Usability testing a travel website: what testers told us

Testers on a travel website we work with knew who it was for within seconds. What they could not tell was what they would get. That gap, not the design, is what held them back. We ran 3 rounds of usability tests to find it, then used A/B tests to check what the testers told us.

How we ran the tests

The site sells extended stays abroad to an older audience. We tested it 3 times: a clickable prototype, a concept page with pricing, and the live site. Each tester matched the audience the site was built for.

Every session followed our usual standard. Each tester completed 3 to 5 realistic tasks, such as finding a location and asking for more information, and talked through what they were thinking as they went. Sessions ran about 15 to 18 minutes. That is long enough to hear real reasoning, and short enough to avoid fatigue.

Talking out loud is the key. Analytics shows you where visitors leave. Hearing someone say “I’m still not sure what you’re offering” tells you why.

Testers knew who the site was for within seconds

Nearly every tester named the audience right away: older travelers. The photos did most of that work. Testers pointed to the people in the images and said the site was meant for people their age. Several said they liked seeing real photos instead of clip art. 1 tester flagged an image that looked computer generated and asked for a real photo instead.

Testers also liked the look of the site. They called it bright, easy to read and easy to navigate. If we had stopped at first impressions, the site would have passed.

Testers could not tell what the service was

Ask testers what the site did, and the answers split. Testers guessed it was an expat community, a short-term rental site, a travel agency or a kind of timeshare. 1 tester said outright that she was not sure what the site was about. Another, a self-described frequent traveler, said she would love to do this but still could not tell what was on offer.

A clear audience is not the same as a clear offer. The headline told testers who belonged. It did not tell them what they would get.

Value questions held testers back more than design did

The biggest blocker in every round was value. Testers asked the same questions again and again:

• What does the fee give me that I cannot find for free online?
• Is airfare included? Are the listed activities included, or just available?
• Does the price cover the whole apartment, and is there staff on site?
• Does this work for a single traveler, or only for couples?
• What does the asterisk on the visa note mean?

Several testers rated the site 3 out of 5 even after praising how it looked. Their reason was the same: they needed more detail before they would pay. 1 tester said she would sign up in a heartbeat if she understood what was included.

Visitors wanted answers before they gave their details

Testers listed what they needed before booking: total cost, cancellation policy, whether English is spoken, transport, and visa rules. 1 tester asked to get more information without entering personal details. Another was met by a pop-up asking for her details the moment she arrived.

Our A/B tests on the same site point the same way. Paid campaign traffic used to land on the homepage. A focused landing page, built around locations, star ratings and a single button, lifted leads 17.8%. The version that led with the lead form did not match it.

On a landing page for solo travelers, links used to send visitors straight into the booking app. Pointing them to location pages first, where they could see what a stay in each place looks like, lifted leads 14.6% over 62 days.

And on the location pages themselves, moving the lead form above the details did not help. Leads measured -3.6% over 13 days, with no positive trend, so we paused the test. Visitors preferred to read about the destination first.

Testers wanted proof from past guests

For a high-priced trip with strangers, testers wanted to hear from people who had already gone. They asked to read reviews or talk to past guests before booking. 1 tester saw a review site rating and wanted to know who had written the reviews. For more on what kinds of proof work, see Which social proof examples actually lift conversion?

What we would change first

• Say in the headline what the service is, not just who it is for.
• List what is included next to every price, and what costs extra.
• Answer cost, cancellation and visa questions before asking for details.
• Offer more information without requiring personal details.
• Show reviews from real guests, and say where they come from.
• A/B test each change, because testers tell you what to try, not what will win.

For how A/B tests performed across the rest of this site, see How travel websites earn more leads: what our A/B tests show. For usability testing in general, see Usability Testing for Better UX and Higher Conversions.

Testers did not doubt the trip. They doubted what they would get.

Frequently asked questions

How do you usability test a travel website?

Recruit testers who match your audience, give each 3 to 5 realistic tasks, such as finding a destination and asking for information, and have them talk through their thinking. Sessions of about 15 to 18 minutes work well.

What do travel website visitors need before they book?

In our tests: total cost, what is included, cancellation policy, visa rules, and proof from people who have already gone. Testers wanted these answers before giving their details.

Can you usability test a prototype?

Yes. Our first round used a clickable prototype. It showed how testers read the audience and the offer, even though some links did not work yet.

Why do visitors like a travel site but still not convert?

Because liking the look is not the same as understanding the offer. Testers praised the design, then rated the site 3 out of 5 because they could not tell what was included.

Should usability findings replace A/B testing?

No. Usability testing tells you what to try. A/B testing tells you what works. Our A/B tests confirmed visitors wanted destination details before the lead form.