Five expert tips for live website testing

Ever wondered how real people actually use your website? You might think you know, but you'd be surprised. While analytics can tell you what's happening on your site, nothing beats watching real users navigate around – it helps you spot and fix problems before they impact your business.

We’ve been chatting with UX professionals and digital leaders at companies like Wrike, Clario, and Looka about their approach to live website testing. They've shared some great tips that we think you'll find useful, whether you’re just getting started or looking to improve your current process.

Let's dive into five key strategies that will help you get more value from your testing.

1. Get clear on your goals

Okay, this might sound obvious, but stick with us – having a clear plan before you start testing can make all the difference. Instead of making random changes to your site, you want to focus on things that actually matter to your business.

Take Stanislav Khilobochenko, VP of Customer Services at Clario, for example. His team had a specific problem to solve: “At Clario, we faced a challenge with our account creation process. Analytics demonstrated 15% of users dropped off before completing it.” Instead of guessing what might help, they set up a test to compare two different solutions.

Here’s what they did:

• They ran an A/B test with two different versions.

• Group A got a streamlined form with fewer fields.

• Group B got the original account creation version, but with helpful tooltips along the way.

• They tested this with 1,000 users over two weeks.

The results? Pretty interesting actually. The simpler form reduced drop-offs by 14%, while the version with tooltips wasn’t far behind at 12%. But here's the cool part – they also noticed that Group A, the people who got the simpler form, spent more time exploring the product afterward. They wouldn't have known that if they hadn’t been measuring multiple things.

So what can we learn from this? A few key things:

• Start by understanding your current situation (like Clario's 15% drop-off rate).

• Test different approaches rather than assuming you know the best solution.

• Keep an eye on more than just your main metric – you might find unexpected wins.

• Make sure you test with enough people to trust your results.

As Stanislav puts it, remember to:

2. Don’t forget to test on different devices

Here’s something we all know but sometimes forget – your website needs to work everywhere, not just on your own computer. A site that looks perfect on your fancy desktop setup might be a nightmare to use on someone’s old phone with a spotty internet connection.

Dave Reeder, Founding Director of Made Simple Media, knows this all too well. As he puts it:

So what's the solution? Dave's team got creative: "As a practical solution, we have purchased a few cheap devices, such as a used Samsung Android phone, so we can test in real-time as a genuine user would be seeing the website, which has occasionally shown issues that weren't visible on the emulators.”

When you're testing across different devices, make sure you’re checking on:

• Real mobile devices, not just your browser's mobile view.

• Different browsers (e.g. Chrome, Safari, Firefox).

• Various internet speeds.

• Different operating systems.

Yes, this takes more time and effort than just checking things on your own computer. But think about it – your users aren’t all sitting at a desk with a perfect internet connection. They might be on their phone on the train, or using an old laptop at a coffee shop. Making sure your site works well in all these situations is worth the extra effort.

3. Create tasks that feel natural

When you’re testing with real users, you want to see how they actually use your site – not how they perform in an artificial test environment. As Elisa Montanari, Head of Organic Growth at Wrike, puts it:

Think big picture here. Instead of focusing on individual elements like button colors (which might be better suited to a preference test or design survey), look at the whole journey. For example, don’t just test your checkout page – watch how someone gets from browsing products all the way through to completing their purchase.

Joe Formica, Design Advocate at Lyssna, emphasizes the importance of thoughtful task design: “Take the time to carefully write your questions and tasks. Make sure your instructions are super clear and that you're not using any leading or biased language that could influence the results.”

He also recommends keeping things focused: “Don't overload users with too many tasks – this keeps the study focused and prevents confusion. Plus, smaller, more focused tests are much easier to analyze and turn into actionable improvements.”

A clever trick Joe suggests is asking about expectations: “Ask users what they expect to happen before they interact with a feature. For example, ‘What would you expect this button to do?’ or ‘What would you expect to see on this page?’ This helps you get unbiased feedback and can reveal gaps between user expectations and the actual experience.”

This approach really works. Just ask Branko Dimitrijevic from DesignRush: “During a live test for a retail client, we noticed users struggling with a filter feature. By monitoring real-time heatmaps and gathering feedback, we quickly identified that unclear labels were causing confusion. We resolved this by refining the language and retesting, which led to a 15% increase in conversions.”

After the test, don’t forget to dig deeper with some relevant follow-up questions. As Elisa suggests, "Ask questions that go beyond a simple yes or no so that users have to demonstrate understanding.”

4. Observe real people using your website

Let’s talk about something every UX professional agrees on – nothing beats watching real people use your website. Sure, technical tools are great, but they can't show you why someone gave up halfway through a purchase or got confused by your navigation.

Joe Formica explains why this is so crucial: “Real people are the ones using your website, so it just makes sense to test with them. When you observe how real users navigate and interact with your site, you get actionable insights that are tough to get from other methods. Watching where they click and hearing their feedback can uncover major pain points and opportunities for improvement."

There’s also a key difference between testing prototypes and live websites. As Joe points out:

A common challenge with live testing is finding the right participants. But Joe suggests a practical approach: “Focus on quality over quantity. A smaller group of highly targeted participants can give you way more valuable insights than casting a wide net.” He also recommends building what he calls a “user garden” – keeping a database of past participants who have provided good feedback, so you're not starting from scratch with each new test.

Remember, you’re not trying to prove your website is perfect – you're trying to understand how real people use it. The more targeted your participants and the more natural the testing process feels, the more valuable insights you'll get.

5. Keep testing and improving

Here's the thing about website testing – it's not about making one big change and calling it a day. It's about making small, smart improvements over time.

Dawson Whitfield, CEO of Looka, has a smart approach to this: “One of my top strategies for successful live website testing is to segment your audience and test incrementally. At Looka, we don't roll out changes across our entire site at once. Instead, we use A/B testing tools like Google Optimize to introduce updates to a small percentage of users first. This allows us to gauge the impact of changes without disrupting the experience for our entire user base."

This careful approach pays off. Max Tang, CMO at GEEKOM, shares a great example: “We examined data indicating cart abandonment spikes during the shipping stage. By testing a redesigned 2-step checkout process that combined shipping and payment, we reduced abandonment by 15% and increased average order value by 8%.”

Want to make iteration work for you? Here's what to keep in mind:

• Start small (no need to change everything at once).

• Keep an eye on your results in real time.

• Be ready to change course if something’s not working.

• Learn from both wins and losses.

• Use what you learn to plan your next test.

Remember, every test teaches you something useful about your users – even if it’s not what you expected to learn.

Getting started with live website testing

Feeling a bit overwhelmed? Don't worry – everyone starts somewhere. As Marin Cristian-Ovidiu, CEO of Online Games puts it: “Do not wait for everything to be perfect before testing. Perfection is boring and slow. Do the test, make mistakes – and learn fast. Every problem you find is a chance to get better.”

The key is to be methodical about it: start with clear goals, check your website on different devices, create natural tasks, observe real user interactions, and keep improving based on what you learn. Each test you run helps you understand your users better and create a website that really works for them.

Remember, website testing isn't a one-off task – it's an ongoing journey of discovering what works best for your users. The most successful teams are those that make testing a regular habit, using what they learn to guide their next steps.

Ready to test your website? Our live website testing templates are a good place to start:

• Test the online shopping experience

• Test navigation on your SaaS website

• Compare competitor product pages