Short answer
To hire a CRO designer, look for someone who can show A/B tests they designed, the hypothesis behind each one and the result in revenue, not only polished screens. Ask how they research, how they decide what to test, how they read results and what they do when a test loses. Avoid anyone who guarantees a lift. A small paid trial, such as one page audit and one test plan, shows more than any interview.
Key takeaways
- A CRO designer is judged on revenue per visitor. A UI designer is judged on the interface. An agency adds a team and a process.
- A strong portfolio shows tests, hypotheses and results in money, including tests that lost.
- Good answers mention research, primary metrics, sample size, guardrails and learning from losses.
- Red flags: guaranteed lifts, no testing experience, and results quoted only as conversion rate.
- A short paid trial project is the best test of fit.
What is the difference between a CRO designer, a UI designer and a CRO agency?
A CRO designer designs changes to raise conversions and proves them with A/B tests. A UI designer designs interfaces. A CRO agency sells a full team and process.
| CRO designer | UI designer | CRO agency | |
|---|---|---|---|
| Main goal | More revenue per visitor from the same traffic | A clear, consistent, usable interface | A testing program run for you |
| Typical output | Researched variants, copy, test plans, readouts | Screens, components, design systems | Research, design, development, analysis and reporting |
| How success is measured | A/B test results in revenue | Usability, consistency, stakeholder approval | Test results and the number of tests run |
| Best when | You buy traffic and the pages are the bottleneck | You are building or rebuilding a product | You want a whole team and have the budget |
| Watch out for | Needs access to data and someone to build or launch tests | Beautiful pages that are never tested | Junior staff doing the work, slower changes |
Many good UI designers can learn CRO, and many CRO designers came from UI. The question is whether the person you hire has designed for tests and read the results. For more on the role, see what a CRO designer is.
What should a CRO designer’s portfolio show?
Tests, not just screens. For each project, a strong portfolio answers four questions:
- What was the problem? A drop-off, a weak first screen, a confusing form, backed by data or research.
- What was the hypothesis? “If you change X, Y will happen, because Z.” Written before the design, not after.
- What was tested? Control and variant side by side, with what changed and why.
- What happened? The result on a primary metric, ideally revenue per visitor, with the test duration and whether it was significant.
Look for honesty too. Someone who has run many tests has had many losses and flat results. A portfolio with only huge wins, and no losses or caveats, is either very short or very selective. Some results will be under NDA, with brands hidden or numbers given as percentages. That is normal, as long as the candidate can talk through the method.
What should you ask in a CRO designer interview?
Ask questions that make the candidate explain their process, not their taste. These are the ten I would ask, with what a good answer sounds like.
| Question | What a good answer sounds like |
|---|---|
| 1. How do you decide what to test first? | Starts from data and research, then ranks ideas by expected impact, confidence and effort, for example with ICE |
| 2. Walk me through a test that lost. | Explains the hypothesis, why it lost, and what the next test was because of it, without blaming the tool |
| 3. What metric do you judge a test on? | Revenue per visitor or value per session as the primary metric, with conversion rate and quality as secondary metrics or guardrails |
| 4. How long do you run a test? | Until a sample size planned before launch is reached, in full weeks, at least two, counted from the last change |
| 5. What research do you do before designing? | Analytics, funnel drop-off, session recordings, reviews, surveys or support tickets, and a check of the ad that sends the traffic |
| 6. How do you write a hypothesis? | A specific change, an expected effect on a named metric, and a reason based on evidence |
| 7. What do you check before a test goes live? | QA on real devices, tracking parity on both versions, a short ramp, and sample ratio checks |
| 8. What do you do with a flat result? | Treats it as an answer, keeps or ships based on reasons, and considers a bolder version rather than extending the test hoping for significance |
| 9. How do you report a winner? | The lift in revenue per visitor turned into yearly revenue at current traffic, with a conservative number next to it |
| 10. What do you need from me to start? | Access to analytics and the testing tool, traffic and revenue numbers, the main ads, and who builds and launches the tests |
You don’t need to know the right answers in detail. You are listening for whether the candidate talks about evidence, metrics and learning, or only about how the page will look.
What are the red flags when hiring a CRO designer?
Promises they can’t keep, no testing experience, and results without money attached.
- Guaranteed lifts. Nobody can promise a test will win. Many don’t. A guarantee usually means the person hasn’t run many tests, or plans to measure in a way that flatters the result.
- No testing experience. Redesigns that went live without a test can’t show whether they helped. Ask which tool they used and how they read results.
- Conversion rate only. A “40% conversion lift” with no mention of order value, lead quality or revenue may have lost money.
- Huge lifts from tiny changes. Possible, but rare. Ask about sample size, duration and whether the result held after launch.
- No questions about your data. A candidate who proposes a design before asking about traffic, drop-off and the ads is guessing.
- Peeking and stopping early. If they call tests the moment they look significant, their wins may not be real. See when to call an A/B test.
What is a good trial project for a CRO designer?
A small, paid project that covers the full thinking, not just a pretty mockup. One that works well:
- Audit one important page. Give access to its analytics and the ad that drives traffic to it. Ask for the top problems, with evidence for each.
- Write three hypotheses. Each with the change, the expected effect on revenue per visitor and the reason.
- Design the strongest one. Mobile first, ready to build as a test variant.
- Write the test plan. Primary metric, guardrails, sample size, duration and what you will do if it wins, loses or comes out flat.
This takes days, not months, and shows how the person thinks. If you have the traffic, launch the test. Then you see the whole loop, including how they report the result.
What should you expect from me?
The same process as above. I am a CRO designer with six years of designing for A/B tests, more than 300 A/B tests designed, and CXL certifications in Conversion Optimization and Ecommerce Marketing. I started at Spiralyze, an A/B testing agency, led CRO and design at G-Plans, and have freelanced since 2020. I work in Figma, GA4, the common A/B testing tools and Claude, and I report every test in revenue per visitor.
For what the work usually costs and how to judge the return, see how much a CRO designer costs. If you want to talk through a page, a funnel or a testing program, book a call.
FAQ
Should I hire a CRO designer or a CRO agency?
Hire a designer when you need someone to research, design and test changes on pages you already buy traffic for, and you can build and launch tests or the designer can. Hire an agency when you want a whole team, including development and analysis, and have the budget for it.
What should a CRO designer portfolio include?
A/B tests with the problem, the hypothesis, the control and variant, and the result on a primary metric such as revenue per visitor. Losses and flat tests are a good sign, because they show real testing experience.
Is a paid trial project fair to ask for?
Yes, as long as it is paid, small and clearly scoped. Auditing one page and writing a test plan for one hypothesis is enough to see how someone works.
What is the biggest red flag when hiring for CRO?
A guaranteed lift. Many tests lose or come out flat, so anyone who promises a win either hasn’t run many tests or plans to measure in a way that flatters the result.
Sources
I’m George (Djordje) Peric, a CRO designer, CXL certified in Conversion Optimization and Ecommerce Marketing. I design ads, landing pages and funnels and A/B test them.