What is conversion rate optimization and what does a CRO engagement include?
CRO is a measurement loop, not a redesign. Here is what a real engagement covers.
Conversion rate optimization is a repeating loop: research what visitors do, form a hypothesis about why they leave, change one thing, measure the result, and repeat. It is not a redesign, and anyone selling it as one is selling a redesign.
The loop
- Research. Quantitative first: where do people enter, where do they drop, which sources convert. Then qualitative: session recordings, scroll data, sales call notes, and the questions your team answers on the phone every day.
- Hypothesis. Written as a sentence with a reason. Because 70 percent of visitors never reach the form, moving it above the case study section should increase submissions. A hypothesis without a because is a guess.
- Test or fix. If traffic supports a test, run one. If it does not, just make the change and monitor.
- Measure. Against qualified leads, not clicks. A change that doubles submissions and halves quality is not a win.
- Repeat. The compounding is the point. One change is a tweak. Twelve changes over a year is a program.
What an engagement includes
A real CRO engagement starts with a tracking audit, because most sites are measuring something slightly wrong and nobody has checked. Then a behavior data setup, a prioritized backlog scored on impact, confidence, and effort, the build and launch of changes or tests, and monthly reporting that connects page level changes to lead volume and quality in the CRM.
It should also include the unsexy fixes: a form that does not route, a broken conversion event, a slow hero image. Those are repairs and they go first.
The honest caveat about traffic
Do the arithmetic before you buy a testing program. Take your monthly visits on the page, multiply by your current conversion rate, and look at the number of conversions. Now imagine a realistic improvement, say from 3 percent to 3.6 percent. If that difference is a handful of leads a month, it will sit inside normal week to week variation and no test will ever resolve it cleanly.
At low volume, skip testing and fix the obvious: speed, clarity, form length, and whether anyone follows up quickly. Those are worth doing regardless of whether you can prove them statistically.
Related: How do you decide what to test first on a CRO backlog?, How do you A/B test a landing page?, What do heatmaps and session recordings show?, and our blog post on CRO strategy.