← All field notes Blog

What is user testing? (And why it's the missing half of CRO)

User testing is watching real, representative people actually use your site — and it's the qualitative 'why' that analytics and A/B tests can't give you. What it is, how it differs from A/B testing, the modern tools (many free), and how many users you really need.

User testing keeps coming up, so let's be clear about what it actually is — and why it's the half of conversion optimization most teams skip.

User testing is improving a website's design and experience by carefully watching real, representative people — folks who could plausibly be your customers — actually use it. Some tests are quantitative, some qualitative, some moderated and some unmoderated, but the objective is always the same: put your site in front of real humans from your target audience and watch where it helps them and where it fails them. If you were a head chef in San Francisco, you wouldn't put a brand-new dish on the menu without feeding it to a few people first. Same idea.

Why user testing is worth it

People have the attention span of a goldfish online. Watching real users shows you exactly what turns them off and what lets them get things done fast — which maps straight onto revenue.

How user testing differs from A/B testing

This is the key distinction. A/B testing is quantitative and tells you what wins — which version converts better, with statistical proof. User testing is qualitative and tells you why — what confused them, what they expected, where they hesitated. They're complementary: user testing generates the insight that becomes your A/B test hypothesis, and the A/B test proves whether the fix actually works at scale. Run only one and you're either guessing at causes or shipping changes you don't understand.

Why don't more companies do it?

  1. They assume it's expensive (it can be — but with the right tools it's nearly free).
  2. Designers, PMs and CEOs have a vision and don't want it "muddled" by user feedback.
  3. It surfaces feedback teams don't want to hear. Pull one thread and the whole sweater can unravel — which is uncomfortable but is exactly the point.

The modern toolkit (much of it free)

You don't need a research lab. Depending on the question:

Whether you run a full moderated study or just turn on Clarity for a week, something beats nothing. Put the energy in early and your site ends up reflecting your customers' needs instead of your team's assumptions — and that's more revenue coming your way. It's a core part of how we run CRO and landing pages.

Questions we get
What is user testing?

User testing is watching real, representative people use your website or product to find out where the experience helps or fails them. It can be moderated or unmoderated, qualitative or quantitative, but the goal is always to replace your team's assumptions with evidence from actual potential customers.

How is user testing different from A/B testing?

A/B testing is quantitative and tells you what wins — which version converts better, with statistical proof. User testing is qualitative and tells you why — what confused or stopped people. They work together: user testing produces the insight, the A/B test proves the fix works at scale. Doing only one leaves you either guessing at causes or shipping changes blind.

What are good user testing tools?

For session recordings and heatmaps, Microsoft Clarity (free) and Hotjar. For unmoderated remote task testing, Lyssna, Maze and UserTesting. For structured usability studies, platforms like Loop11. Plus five-second tests and on-site surveys for quick first-impression and messaging feedback. You can start meaningfully for free with Clarity alone.

How many users do you need to test?

Fewer than people expect. For qualitative usability testing, around five representative users typically surfaces the large majority of the serious problems, which is why small, frequent rounds beat one big study. Quantitative methods (like large unmoderated tests or A/B tests) need far more participants to reach statistical confidence.

Is user testing expensive?

It can be, but it doesn't have to be. Free tools like Microsoft Clarity let you watch real session recordings at no cost, and lightweight unmoderated tests are inexpensive. The cost scales with how formal and moderated you go — but even a week of session recordings usually reveals issues worth far more than they cost to find.

I
Igor Belogolovsky
CEO

Igor co-founded Clever Zebo in 2011 and has run paid acquisition for everyone from seed-stage SaaS to DTC brands. He’s allergic to vanity metrics and very fond of clean attribution.

How we'd help
enough reading —

Want this run on your account?

Say hello