The Method · Performance & Growth
The conversion rate optimization agency that refuses to lie with statistics
Every visitor you already pay for — through search, ads or content — is an asset most sites waste. CRO is how you stop wasting them: research, hypotheses, disciplined tests, and gains that compound on every channel at once.
The Standard · Pricing
Pricing — and exactly what you get
CRO Audit
$499 one-time
It is the first month of research priced as a document — credited in full against any program started within 30 days.
- Heuristic review by a senior optimizer
- Analytics forensics on your funnel
- Session-recording analysis
- Prioritized test backlog, scored impact × effort
- Credited in full against any program within 30 days
Program
$999 /mo
Boutique CRO shops run $5,000–$8,000/month and flat-fee leaders charge $25,000 a quarter — same statistics, honest price.
- 2–3 tests per month, hypothesis to documentation
- Sample sizes computed before launch — no early stopping
- Ongoing research: recordings, polls, funnel analysis
- Losers and inconclusives reported next to winners
- Wins re-validated against revenue a month later
Growth
$1,999 /mo
The full cycle at velocity — dedicated research, design and development so tests never stall in someone else’s backlog.
- 4–6 tests per month
- Dedicated research + design + development
- Landing-page builds for paid traffic
- Quarterly conversion-strategy review
- Same honest statistics discipline throughout
The discipline
Research decides what to test. Statistics decide what won.
The rules we won't break
- → Sample size computed before launch, not discovered after
- → No early stopping when a variant "looks like" a winner
- → Losers and inconclusives reported next to winners
- → Wins re-validated against revenue, not just clicks
- → Low traffic? Research-driven CRO — never fake tests
The dirty secret of the testing industry is that most "wins" are noise, harvested by checking results daily and stopping the moment the graph blushes green. It photographs well in a case study and quietly does nothing for revenue — which is why so many companies have run fifty tests and can't point to the money.
We run the boring, correct version. Research first: session recordings, funnel analytics, exit polls, user tests — because the best hypotheses come from watching real people struggle, not from a "100 CRO ideas" listicle. Then each test is pre-registered with a hypothesis, a primary metric and a computed sample size, and it runs until the mathematics — not the mood — says stop.
The compounding matters more than any single win. A 6% lift on a checkout multiplies every channel you run — the SEO program, the ad account, the email flows — simultaneously and forever. That's why CRO sits in the growth zone of the system: it's the multiplier on everything else.
Anatomy of a test
One test, end to end
It starts in the research file: session recordings show visitors stalling on the pricing table; exit polls say they can't tell which plan fits. That's a hypothesis worth money — "clarifying plan-fit guidance will raise checkout starts" — not a hunch about button colours.
Before launch, the test gets registered: primary metric (checkout starts), guardrails (revenue per visitor, refund rate), minimum sample computed from current traffic, expected runtime. Design and build follow — the variant produced by the same design system as your site, so the test measures the idea, not a styling glitch.
Then the hard part: nothing. The test runs to its computed sample even when the early graph looks exciting, because early graphs always look exciting. When it closes, the analysis reads out the primary metric, the guardrails and the segments — and the result gets documented either way, because a well-run loser teaches the backlog as much as a winner.
Winners ship to 100% and get re-measured against revenue a month later — the honesty check most agencies skip, and the reason our documented wins stay won.
Questions we hear most
Conversion rate optimization — FAQ
How much traffic do I need for CRO to make sense?
Enough conversions, not enough visitors — the unit that powers a test is the conversion event. As a rough bar: 200+ conversions a month per tested page supports clean A/B testing. Below that we do not fake it with underpowered tests; we run research-driven CRO instead — heuristic review, session recordings, user testing and copy work — which improves low-traffic sites honestly.
What does "honest statistics" mean in practice?
Pre-registered hypotheses, minimum sample sizes computed before launch, no peeking-and-stopping the moment a variant lucks into significance, and reporting that includes losers and inconclusives. Roughly one in four well-built tests wins. An agency reporting an 80% win rate is measuring something other than reality, and you are paying for the fiction.
Which tools do you run tests on?
Whatever fits your stack and budget: GrowthBook or PostHog for engineering-friendly teams, VWO or Convert on the commercial side, plus GA4, Clarity or Hotjar for the research layer. Tool cost is separate and transparent; we have no reseller margins baked in, so the recommendation is what fits, not what pays us.
Why is the audit credited against programs?
Because it is the first month of work, not a sales brochure. The $499 CRO audit produces the research file and the prioritised test backlog a program starts from — so if you continue within 30 days, you have already paid for that work and we credit it in full. Same policy as our $349 Deep SEO Audit.
Stop guessing what your visitors want
Start with the $499 audit — a real research file and a prioritised backlog, credited in full if you continue. Honest maths from day one.