A synthetic audience for your landing page
See which visitors bounce, and why, before you launch
1,000 synthetic visitors read your landing page, each arriving from a real traffic source: search, Hacker News, Reddit, your ads. In about a minute you see who leaves at the hero, what stops each group, and what to fix first.
From $29/month. 1,000 visitors per run. Rerun after every change.
The problem
Real traffic is an expensive way to test a page
You find out a page does not work by losing the visitors it was for.
Launches happen once
A Show HN post or a Product Hunt day sends its crowd one time. If the hero loses them, they do not come back.
Ads bill by the click
Every visitor who bounces from a paid click is budget spent on a page that did not work.
Analytics arrive late
A week later you learn the bounce rate. You do not learn who left, or why.
How it works
A test run before the real one
- 01
Paste a URL
We fetch the page and pull out what a visitor actually reads: headline, promise, proof, pricing, calls to action.
- 02
Check the audience
Claude proposes who lands on this page and where they come from. Edit the segments, sources and weights before anything runs.
- 03
Read the readout
1,000 visitors each decide to bounce, scroll, click or sign up. You get distributions, objections and a fix list in about a minute.
The readout
Numbers, not a persona chat
Every visitor makes the decisions real traffic makes. The report counts them.
- Bounce by traffic source
- Which arrivals leave at the hero, and which stay to read.
- Comprehension
- Whether visitors can say what the product does after the first screen.
- Objections by segment
- The top reasons each segment hesitates, as shares, not quotes.
- Section by section
- Which sections move visitors forward and which are dead weight.
- Prioritized fix list
- Concrete changes, each citing the number it came from. Copy as markdown.
- Rerun diffs
- Change the headline, rerun, and see exactly which numbers your edit moved.
Sample report
See a real report
We ran Mock Audience on screenies.app, an App Store screenshot generator. 1,000 synthetic visitors read it on 2026-10-02.
- Bounced
- 50%
- 93% of off-target visitors, 22% of on-target
- Understood the product
- 78%
- Could say what it does after the first screen
- Top objection
- 58%
- "I cannot try it before paying $9"
- Sign-up intent
- 8.4%
- Compare it across sources and reruns
Top fix: Add a free preview above the pricing cards: 58% of readers could not try a generation before paying.
Rerun
Fix one thing, rerun, see what moved
Then we applied one fix from the list and ran the same audience again.
Change: Rewrote the subheadline to start with "For app teams, agencies, and app builders".
| Number | Before | After | Change |
|---|---|---|---|
| Understood the product | 78.1% | 81.6% | +3.5 pts |
| "Feels aimed at someone else" | 25.1% | 32.1% | +7.0 pts |
| Bounced | 49.7% | 49.7% | no real change |
| Sign-up intent | 8.4% | 7.9% | no real change |
Almost all of the rise in "aimed at someone else" came from off-target visitors and hobbyists. The right visitors understood more, and the wrong ones saw sooner that it was not for them.
Small shifts are grayed out, so you only see what your edit actually changed.
Alternatives
What you would do instead
| Question | Ask ChatGPT | Post a roast thread | Launch and watch analytics | Mock Audience |
|---|---|---|---|---|
| Who reads the page | One model, one opinion | Whoever replies | Real visitors, once | 1,000 visitors from your traffic sources |
| What you get back | General advice | Comments and hot takes | A bounce rate, no reasons | Bounce by source, objections by segment, a fix list |
| When | Before launch | Hours to days | After launch | Before launch, in about a minute |
| After you edit the page | New chat, new answer | Ask again | Wait for more traffic | Rerun the same audience and see what moved |
Pricing
Pick the allowance that fits how often you ship
Subscriptions only. Every plan runs 1,000 visitors per page and includes reruns with diffs and share links.
Every run simulates 1,000 visitors on every plan. Prices exclude tax; tax is added at checkout. Allowances reset monthly and do not carry over. No free plan; cancel any time.
Questions
Before you ask
- Are the numbers real conversion rates?
- No. They show how your page compares: which sources bounce most, which sections do the work, and whether an edit helped. Use them to decide what to fix first.
- Does it send traffic to my site?
- No. We fetch the page once per run, like a browser does. The 1,000 visitors are simulated, so your analytics and ad budget are untouched.
- Why not paste my page into ChatGPT?
- A chat gives one opinion, and a new chat gives a different one. Mock Audience asks a decision model about each of 1,000 visitors, each with a source, role and intent, and counts the answers. You get shares you can compare and rerun.
- How is this different from persona tools?
- Persona tools interview 8 to 12 personas and return quotes. Mock Audience visitors make the decisions real traffic makes, in volume, so you get distributions instead of anecdotes.
- Does it work on any page?
- Any publicly reachable landing page rendered in the server HTML. Pages that block fetches or ship an empty shell are reported as unusable rather than guessed at.
- What do you keep from my page?
- The page text we fetch, the page summary and audience we generate, your edits and the run results. You can delete your account and its data at any time.
- Can I share a report?
- Yes. Each report can be published as a read-only link. Pro and Agency links carry no Mock Audience branding.
- Can I cancel?
- Yes, at any time from the billing page. Your plan stays active until the end of the period you paid for.
Run 1,000 visitors through your page before the real ones arrive
Paste a URL, check the audience, and read the fix list in about a minute.
Test your landing page