Your Growth Prioritization Engine
Get actionable insights out of your analytics, heatmaps, surveys, reviews, form analytics, and calls. All analyzed with the same processes that Conversion Optimization experts use. Track your progess/results (win rate and velocity), run experiments, and gather insights all in one platform.
The story behind the platform
Built by a consultant who did this work by hand
I'm Ryan Johnson. I spent years as a CRO and SEO consultant — personally coached by Peep Laja, the most influential conversion optimization expert in the world. I helped brands like Silver Jeans, Fresh Tracks Canada, and Adventures Abroad generate tens of millions in additional revenue.
Every engagement ended the same way: I'd hand my client a prioritized list of everything they needed to do to grow their site. They started calling it their "Growth Roadmap."
I built this platform to automate that entire methodology — the research, the analysis, the prioritization — so any business can get the same roadmap without hiring a consultant.
The problem
The Real Cost of No Synthesis Layer
Most CRO teams aren't failing because they lack data. They're failing because nobody connects it.
Your tools don't talk to each other
GA4 says bounce rate is 68%. Heatmaps show a rage click on the CTA. Your survey says "confusing." Three tools, three tabs, no one has time to cross-reference them into a hypothesis.
Without synthesis, every test is a guess
"Let's test the CTA color" is an opinion. Opinions win as often as they lose. A finding backed by GA4 drop-off data, a heatmap rage cluster, and a survey quote is a hypothesis — and hypotheses have a reason to win.
Tests you can't trust
Standard p-values lie when you peek early. Every time you check results mid-test, you inflate your false-positive rate. Most teams ship "winners" that were noise.
Two subscriptions. One CRO system.
Research What to Test. Prove It Works.
You can subscribe to either on its own. They work better together.
Conversion Research
Starts at $279/mo (annual)AI runs 30 analysis lenses across your GA4, heatmaps, surveys, form analytics, and device data. It reads them together — the way a real analyst would, not in six separate tabs — and surfaces a prioritized list of findings, each one tagged with what to do next: test it, just ship the fix, or dig deeper first.
Every finding links to the source data behind it. Not "your CTA has low contrast." The GA4 exit report, the heatmap rage cluster, and the survey quote that all point the same direction.
- 30+ GA4 reports cross-referenced (funnels, exits, landing pages, device split, geo, source/medium)
- Heuristic audit on 8 dimensions: Clarity, Relevance, Value Proposition, Distraction, Urgency, Friction, Anxiety, Motivation
- Five-persona Buyer Panel walks your funnel as a Ready Buyer, Skeptic, Price-Conscious Buyer, Confused Visitor, and Comparison Shopper
- Paid-traffic scent check: compares the promise on your ad source to the promise on the landing page it hits
- Per-field form analytics (drop-off, time-to-complete, refill patterns)
- Each finding tagged with a PIE score, a PXL score, a severity rating, and one of five action buckets (test / just ship it / instrument / hypothesize / investigate)
- One-time audit option available — no subscription required
$279/mo for up to 50K visits/mo (annual). Monthly option from $349. One-time audits from $698.
Testing & Insights
Free to 10K MTUs/moA Bayesian engine (50,000 simulations per metric) tells you the real probability your variant is winning — not a yes/no significance flag. mSPRT keeps the math valid even if you check results every hour. An AI chat writes the variant code from plain English. Per-variant heatmaps and follow-up surveys are wired in.
Drop one script tag. That's the whole install.
- Bayesian probability to beat control (50,000 simulations), mSPRT for always-valid sequential testing, χ² SRM check on every run
- AI variant editor: describe the change in English, AI writes JS/CSS against your live page, you review the diff before it ships
- Heatmaps, on-site surveys, and form analytics included — same script tag
- GA4 segment slicing: slice results by device, source, audience, or any custom attribute
- Traffic ramping, scheduling, preview URLs, audience targeting rules
- CUPED available as an opt-in for high-recurrence products (streaming/SaaS) — reduces traffic needed to confirm lift by up to 50%
- Server-side SDK for Next.js, Node, Express, Remix, edge functions
Free to 10K MTUs/mo. $239/mo for 50K MTUs (annual). Monthly from $299.
Closed-loop CRO
Research Finds It. Testing Proves It.
Most research dies in a slide deck. Most tests start with a guess. When the two subscriptions are active, findings flow directly into experiments.
Conversion Research surfaces a finding — a drop-off in the checkout form, a rage cluster on the CTA, a message-match gap from paid traffic.
One click promotes it to a draft A/B test. The hypothesis, target page, and variant idea are pre-filled. The AI Test Coder writes the variant code.
The test runs. The result writes back to the finding and to the team knowledge base. The next Conversion Research run reads what's already been tested — so it doesn't send you back to a hypothesis you already answered.
Trusted by growing businesses
How It Works
Connect your data (5 minutes)
Link GA4 (read-only) and drop the SDK on your site. One script tag covers heatmaps, surveys, form analytics, and A/B experiment assignment. No separate installs.
Run Conversion Research
Hit Run. The AI analyst pulls your data, runs 30 lenses, cross-references everything, and emails you when the prioritized list is ready — usually within the hour.
Promote the top finding to an A/B test
One click. The hypothesis, target page, and variant idea are pre-filled. The AI Test Coder writes the variant code. You review before it ships.
Results compound
When a test concludes, the result writes back to the originating finding and the team knowledge base. Every future run learns from what's already been tested.
Who it's for
The analysis adapts to your funnel
The lenses, scoring, and finding format are the same. What they surface depends on your business model.
DTC ecommerce conversion research
Mobile PDP friction, add-to-cart drop-off, checkout abandonment, and shipping cost reveal — the issues that quietly kill ecommerce conversion rates.
See how it works for DTC ecommerce
SaaS self-serve conversion research
Pricing page confusion, trial-to-paid drop-off, activation blockers, and upgrade friction — the leaks that keep MRR growth flat.
See how it works for SaaS self-serve
Lead generation conversion research
Form field drop-off, trust-signal gaps, and landing page message-match for paid traffic — the barriers between a click and a booked call.
See how it works for lead generation
Common Questions
How do I know the AI isn't just generating findings that sound good?
Every finding links to the source data behind it: the GA4 report, the heatmap clip, the survey quote, the form drop-off number, the page screenshot. If the evidence doesn't hold up, you don't ship the test. You can also feed your own past winning experiments into the team knowledge base, which the AI reads on every run — so the recommendations get sharper as your library grows.
Do I need both subscriptions?
No. Conversion Research is for teams who want the diagnosis — what's leaking, where, and what to do about it. Testing & Insights is for teams who want to run the experiments themselves. Each works on its own. When you use both, findings flow directly into experiments with one click, and test results write back into the research.
Do we have enough traffic for the findings to be useful?
Research and testing have different traffic requirements. Conversion Research works at any volume — heuristic analysis, Buyer Panel walkthroughs, heatmap sessions, and form analytics all surface friction without needing statistical significance. A/B testing is where traffic matters. The sample-size calculator in Testing & Insights tells you the days-to-significance estimate from your real conversion rate before you start, so you don't waste three weeks on a test your site can't reach significance on.
Do I need a developer to write variant code?
No. The AI variant editor in Testing & Insights takes plain-English requests — "make the CTA orange and change the copy to Get started in 60 seconds" — and writes the JS/CSS for you. It reads your live page HTML to pick the right selectors. You review the diff before anything ships. A developer is only needed for server-side rendering with the npm SDK.
I don't want to get locked in.
Pay monthly, cancel anytime. Annual plans get a discount. Your data stays yours.
Start with a Free Audit on Your Own Site
Paste any page from your site. In about a minute, you'll get 3–5 prioritized findings on that exact page — the same output as a full Conversion Research run, scoped to one page. No signup. No install.
Conversion Research from $279/mo · Testing & Insights free to 10K MTUs