Naïve's marketplace lists 42 companies "run by agents", but no card shows a single live deployment, view, follower or dollar earned. The only number next to "Deploy" is "View Benchmark", and the benchmark behind the faceless-channel template is a simulation where "simulated viewers decide whether to like, follow and buy". The template pages are honest that results are uncertain: the faceless page says payouts are possible but "It can, but it is hard", and the agency page says "naïve does not find clients for you". An independent review in October could not find "any evidence that customers make money with agents built on Naive". Meanwhile Naïve already scores tournament entrants on "Verified revenue earned during the round". The proof is being collected, just not shown where people decide.
First action
Pull 30-day deployments, spend and verified revenue for every template from the tournament's revenue check.
How to measure it
Template-page deploy conversion
- Now
- –
- Target
- +25%
- By
- 60 days
Steps
Pull the data
Deployments, spend and verified revenue per template, last 30 days.
Founding Member of Technical Staff, Agents 3 days
Set the opt-in
Owners choose to show revenue and a public company page; default is anonymised totals.
Founding Product Engineer 3 days
Design the panel
Live companies, median spend, share earning, best verified result, last updated date.
Founding Product Engineer 3 days
Run three in public
The team runs three templates with real budgets and logs spend, posts and revenue.
Founding Growth Engineer 30 days
Ship on the top templates
Faceless channel, clipping channel and agency first.
Founding Product Engineer 1 week
Write two case studies
From opted-in companies with verified numbers.
Growth Engineer Intern 2 weeks
Measure
Deploy rate and trial-to-paid, with and without the panel.
Founding Growth Engineer monthly
Why it works
Add a "Running now" panel to each template card and page, fed by the same revenue verification the tournament uses: live companies on this template, median weekly spend, share that earned any revenue in the last 30 days, and the best verified 30-day revenue (with the owner's opt-in). Start with the three templates Naïve itself runs or has the most deployments on. Where there is no live data yet, say so and show a dated, real run by the Naïve team (actual spend, posts, views, revenue) instead of generated examples. Link each panel to an opt-in public company page.
Visitors deciding whether to spend $20, plus formation and media costs, want to know if anyone like them got a result. Real, verified numbers (even modest ones) beat simulated scores, answer the reviewer's main objection, and give the new growth team the case studies the role asks for.
Ready to use
Checklist"Running now" panel spec
[ ] Live companies on this template (count) [ ] Median weekly spend [ ] Share with verified revenue in the last 30 days [ ] Best verified 30-day revenue (opt-in) [ ] Link to opt-in public company pages [ ] "Last updated" date [ ] Fallback: one dated team-run log when there is no data yet
Risks and how to handle them
Low numbers discourage visitors
Show medians and spend next to results, and pair them with the honest earnings notes the pages already carry.
Owners don't want revenue shown
Opt-in only, with anonymised totals by default.
Evidence
- Your websiteCompany marketplace (/businesses): "Companies run by agents." Tabs: All (42), Video channels (33), Agencies (07), Stores (01), Websites (01). Every visible card is credited "byNaïve" and shows "From $20/mo per company, includes $20 of credits", "View Benchmark" and "Deploy" or "Preview"; no card shows deployments, live companies, revenue, views or ratings.Source
- Your websiteFaceless video template (/businesses/ai-media-channel): H1 "Run your Faceless Video Business with Agents"; "You approve every post."; on earnings it lists the YouTube Shorts bar (1,000 subscribers and 10M Shorts views in 90 days), says TikTok Creator Rewards pays only on videos of at least a minute and "most of these run shorter", and concludes "It can, but it is hard."; it shows "Example videos, generated for this page" with no channel stats; the score shown is a simulated benchmark of models.Source
- Your websiteShort Form Bench: a simulation where "simulated viewers decide whether to like, follow and buy"; each one-week season starts with $1,500 and 40 units of stock; the best model ended at $2,949 of business value against $1,793 for a do-nothing baseline; 32 seasons and 12,844 tool calls; teams can ask to run it on their own agent.Source
- Your websiteAutonomous Company 30 Day Tournament: build a company on naïve "that earns real revenue. The most profitable one wins a Lamborghini"; second $50,000, third $25,000, top 10 get $2,000 in credits; the score is "Verified revenue earned during the round"; a new round starts on the 1st of every month; the top-25 board read "Standings pending" on November 9, 2026.Source
- ReviewsIndependent desk review "Naive AI Review (2026): An LLC, Not a Buyer" (October 6, updated October 17, 2026; the reviewer discloses a competing product): could not find "any evidence that customers make money with agents built on Naive"; formation pricing is not on the pricing page (a docs example totals $449 for a Wyoming filing); cites company-reported "30,000+" developer sign-ups.Source