the browser option.

      Frontend Engineer 1 day

    4. 4

      Label previews

      Mark Windows and Linux features that are still in alpha.

      Product Lead 1 hour

    5. 5

      Copy to other pages

      /ai-video and the comparison pages.

      Frontend Engineer 1 day

    6. 6

      Measure

      Downloads by OS, week on week.

      Head of Growth weekly

    First result
    1 week
    Cost
    $0, engineering time only
    Owner
    Frontend Engineer

    Smallest test

    Add the platform line today.

    Why it works

    Detect the visitor's operating system and show the matching button ("Download for Windows", "Download for Linux"), with "Also on macOS, Windows and Linux" underneath. For Intel Macs and phones, show the browser generator and a reminder option instead (ideas 1 and 23). Do the same on /ai-video and the comparison pages.

    Visitors look for their own platform in the first button. Naming it removes a silent reason to leave.

    Risks and how to handle them

    • Windows users hit alpha limits

      Say what works today on Windows next to the button.

    Evidence

    • Your websiteHomepage: title "Voyager: the open harness for creative work", headline "The faster, cheaper, better harness for creative work."; claims "100+ Creative tools"; the hero button reads "Download for macOS ↓" (to /download/mac), and the homepage text does not mention Windows or Linux; "Free · 1,000 credits a month · or bring your own Claude or ChatGPT"; social proof is six testimonials quoted from X posts, with no user counts, ratings or customer stories.Source
    • Your websiteDownload page offers macOS, Windows and Linux ("Choose your computer. We'll help you get started.") with no email capture. The FAQ lists Apple silicon Macs on macOS 15+ ("Intel isn't supported yet"), Windows 11 (x64) and Linux (x64), and says "Mobile apps are coming soon."Source
  • 04

    The homepage claims "100+ Creative tools". The apps hub lists 10 apps ("facts checked 23 October 2026"), and the Voyager vs Claude page says skills cover "9+ creative apps". The launch review on explainx.ai pointed out the same gap. For a brand built on taste and honest pricing, a number reviewers can't check costs trust.

    Medium impactLow effort First result in 2 weeks Content Lead

    First action

    Count the apps Voyager can actually drive today, by support level.

    How to measure it

    Pages with a matching count

    Now
    –
    Target
    100%
    By
    14 days

    Steps

    1. 1

      Count

      Native scripting, CLI, computer control and MCP, counted separately.

      Product Lead 1 day

    2. 2

      Decide

      Publish the full list or change the claim.

      Founder & CEO 1 day

    3. 3

      Build the list

      Support level per app, linked to app pages where they exist.

      Content Lead 3 days

    4. 4

      Align copy

      Homepage, comparison pages and FAQ.

      Content Lead 1 day

    5. 5

      Add pages

      Next app pages in order of search demand (Photoshop, Ableton Live, Final Cut Pro).

      Content Lead ongoing

    6. 6

      Review

      Recheck the count each month.

      Content Lead monthly

    First result
    2 weeks
    Cost
    Staff time only
    Owner
    Content Lead

    Smallest test

    Change the homepage stat this week.

    Why it works

    Either publish a /resources/apps/all list of every app and service the agent can use (native scripting, computer control, MCP), grouped by support level, or change the homepage to a number that matches the apps hub, such as "Works in 10 pro apps, plus any app with computer control". Use the same number on the homepage, the comparison pages and the FAQ.

    Claims that match across pages turn a reviewer's caveat into a proof point, and the full list also earns long-tail "<app> AI agent" searches.

    Risks and how to handle them

    • A smaller number looks weaker

      Show depth: what the agent can do in each app.

    Evidence

    • Your websiteApps hub: "10 apps · facts checked 23 October 2026" (After Effects, Blender, CapCut, DaVinci Resolve, Godot, Houdini, Premiere Pro, Roblox Studio, Unity, Unreal Engine), while the homepage says "100+ Creative tools" and the Voyager vs Claude page says skills cover "9+ creative apps". The Blender page lists "Try asking" prompts but has no video demo, gallery or downloadable project.Source
    • Pressexplainx.ai review (October 9, 2026): Voyager launched October 8, 2026 from Nullframe, the company behind the design tool Moda; it reports a founder claim of a Moda user base of "100k+ people" that it could not verify; it flags the "100+" tools claim against the roughly two dozen apps named in the FAQ and "9+" on a comparison page, notes the app is not open source despite the "open harness" name, and found no Hacker News discussion of the launch.Source
  • 05

    The pricing page says annual plans are "15% off" but shows only Creator's annual price ($41.65 a month). Team plans (2 to 25 people, shared credits, admin roles) and Enterprise have no prices at all. A small studio comparing Voyager with rivals that publish their prices has to email support to find out.

    Medium impactLow effort First result in 2 weeks Founder & CEO

    First action

    Fill in the annual prices for Starter, Ultra and GigaMax.

    How to measure it

    Annual share of new paid plans

    Now
    –
    Target
    +10 points
    By
    60 days

    Steps

    1. 1

      Fill annual prices

      All four paid tiers.

      Frontend Engineer 1 hour

    2. 2

      Set Team pricing

      Per seat, plus the 10% annual discount.

      Founder & CEO 1 day

    3. 3

      Seat calculator

      Seats, plan, credits and total.

      Frontend Engineer 2 days

    4. 4

      Enterprise line

      "From" price, SSO and spend controls.

      Founder & CEO 1 hour

    5. 5

      Self-serve Team checkout

      Team plans bought without email.

      Product Lead 1 week

    6. 6

      Measure

      Annual mix and Team starts.

      Head of Growth monthly

    First result
    2 weeks
    Cost
    $0
    Owner
    Founder & CEO

    Smallest test

    Show all annual prices this week.

    Why it works

    Show the annual price for every tier on the toggle, publish a per-seat Team price with its 10% annual discount, and add "Enterprise from" with the SSO and spend-control features named. Add a short calculator: seats × plan = monthly credits and cost.

    Studios buy when they can budget. Prices on the page remove an email round-trip, and annual prices that are easy to see pull more people onto annual plans.

    Risks and how to handle them

    • Committing to Team prices too early

      Label them launch prices and review after a quarter.

    Evidence

    • Your websitePricing page: Free $0 with 1,000 credits a month (up to 500 per 5 hours); Starter $19 (1,800 credits, 2,700 for the first 3 months); Creator $49, "Most popular"; Ultra $149; GigaMax $500; annual plans "15% off", but only Creator's annual price ($41.65 a month) is shown; Team plans for 2 to 25 people (annual 10% off) and Enterprise for 26+ have no listed prices; credit packs "from $11" at 1.1¢ a credit; billing is prepaid; Free pauses when credits run out.Source
    • Your websiteFAQ: made by Nullframe, Inc.; for "Creators, editors, motion designers, marketers, and small studios"; users can bring their own chat subscription ("Voyager never bills those subscriptions") or API keys; generation is billed at provider cost, 1 credit = 1¢, with no markup; Team plans cover 2 to 25 people, and Enterprise adds SSO and admin spend controls; payments are generally non-refundable; Voyager ships about 40 skills.Source
  • 06

    Voyager publishes several articles a day (161 so far), and every one ends at "Get Voyager". There is no newsletter signup on the blog and no email capture on the download page. Readers who aren't ready to install, or are on the wrong computer, leave with no way back.

    Medium impactLow effort First result in 2 weeks Lifecycle Marketer

    First action

    Add a one-field signup to the end of blog posts.

    How to measure it

    Newsletter subscribers

    Now
    –
    Target
    2,000
    By
    90 days

    Steps

    1. 1

      Pick a tool

      One that works with the existing account system.

      Lifecycle Marketer 1 day

    2. 2

      Add forms

      End of blog posts, model pages and the download page.

      Frontend Engineer 2 days

    3. 3

      Phone option

      "Email me the download link".

      Frontend Engineer 1 day

    4. 4

      First issue

      Price moves, one guide, one community result.

      Content Lead 1 day

    5. 5

      Send weekly

      Same weekday every week.

      Lifecycle Marketer weekly

    6. 6

      Measure

      Signups, opens and downloads from email.

      Head of Growth monthly

    First result
    2 weeks
    Cost
    Email tool, under $100 a month at first
    Owner
    Lifecycle Marketer

    Smallest test

    A signup box on the 10 most-read posts for two weeks.

    Why it works

    Add a "Weekly model prices and new workflows" signup to the blog, the model pages and the download page. Send a short weekly email: price changes from the tested-models catalogue, one new workflow guide, and one result from the community. Add an "Email me the download link" option for visitors on a phone.

    Model prices change every week, which gives creators a reason to subscribe that a generic newsletter lacks. Email brings back the visitors a desktop download loses.

    Risks and how to handle them

    • Low-quality signups

      Double opt-in, and track what share go on to download.

    Evidence

    • Your websiteBlog: 161 articles (Techniques 61, Agents 34, Generators 31, Editors 27, Models 4, Voyager 4), several published a day in November 2026, including Roblox Studio game guides and statistics roundups; every post ends at "Get Voyager", and there is no newsletter signup.Source
    • Your websiteDownload page offers macOS, Windows and Linux ("Choose your computer. We'll help you get started.") with no email capture. The FAQ lists Apple silicon Macs on macOS 15+ ("Intel isn't supported yet"), Windows 11 (x64) and Linux (x64), and says "Mobile apps are coming soon."Source

This quarter 12

The core plays for the next three months.

  • 07

    Voyager skills are folders with a SKILL.md file. Users can create them, import them, and get "Learned" skills saved from finished work, and the app ships about 40. But the docs describe no way to share a skill with anyone else, and there's no public library. Each user's best workflows stay locked on their own machine.

    High impactMedium effort First result in 6 weeks Product Lead

    First action

    Publish the six homepage workflow guides as installable skills.

    How to measure it

    Library installs

    Now
    –
    Target
    5,000
    By
    90 days

    Steps

    1. 1

      Define the format

      SKILL.md plus an example output and app requirements.

      Product Lead 3 days

    2. 2

      One-click install

      A deep link opens the skill in Voyager.

      Frontend Engineer 1 week

    3. 3

      Seed

      Bundled skills and homepage workflows.

      Content Lead 1 week

    4. 4

      Publish flow

      Publish from the Learned and Added tabs.

      Product Lead 2 weeks

    5. 5

      Review queue

      Text-only checks, matching the import rule that scripts don't run.

      Product Lead ongoing

    6. 6

      Invite creators

      The first 20 publishers from Discord.

      Community Manager 2 weeks

    7. 7

      Measure

      Installs, and first generations after an install.

      Head of Growth monthly

    First result
    6 weeks
    Cost
    About six weeks of product and engineering time
    Owner
    Product Lead

    Smallest test

    Publish 10 skills as pages with copy-to-import instructions.

    Why it works

    Launch voyager.so/skills: a public, searchable library where users publish a skill (with an example output), and others install it in one click through a deep link. Seed it with the about 40 bundled skills and the six homepage workflows, credit authors on each page, and rank skills by installs.

    Shared skills make every power user a distributor. Each skill page is an indexable "how to make X" page with a working answer, and installing someone else's skill is a fast path to a first good result.

    Risks and how to handle them

    • Low-quality or unsafe skills

      Moderation, text-only skills and visible install counts.

    Evidence

    • Your websiteSkills guide: users create skills or import a folder with a SKILL.md file; skills are grouped as Bundled, Added, Learned ("saved from your finished work") and External; the guide describes no way to share skills with other people and no public library.Source
    • Your websiteFAQ: made by Nullframe, Inc.; for "Creators, editors, motion designers, marketers, and small studios"; users can bring their own chat subscription ("Voyager never bills those subscriptions") or API keys; generation is billed at provider cost, 1 credit = 1¢, with no markup; Team plans cover 2 to 25 people, and Enterprise adds SSO and admin spend controls; payments are generally non-refundable; Voyager ships about 40 skills.Source
  • 08

    Voyager's main argument is that it delivers "an editable project", not just clips. Yet the site has no showcase, gallery or customer stories: the 373-URL sitemap has none, the app pages have no gallery, and the homepage's social proof is six quotes from X.

    High impactMedium effort First result in 1 month Content Lead

    First action

    Turn the six homepage workflow guides into the first showcase entries, with credits spent.

    How to measure it

    Showcase-to-sign-up rate

    Now
    –
    Target
    5%
    By
    90 days

    Steps

    1. 1

      Template

      Brief, final cut, apps, models, tries, credits, project.

      Designer 3 days

    2. 2

      First six

      The homepage workflows.

      Content Lead 1 week

    3. 3

      Call for work

      Discord and X, with credits as thanks.

      Community Manager 2 weeks

    4. 4

      Permissions

      Written consent from each featured creator.

      Community Manager ongoing

    5. 5

      Link from app pages

      A gallery on each /resources/apps page.

      Frontend Engineer 1 week

    6. 6

      Feature weekly

      In the newsletter and on social.

      Content Lead weekly

    First result
    1 month
    Cost
    Credits for featured creators, plus content time
    Owner
    Content Lead

    Smallest test

    Publish the first six entries.

    Why it works

    Publish /showcase: real pieces made with Voyager, each with the brief, the final video, the apps and models used, the number of tries, the credits spent and a downloadable project (or a timeline screenshot). Start with the team's own six homepage workflows and add work from users on Discord who agree to be featured.

    Showing the project file and the cost proves both claims rivals can't easily copy: editable work and no markup. It gives creators something to share and gives each app page a gallery.

    Risks and how to handle them

    • Rights in generated media

      Feature only work whose creator confirms they own it, and keep provider labels.

    Evidence

    • Your websiteSitemap (373 URLs, read November 9, 2026): 187 /resources pages, 168 blog URLs, 8 comparison pages (Higgsfield, Flora, Weavy, OpenArt, Krea, Runway, Claude, Codex), 10 app pages, /ai-video and 2 vork pages; no showcase, gallery, templates, customer-story, use-case or referral pages.Source
    • Your websiteHomepage: title "Voyager: the open harness for creative work", headline "The faster, cheaper, better harness for creative work."; claims "100+ Creative tools"; the hero button reads "Download for macOS ↓" (to /download/mac), and the homepage text does not mention Windows or Linux; "Free · 1,000 credits a month · or bring your own Claude or ChatGPT"; social proof is six testimonials quoted from X posts, with no user counts, ratings or customer stories.Source
    • Your websiteApps hub: "10 apps · facts checked 23 October 2026" (After Effects, Blender, CapCut, DaVinci Resolve, Godot, Houdini, Premiere Pro, Roblox Studio, Unity, Unreal Engine), while the homepage says "100+ Creative tools" and the Voyager vs Claude page says skills cover "9+ creative apps". The Blender page lists "Try asking" prompts but has no video demo, gallery or downloadable project.Source
  • 09

    vork, Voyager's headless 3D engine for agents, has 1.3k stars and 97 forks, far more than the voyager-so organisation's 29 followers or the desktop-releases repo's 22 stars. Developers are paying attention to vork, but rendering on Voyager cloud GPUs needs a paid Starter workspace, so the free path ends at a local install.

    High impactMedium effort First result in 1 month Developer Relations

    First action

    Add a Voyager and Vork Cloud section to the vork README.

    How to measure it

    GitHub-sourced accounts

    Now
    –
    Target
    500
    By
    90 days

    Steps

    1. 1

      README section

      What Vork Cloud is, and a tracked link.

      Developer Relations 1 day

    2. 2

      Org profile

      Link the voyager-so organisation to Voyager and vork.

      Developer Relations 1 hour

    3. 3

      Free allowance

      Render-lite minutes for each new account.

      Product Lead 1 week

    4. 4

      One-command renders

      In every example scene.

      Developer Relations 2 weeks

    5. 5

      Showcase

      Monthly community renders.

      Community Manager monthly

    6. 6

      Convert

      Starter offer when the allowance runs out.

      Lifecycle Marketer ongoing

    7. 7

      Measure

      Stars to accounts to paid.

      Head of Growth monthly

    First result
    1 month
    Cost
    GPU minutes for the free allowance
    Owner
    Developer Relations

    Smallest test

    Add the README section and watch the tracked link for two weeks.

    Why it works

    Add a "Render this on a cloud GPU" section to the vork README and docs, with a one-time free allowance of Render-lite minutes for new accounts. Each example scene ships with a one-command cloud render. Point the GitHub organisation profile at Voyager and the Vork Cloud page, and publish monthly vork showcase renders.

    A first render on a big GPU, at no cost, shows exactly what the paid workspace is for. GitHub is where these agent builders already are.

    Risks and how to handle them

    • GPU abuse

      Verified email, small caps, Render-lite only.

    Evidence

    • GitHubvork ("Headless 3D and VFX engine for agents. Research preview."): 1.3k stars and 97 forks; the voyager-so organisation has 29 followers, and its desktop-releases repo has 22 stars. The vork page calls it "Source-available and free during the preview", with Voyager cloud GPUs billed by the second from a Starter or higher workspace.Source
    • Your websiteVork Cloud page: GPU presets Render (RTX PRO 6000, 96 GB), Render lite (L4, 24 GB) and Compute (H100, 80 GB); "billed by the second", and idle machines pause after 15 minutes; no prices on the page ("Live price on the dashboard"), only a CLI estimate command.Source
  • 10

    Vork Cloud bills by the second for RTX PRO 6000, L4 and H100 machines, but the page lists no rates. It says "Live price on the dashboard" and offers only a CLI estimate command. For a company whose pitch is published, no-markup pricing, the GPU product is the one place a buyer can't see the price.

    Medium impactLow effort First result in 2 weeks Product Lead

    First action

    Add a three-row price table to /vork/cloud.

    How to measure it

    Page-to-lease conversion

    Now
    –
    Target
    +30%
    By
    60 days

    Steps

    1. 1

      Price table

      Three presets, per hour and per second.

      Product Lead 1 day

    2. 2

      Web estimator

      The same logic as the CLI estimate.

      Frontend Engineer 3 days

    3. 3

      Worked example

      One scene, cost and time.

      Developer Relations 2 days

    4. 4

      Idle note

      Explain the 15-minute auto-pause.

      Product Lead 1 hour

    5. 5

      Keep in sync

      Pull prices from the live source.

      Frontend Engineer 2 days

    6. 6

      Measure

      Leases from page visits.

      Head of Growth monthly

    First result
    2 weeks
    Cost
    $0
    Owner
    Product Lead

    Smallest test

    Add the static price table this week.

    Why it works

    Show each preset's price per hour and per second in credits and dollars, with a web estimator: frames × resolution = estimated time and cost, the same calculation the CLI does. Compare one standard scene against the cost of a local GPU.

    Visible prices match the no-markup brand and let developers budget before they sign up.

    Risks and how to handle them

    • GPU prices move

      Pull them live, with a date stamp.

    Builds on

    Evidence

    • Your websiteVork Cloud page: GPU presets Render (RTX PRO 6000, 96 GB), Render lite (L4, 24 GB) and Compute (H100, 80 GB); "billed by the second", and idle machines pause after 15 minutes; no prices on the page ("Live price on the dashboard"), only a CLI estimate command.Source
    • Your websiteFAQ: made by Nullframe, Inc.; for "Creators, editors, motion designers, marketers, and small studios"; users can bring their own chat subscription ("Voyager never bills those subscriptions") or API keys; generation is billed at provider cost, 1 credit = 1¢, with no markup; Team plans cover 2 to 25 people, and Enterprise adds SSO and admin spend controls; payments are generally non-refundable; Voyager ships about 40 skills.Source
  • 11

    The app pages are detailed (the Blender page lists "Try asking" prompts and 15 agent jobs), but they have no video demo, gallery or downloadable project. The thing people most want to see, the agent working inside their own app, is described rather than shown.

    Medium impactMedium effort First result in 1 month Content Lead

    First action

    Record the Blender turntable prompt from the Blender page.

    How to measure it

    App-page sign-up rate

    Now
    –
    Target
    +25%
    By
    90 days

    Steps

    1. 1

      Pick prompts

      One "Try asking" prompt per app.

      Content Lead 1 day

    2. 2

      Record

      Uncut, 30 seconds.

      Content Lead 1 week

    3. 3

      Embed light

      Poster image, click to play (see idea 2).

      Frontend Engineer 2 days

    4. 4

      Starter files

      A downloadable project per app.

      Content Lead 2 weeks

    5. 5

      Distribute

      X, Discord and the newsletter.

      Community Manager weekly

    6. 6

      Measure

      Plays and sign-ups per page.

      Head of Growth monthly

    First result
    1 month
    Cost
    Staff time only
    Owner
    Content Lead

    Smallest test

    One clip on the Blender page.

    Why it works

    Record one uncut 30-second clip per app: a "Try asking" prompt going in, the agent working in Blender, Resolve or After Effects, and the finished file coming out. Put it at the top of each app page, post it as a short on X and in the newsletter, and add the starter project as a download.

    Uncut recordings answer the main doubt ("does it really drive my app?") and work as both page content and social posts.

    Risks and how to handle them

    • Clips go stale as the app changes

      Re-record every quarter.

    Builds on

    Evidence

    • Your websiteApps hub: "10 apps · facts checked 23 October 2026" (After Effects, Blender, CapCut, DaVinci Resolve, Godot, Houdini, Premiere Pro, Roblox Studio, Unity, Unreal Engine), while the homepage says "100+ Creative tools" and the Voyager vs Claude page says skills cover "9+ creative apps". The Blender page lists "Try asking" prompts but has no video demo, gallery or downloadable project.Source
  • 12

    New users choose among several setups: Voyager's own agent and models, their own chat subscription, API keys, or the browser dashboard. The docs walk them through "Make your first video" and "Find your finished file". The setup choice probably decides who succeeds, and without measuring each path it's guesswork.

    High impactMedium effort First result in 1 month Product Lead

    First action

    Log setup-path choice and first export as events.

    How to measure it

    24-hour activation

    Now
    –
    Target
    +15 points
    By
    1 quarter

    Steps

    1. 1

      Events

      Install, setup path, first generation, first export.

      Frontend Engineer 1 week

    2. 2

      Funnel by path

      Drop-off at each step.

      Head of Growth 2 weeks

    3. 3

      Fix the worst step

      One change per sprint.

      Product Lead ongoing

    4. 4

      Starter brief

      A one-click first project.

      Content Lead 1 week

    5. 5

      Nudge

      An in-app and email prompt after an hour.

      Lifecycle Marketer 1 week

    6. 6

      Review

      Activation by path each week.

      Product Lead weekly

    First result
    1 month
    Cost
    Engineering time
    Owner
    Product Lead

    Smallest test

    Measure one week of activation by setup path.

    Why it works

    Define activation as a finished file exported within 24 hours of installing. Track it for each setup path, find the path with the most drop-off, and fix its slowest step first. Send a one-click "first project" brief (from idea 8's showcase) to anyone who installs but doesn't generate within an hour.

    A credits business grows when people finish something early. Measuring each path shows where the setup loses them.

    Risks and how to handle them

    • Usage statistics are switched off

      Use aggregate counts and respect the opt-out.

    Evidence

    • Your websiteDocs (124 pages): setup guides for connecting existing chat subscriptions, "Make your first video" and "Find your finished file", a Voyager API with 49 reference pages, cloud machines, "Make browser games", "Share a video for review", "Work as a team" and brand kits in the Library; nothing on referrals, affiliates or education pricing.Source
    • Your websiteFAQ: made by Nullframe, Inc.; for "Creators, editors, motion designers, marketers, and small studios"; users can bring their own chat subscription ("Voyager never bills those subscriptions") or API keys; generation is billed at provider cost, 1 credit = 1¢, with no markup; Team plans cover 2 to 25 people, and Enterprise adds SSO and admin spend controls; payments are generally non-refundable; Voyager ships about 40 skills.Source
  • 13

    Voyager has no referral or affiliate programme: no page in the sitemap, nothing in the 124 docs pages. Its credits are worth exactly 1¢ of provider cost, so a referral reward costs Voyager a known, small amount, and Free users who run out today simply wait for the reset.

    Medium impactLow effort First result in 1 month Lifecycle Marketer

    First action

    Set the reward and the anti-abuse rules.

    How to measure it

    Referral share of sign-ups

    Now
    –
    Target
    10%
    By
    90 days

    Steps

    1. 1

      Rules

      Reward on the friend's first generation, verified email.

      Product Lead 2 days

    2. 2

      Links

      A per-user link in the dashboard.

      Frontend Engineer 1 week

    3. 3

      Placement

      The out-of-credits screen and export screens.

      Product Lead 3 days

    4. 4

      Announce

      Discord, X and the newsletter.

      Community Manager 1 day

    5. 5

      Abuse checks

      Device and email rules, as on Free.

      Product Lead ongoing

    6. 6

      Measure

      Referral share and cost per account.

      Head of Growth monthly

    First result
    1 month
    Cost
    $10 of generation per successful referral
    Owner
    Lifecycle Marketer

    Smallest test

    Offer referral links to Discord members for a month.

    Why it works

    Give each user a referral link. When a friend signs up and makes a first generation, both get 500 credits, which last 12 months like pack credits. Show the link at the moment a Free user runs out of credits, and on export screens.

    The reward is the product itself, it has a fixed cost, and it appears at the moment users most want more credits.

    Risks and how to handle them

    • Self-referral farming

      Use the existing one-free-plan-per-person-and-device rule.

    Evidence

    • Your websitePricing page: Free $0 with 1,000 credits a month (up to 500 per 5 hours); Starter $19 (1,800 credits, 2,700 for the first 3 months); Creator $49, "Most popular"; Ultra $149; GigaMax $500; annual plans "15% off", but only Creator's annual price ($41.65 a month) is shown; Team plans for 2 to 25 people (annual 10% off) and Enterprise for 26+ have no listed prices; credit packs "from $11" at 1.1¢ a credit; billing is prepaid; Free pauses when credits run out.Source
    • Your websiteDocs (124 pages): setup guides for connecting existing chat subscriptions, "Make your first video" and "Find your finished file", a Voyager API with 49 reference pages, cloud machines, "Make browser games", "Share a video for review", "Work as a team" and brand kits in the Library; nothing on referrals, affiliates or education pricing.Source
  • 14

    Free pauses when its 1,000 credits (500 per 5 hours) run out, and Free can't buy packs. Starter is $19 with 1,800 credits, or 2,700 for the first three months, and every credit maps to 1¢ of provider cost. Users hit the wall at the moment they're most engaged, and the price question has an exact answer.

    Medium impactLow effort First result in 1 month Product Lead

    First action

    Write the message using one user's real month.

    How to measure it

    Upgrade rate at the pause screen

    Now
    –
    Target
    +30%
    By
    60 days

    Steps

    1. 1

      Draft

      Message and calculation.

      Product Lead 1 day

    2. 2

      Build

      Usage summary on the pause screen.

      Frontend Engineer 1 week

    3. 3

      Add options

      Upgrade, refer, or wait for the reset.

      Product Lead 2 days

    4. 4

      Email

      The same message for users who leave.

      Lifecycle Marketer 2 days

    5. 5

      A/B test

      Against the current pause screen.

      Head of Growth 4 weeks

    6. 6

      Iterate

      Try the annual plan as the default.

      Product Lead monthly

    First result
    1 month
    Cost
    Engineering time
    Owner
    Product Lead

    Smallest test

    Send the email version to last month's Free users who ran out.

    Why it works

    Replace the generic pause with a personal message: "You've made N images and N clips this month. On Starter that's $X; you'd have Y credits left", plus the launch bonus and a one-click upgrade. Offer the referral link (idea 13) as the free option.

    Showing what the work the user just did would cost turns a dead end into an upgrade, and it fits the no-markup brand.

    Risks and how to handle them

    • Feels pushy

      Always show the free wait option and the reset time.

    Builds on

    Evidence

    • Your websitePricing page: Free $0 with 1,000 credits a month (up to 500 per 5 hours); Starter $19 (1,800 credits, 2,700 for the first 3 months); Creator $49, "Most popular"; Ultra $149; GigaMax $500; annual plans "15% off", but only Creator's annual price ($41.65 a month) is shown; Team plans for 2 to 25 people (annual 10% off) and Enterprise for 26+ have no listed prices; credit packs "from $11" at 1.1¢ a credit; billing is prepaid; Free pauses when credits run out.Source
    • Your websiteHelp docs, "Generate in your browser": "You can make a video clip, an image or a piece of audio on voyager.so/dashboard without installing anything." "Voyager Free includes some image and voice models; video needs a paid plan."Source
  • 15

    Voyager's comparison page notes that Voyager can sign in with Higgsfield to spend Higgsfield credits, which Higgsfield can't do in reverse. Higgsfield claims 40M+ creators, and Voyager's own page concedes Higgsfield's preset-driven direction and unlimited windows. Those users don't have to switch: they can bring their credits and get editable projects on top.

    Medium impactLow effort First result in 1 month Head of Growth

    First action

    Write the connect-and-finish guide in the docs.

    How to measure it

    Higgsfield-connected accounts

    Now
    –
    Target
    1,000
    By
    90 days

    Steps

    1. 1

      Guide

      Connect, generate, finish.

      Content Lead 3 days

    2. 2

      Landing page

      Use case first, with one before-and-after.

      Content Lead 3 days

    3. 3

      Internal links

      From the comparison and alternatives posts.

      Content Lead 1 day

    4. 4

      Track

      Connected-account events.

      Frontend Engineer 2 days

    5. 5

      Short demo

      A 30-second clip for social.

      Community Manager 1 week

    6. 6

      Measure

      Connects, then upgrades.

      Head of Growth monthly

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

    Smallest test

    Publish the guide and link it from the comparison page.

    Why it works

    Publish a page and a short guide: connect a Higgsfield account, generate with existing credits, then cut, caption and score the result in Voyager's editors as an editable project. Target "Higgsfield editor", "edit Higgsfield video" and similar searches, and link it from the Higgsfield comparison and alternatives posts.

    It asks for no switch and no new spend. It's an add-on for an audience many times Voyager's size, built on a feature that already exists.

    Risks and how to handle them

    • The integration changes

      Date the page, and check the connection every month.

    Evidence

    • Your websiteComparison page "Voyager vs Higgsfield" (last verified 8 October 2026): lists Higgsfield's strengths as "Browser access with nothing to install" and preset-driven direction; shows Voyager cheaper per image on every matched plan for one model (4.2¢ against 10.6¢ at the $19 tier); says Voyager can "sign in with Higgsfield to spend Higgsfield credits"; the main call to action is "Download Voyager".Source
    • Competitor researchHiggsfield: "the generative platform for creative production", "40M+ creators", "35+ frontier models", in the browser. Krea: "Trusted by over 30,000,000 users", Free at 100 units a day, Pro at $35 a month. FLORA: "Your creative workspace", "50+ models", "Teams from Pentagram to Lionsgate". Figma Weave (formerly Weavy): a node-based AI workflow canvas.Source
  • 16

    Voyager's Discord has about 327 members, with 56 online. It's small, but these are the most engaged early users. With no showcase and no customer stories on the site, the community is where proof should come from.

    Medium impactLow effort First result in 2 weeks Community Manager

    First action

    Post the first brief this Monday.

    How to measure it

    Entries a week

    Now
    –
    Target
    50
    By
    90 days

    Steps

    1. 1

      Format

      Brief, rules, deadline and prize.

      Community Manager 1 day

    2. 2

      First brief

      A product reveal.

      Community Manager 1 day

    3. 3

      Judge

      Taste, cost and editability.

      Founder & CEO weekly

    4. 4

      Reward

      Credits plus a showcase spot.

      Community Manager weekly

    5. 5

      Grow

      Invite links in the app and the newsletter.

      Lifecycle Marketer ongoing

    6. 6

      Measure

      Members, entries and retention.

      Community Manager monthly

    First result
    2 weeks
    Cost
    About $50 of credits a week
    Owner
    Community Manager

    Smallest test

    Run one brief and count the entries.

    Why it works

    Each Monday, post one brief (for example, "a 15-second product reveal"). Members share the result with their credits spent and the skill they used. The team picks three winners, who get credits, a showcase spot (idea 8) and a mention in the newsletter. The best skills go to the library (idea 7).

    A shared weekly brief gives people a reason to come back, produces comparable proof, and fills the showcase and the skills library.

    Risks and how to handle them

    • Too few entries at first

      Invite 10 known creators directly.

    Builds on

    Evidence

    • CommunityDiscord server "Voyager": about 327 members, 56 online (November 9, 2026).Source
    • Your websiteSitemap (373 URLs, read November 9, 2026): 187 /resources pages, 168 blog URLs, 8 comparison pages (Higgsfield, Flora, Weavy, OpenArt, Krea, Runway, Claude, Codex), 10 app pages, /ai-video and 2 vork pages; no showcase, gallery, templates, customer-story, use-case or referral pages.Source
  • 17

    Voyager comes from Nullframe, the company behind the design tool Moda ("Trusted by thousands of teams"; the launch review reports a founder claim of 100k+ Moda users). Voyager's sign-up already runs through a Voyager-branded account page on moda.app, so the two products share sign-in. Yet Moda's homepage doesn't mention Voyager.

    High impactLow effort First result in 1 month Founder & CEO

    First action

    Add a Voyager link to Moda's export menu.

    How to measure it

    Moda-sourced activations

    Now
    –
    Target
    2,000
    By
    90 days

    Steps

    1. 1

      Pick the moment

      Export of a social or ad asset.

      Product Lead 2 days

    2. 2

      Card and link

      "Animate this in Voyager".

      Frontend Engineer 1 week

    3. 3

      Asset hand-off

      Open the design in Voyager Design.

      Frontend Engineer 2 weeks

    4. 4

      Email

      To Moda users who export ads.

      Lifecycle Marketer 1 week

    5. 5

      Bonus

      Launch credits for Moda subscribers.

      Founder & CEO 1 day

    6. 6

      Measure

      Moda-to-Voyager activation.

      Head of Growth monthly

    First result
    1 month
    Cost
    Engineering time and launch credits
    Owner
    Founder & CEO

    Smallest test

    One email to Moda users who export ads.

    Why it works

    Add a Voyager card to Moda: "Turn this design into a motion graphic or video with Voyager", one click with the same account, the asset carried over. Email Moda users who make social or ad assets, and give Moda subscribers a launch credit bonus.

    Cross-selling to existing, signed-in users of a sister product is the cheapest acquisition there is, and turning a static design into motion is an obvious next step.

    Risks and how to handle them

    • Distracts Moda users

      Show it only after an export, and let users dismiss it.

    Evidence

    • Your websiteModa, Nullframe's design tool: "Create fully-editable, on-brand visual assets on a real canvas you control", "Trusted by thousands of teams"; the homepage does not mention Voyager. Voyager's /sign-up redirects to a Voyager-branded account page on moda.app, so the two products share sign-in.Source
    • Pressexplainx.ai review (October 9, 2026): Voyager launched October 8, 2026 from Nullframe, the company behind the design tool Moda; it reports a founder claim of a Moda user base of "100k+ people" that it could not verify; it flags the "100+" tools claim against the roughly two dozen apps named in the FAQ and "9+" on a comparison page, notes the app is not open source despite the "open harness" name, and found no Hacker News discussion of the launch.Source
  • 18

    The FAQ names small studios and marketers as core users. Team plans exist (2 to 25 people, a shared credit pool, admin roles) and the Library holds brand kits, voices and elements, but Team is sold only as one bill and a credit pool, with no price shown.

    High impactMedium effort First result in 2 months Founder & CEO

    First action

    Interview five paying users who work in studios or agencies.

    How to measure it

    Paid Team seats

    Now
    –
    Target
    300
    By
    2 quarters

    Steps

    1. 1

      Interviews

      Five studio users.

      Founder & CEO 1 week

    2. 2

      Shared Library

      Brand kits, voices and skills at workspace level.

      Product Lead 4 weeks

    3. 3

      Spend controls

      Per-person allowances.

      Product Lead 2 weeks

    4. 4

      Team page

      Studio walkthrough and price.

      Content Lead 1 week

    5. 5

      Trial

      14 days with pooled credits.

      Product Lead 1 week

    6. 6

      Outreach

      Users whose workspace has 2+ people.

      Founder & CEO ongoing

    First result
    2 months
    Cost
    Product time and trial credits
    Owner
    Founder & CEO

    Smallest test

    Offer five studios a manual Team setup.

    Why it works

    Sell Team on what a studio actually needs: one brand kit, voices and skills shared across the team, review links for clients ("Share a video for review"), and spend controls per person. Publish a Team page with a studio walkthrough and a 14-day Team trial with pooled credits.

    Shared assets and skills give studios a reason to bring every seat onto one plan, which raises revenue per account and makes it less likely they leave.

    Risks and how to handle them

    • Studios need SSO first

      Point them to Enterprise, which already lists SSO.

    Builds on

    Evidence

    • Your websiteFAQ: made by Nullframe, Inc.; for "Creators, editors, motion designers, marketers, and small studios"; users can bring their own chat subscription ("Voyager never bills those subscriptions") or API keys; generation is billed at provider cost, 1 credit = 1¢, with no markup; Team plans cover 2 to 25 people, and Enterprise adds SSO and admin spend controls; payments are generally non-refundable; Voyager ships about 40 skills.Source
    • Your websiteDocs (124 pages): setup guides for connecting existing chat subscriptions, "Make your first video" and "Find your finished file", a Voyager API with 49 reference pages, cloud machines, "Make browser games", "Share a video for review", "Work as a team" and brand kits in the Library; nothing on referrals, affiliates or education pricing.Source
    • Your websitePricing page: Free $0 with 1,000 credits a month (up to 500 per 5 hours); Starter $19 (1,800 credits, 2,700 for the first 3 months); Creator $49, "Most popular"; Ultra $149; GigaMax $500; annual plans "15% off", but only Creator's annual price ($41.65 a month) is shown; Team plans for 2 to 25 people (annual 10% off) and Enterprise for 26+ have no listed prices; credit packs "from $11" at 1.1¢ a credit; billing is prepaid; Free pauses when credits run out.Source

Big bets 3

Larger plays that can change the trajectory.

  • 19

    The FAQ says "Mobile apps are coming soon". Agent work in Voyager runs for minutes or longer, with approvals and spending controls, and the docs already cover sharing a video for review. A creative director doesn't need to edit on a phone, but they do need to approve a cut, an expensive generation or a client review while away from the desk.

    High impactHigh effort First result in 2 quarters Product Lead

    First action

    Count how often desktop jobs sit waiting for an approval.

    How to measure it

    Approval wait time

    Now
    –
    Target
    -50%
    By
    2 quarters after launch

    Steps

    1. 1

      Measure waits

      Time jobs spend waiting for the user.

      Product Lead 2 weeks

    2. 2

      Scope

      Progress, approvals and review only.

      Product Lead 2 weeks

    3. 3

      Design

      Three screens.

      Designer 3 weeks

    4. 4

      Build

      On the existing review and approvals features.

      Mobile Engineer 2 months

    5. 5

      Beta

      Discord members first.

      Community Manager 1 month

    6. 6

      Launch

      Store listings and a launch post.

      Founder & CEO 2 weeks

    First result
    2 quarters
    Cost
    One mobile engineer for about a quarter
    Owner
    Product Lead

    Smallest test

    Push notifications for approvals through the browser dashboard first.

    Why it works

    Scope the first phone release around three jobs: watch progress on a running project, approve spending and choices, and review or share drafts with comments. Leave creation to the desktop and browser. Use push notifications for "draft ready" and "approval needed".

    Approvals from a phone keep long agent jobs moving, which means more finished projects and more credit use, and push notifications bring people back without email.

    Risks and how to handle them

    • Users expect creation on mobile

      Position it clearly as a companion and link to the browser generator.

    Evidence

    • Your websiteDownload page offers macOS, Windows and Linux ("Choose your computer. We'll help you get started.") with no email capture. The FAQ lists Apple silicon Macs on macOS 15+ ("Intel isn't supported yet"), Windows 11 (x64) and Linux (x64), and says "Mobile apps are coming soon."Source
    • Your websiteDocs (124 pages): setup guides for connecting existing chat subscriptions, "Make your first video" and "Find your finished file", a Voyager API with 49 reference pages, cloud machines, "Make browser games", "Share a video for review", "Work as a team" and brand kits in the Library; nothing on referrals, affiliates or education pricing.Source
  • 20

    Voyager already tests every model the same way ("Twelve pictures for every image model, nine clips for every video model"), publishes per-second prices, and argues on /ai-video that the real cost is tries per usable clip ($9.45 for 4 tries against $92.44 for 20 tries elsewhere). No rival publishes that number across models.

    High impactMedium effort First result in 2 months Content Lead

    First action

    Define "usable" and score last month's video tests.

    How to measure it

    New referring domains

    Now
    –
    Target
    50
    By
    2 quarters

    Steps

    1. 1

      Method

      A usability rubric reviewed by two people.

      Content Lead 1 week

    2. 2

      Score

      The existing test outputs.

      Content Lead 2 weeks

    3. 3

      Page

      Sortable table, method and raw outputs.

      Frontend Engineer 1 week

    4. 4

      Chart

      One shareable chart each month.

      Designer monthly

    5. 5

      Outreach

      Creator newsletters and tech press.

      Head of Growth monthly

    6. 6

      Update

      New models and price changes.

      Content Lead monthly

    First result
    2 months
    Cost
    Generation credits for tests plus content time
    Owner
    Content Lead

    Smallest test

    Publish the index for the top 10 video models.

    Why it works

    Each month, publish an index: for each video and image model, the share of test outputs that were usable, the price per second or image, and the resulting cost per usable result, with the method and raw outputs linked. Release it as a page, a chart for social, and a short report journalists and creators can cite.

    A neutral, repeatable benchmark earns links and citations every month, and it makes Voyager's core claim (fewer tries, provider cost) measurable.

    Risks and how to handle them

    • Seen as biased

      Publish the raw outputs and the method, and invite challenges.

    Evidence

    • Your websiteModels hub "AI image, video and voice models, tested": 79 tested model cards (38 image, 31 video, 10 voice) with real test outputs and prices; its only call to action is "Get Voyager" to the download page. Model pages such as "Veo 3.1: Pricing, Free Access and Real Clips (2026)" and "Kling 4.0: Pricing, 30-Second Clips and Real Tests (2026)" offer "Get Voyager" (to /download) and "Sign up", with no link to try the model in the browser.Source
    • Your website/ai-video page: "Voyager: the most cost-effective platform for AI video"; per-second prices against fal, Runway and Higgsfield (e.g. Seedance 2.5 at 720p, $0.2362 a second, "−50% vs fal"); "One clip you keep" at $9.45 for 4 tries against $92.44 for 20 tries elsewhere; a "Show all 127 models" table; "Download for macOS ↓" as the main call to action.Source
  • 21

    Voyager's docs already cover adding Voyager to three chat and coding agents, and the FAQ says those subscriptions are never billed by Voyager. People who already pay for an agent are Voyager's cheapest users to serve, but today they have to find Voyager first and then follow setup docs.

    High impactMedium effort First result in 1 quarter Developer Relations

    First action

    List each client's directory rules and submission requirements.

    How to measure it

    Directory-sourced accounts

    Now
    –
    Target
    3,000
    By
    2 quarters

    Steps

    1. 1

      Requirements

      Each directory's rules.

      Developer Relations 1 week

    2. 2

      Package

      Connector, icon, description and first task.

      Developer Relations 3 weeks

    3. 3

      Review

      Security and permission scopes.

      Product Lead 2 weeks

    4. 4

      Submit

      One directory at a time.

      Developer Relations 1 month

    5. 5

      Onboard

      A guided first task after install.

      Product Lead 2 weeks

    6. 6

      Measure

      Installs and conversion to paid.

      Head of Growth monthly

    First result
    1 quarter
    Cost
    Developer relations time
    Owner
    Developer Relations

    Smallest test

    Submit to the one directory with the simplest process.

    Why it works

    Package the existing connector as listed apps in each agent client's directory or marketplace, with a clear first task ("make a 15-second product video from this image"). Track installs from each directory, and use the listing pages to promote the browser generator and the desktop app.

    Directories put Voyager where agent users already look for new abilities, and the setup is already documented.

    Risks and how to handle them

    • Directory rules change or reject

      Keep the direct setup docs as the fallback.

    Evidence

    • Your websiteDocs (124 pages): setup guides for connecting existing chat subscriptions, "Make your first video" and "Find your finished file", a Voyager API with 49 reference pages, cloud machines, "Make browser games", "Share a video for review", "Work as a team" and brand kits in the Library; nothing on referrals, affiliates or education pricing.Source
    • Your websiteFAQ: made by Nullframe, Inc.; for "Creators, editors, motion designers, marketers, and small studios"; users can bring their own chat subscription ("Voyager never bills those subscriptions") or API keys; generation is billed at provider cost, 1 credit = 1¢, with no markup; Team plans cover 2 to 25 people, and Enterprise adds SSO and admin spend controls; payments are generally non-refundable; Voyager ships about 40 skills.Source

Wildcards 2

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

  • 22

    Voyager's apps hub includes Roblox Studio, Godot, Unity and Unreal Engine, the docs have a "Make browser games" guide, and recent blog posts walk through a Roblox survivors-like game and a third-person platformer. Game makers are an audience Voyager already writes for, and its creative-video rivals don't.

    Medium impactMedium effort First result in 6 weeks Community Manager

    First action

    Pick the dates and the prize credits.

    How to measure it

    Jam sign-ups

    Now
    –
    Target
    500 entrants
    By
    First jam

    Steps

    1. 1

      Rules

      Theme, engines, length and judging.

      Community Manager 1 week

    2. 2

      Starter kits

      A skill and a template per engine.

      Developer Relations 2 weeks

    3. 3

      Promote

      Game-dev communities and the blog.

      Community Manager 2 weeks

    4. 4

      Run

      Two weeks, with daily help in Discord.

      Community Manager 2 weeks

    5. 5

      Judge

      Playable entries only.

      Founder & CEO 1 week

    6. 6

      Showcase

      Winners' games and skills.

      Content Lead 1 week

    First result
    6 weeks
    Cost
    Prize and starter credits, about $2,000
    Owner
    Community Manager

    Smallest test

    A one-weekend jam in Discord.

    Why it works

    Host a two-week online jam: build a small game where the agent writes the code and generates the art, sound and music. Give entrants starter credits, judge on fun and craft, feature the winners in the showcase, and publish each winner's skills to the library.

    Jams produce shareable, playable proof and reach young creators where the creative-video rivals aren't competing.

    Risks and how to handle them

    • Young entrants and platform rules

      Follow each engine's terms and set a minimum age for prizes.

    Builds on

    Evidence

    • Your websiteApps hub: "10 apps · facts checked 23 October 2026" (After Effects, Blender, CapCut, DaVinci Resolve, Godot, Houdini, Premiere Pro, Roblox Studio, Unity, Unreal Engine), while the homepage says "100+ Creative tools" and the Voyager vs Claude page says skills cover "9+ creative apps". The Blender page lists "Try asking" prompts but has no video demo, gallery or downloadable project.Source
    • Your websiteBlog: 161 articles (Techniques 61, Agents 34, Generators 31, Editors 27, Models 4, Voyager 4), several published a day in November 2026, including Roblox Studio game guides and statistics roundups; every post ends at "Get Voyager", and there is no newsletter signup.Source
    • Your websiteDocs (124 pages): setup guides for connecting existing chat subscriptions, "Make your first video" and "Find your finished file", a Voyager API with 49 reference pages, cloud machines, "Make browser games", "Share a video for review", "Work as a team" and brand kits in the Library; nothing on referrals, affiliates or education pricing.Source
  • 23

    The FAQ says the Mac app needs Apple silicon and macOS 15 or later ("Intel isn't supported yet"), the homepage hero offers "Download for macOS", and the download page has no email capture. An Intel Mac owner who clicks download hits a wall with no way to stay in touch.

    Low impactLow effort First result in 2 weeks Frontend Engineer

    First action

    Count Intel Mac visits to /download in analytics.

    How to measure it

    Intel visitor save rate

    Now
    –
    Target
    30%
    By
    60 days

    Steps

    1. 1

      Size it

      Intel Mac share of download visits.

      Head of Growth 1 day

    2. 2

      Detect

      Intel Macs where the browser reports it.

      Frontend Engineer 1 day

    3. 3

      Swap the button

      Browser generator first.

      Frontend Engineer 1 day

    4. 4

      Waitlist

      One email field.

      Lifecycle Marketer 1 day

    5. 5

      Decide

      Whether an Intel build is worth it, from the list size.

      Product Lead 1 quarter

    6. 6

      Notify

      Email the list at launch.

      Lifecycle Marketer at launch

    First result
    2 weeks
    Cost
    $0
    Owner
    Frontend Engineer

    Smallest test

    Add a line under the Mac button: "On an Intel Mac? Use the browser."

    Why it works

    Where the browser reports an Intel Mac, swap the download button for "Generate in your browser now" (idea 1) and a one-field "Email me when Intel Macs are supported" form. Use the list size to decide whether an Intel build is worth it.

    It saves a share of visitors who'd otherwise leave, and the list gives a measured answer to a roadmap question.

    Risks and how to handle them

    • Detection is imperfect

      Always show a link to every download.

    Builds on

    Evidence

    • Your websiteDownload page offers macOS, Windows and Linux ("Choose your computer. We'll help you get started.") with no email capture. The FAQ lists Apple silicon Macs on macOS 15+ ("Intel isn't supported yet"), Windows 11 (x64) and Linux (x64), and says "Mobile apps are coming soon."Source
    • Your websiteHelp docs, "Generate in your browser": "You can make a video clip, an image or a piece of audio on voyager.so/dashboard without installing anything." "Voyager Free includes some image and voice models; video needs a paid plan."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 #9 (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)
More example playbooks and reviews
ProgrammableDigital

Show Voyager’s product working

Turn screenshots or the website into an animated product demo in its own brand, for the homepage, launch posts and outreach in this playbook.

Make a demo (opens programmable.digital in a new tab)

More growth playbooks

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

  • Veritus growth strategyAn AI communications platform for financial services: compliant voice, SMS, email and chat agents for origination, servicing and…23 growth ideas
  • Datadog growth strategyAn observability and security platform: infrastructure and application monitoring, logs, security, digital experience, software delivery…23 growth ideas
  • Agiloft growth strategyA no-code, data-first contract lifecycle management platform with Astra AI contract intelligence (including Screens), an integration hub…22 growth ideas
  • Gratwick growth strategyAn independent studio and creator platform: productions surrounded by live interactive shows, audience casting, trivia, podcasts, a…21 growth ideas
  • Talkspace growth strategyOnline therapy and psychiatry with licensed providers by messaging, video and voice, covered by most major insurance plans, plus Talkspace…21 growth ideas
  • January growth strategyAn AI-led debt collection agency (formerly Debtsy) that recovers charged-off and late consumer debt for banks, credit unions, fintechs…21 growth ideas
  • Shive-Hattery growth strategyA privately held architecture and engineering firm founded in 1895: architecture, interiors, civil, structural, mechanical, electrical…20 growth ideas
  • Cabify growth strategyA ride-hailing and urban mobility platform: VTC and taxi rides, airport transfers, motorbikes, deliveries and logistics, corporate…23 growth ideas

See every playbook and website review

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

Something wrong, or do you represent Voyager? 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 307 ideas

    Browser trial on model pages, a lighter homepage, every platform in the hero, claims that match the app pages, full pricing, email capture and an Intel Mac waitlist.

  2. Days 31 to 6012 ideas

    Loops and proof: shared skills, a showcase, vork to Vork Cloud, GPU prices, app demos, activation, referrals, upgrade prompts, Higgsfield-credit users, Discord, the Moda cross-sell and Team.

  3. Days 61 to 904 ideas

    Commit to the bigger bets: phone review, a cost-per-usable-clip index, agent-client directories and a game-making wildcard.

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 editable projects instead of one-off clips. The most cited problem: paying for many tries per usable clip.

  • Provider-cost pricing with no markup
  • Jumping between generators and editors
  • In their words: “harness”

Market

Competitive landscape

4 competitors compared: Higgsfield, Krea, FLORA and Figma Weave.

  • Gap to exploit: An agent that finishes work in the apps creatives already own
  • Higgsfield: Rendered clips, not editable projects
  • Krea: No control of desktop creative apps

From the team that wrote this playbook

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

© 2026 AI-Powered Growth Intelligence