We're looking to redesign a destinations section of a travel website as the current is basically an import of our brochure and doesn't work for web. We have all the analytics, customer research videos, heatmaps etc - so a lot of research on how what we have is not working.

Coming further into the project, I'm wondering would we need to map out the customer journey, when destinations are only 10% of the overall customer journey of the website? Would it be feasible to start moving into the sketching phase of the design process and conduct usability tests on the new, potential design going forward?

What would be the best practice in this situation?

4 Answers 4


In your situation, I'd start by defining the user goals for this section of your site.

Once I had those, I'd work out what sub-journeys were strictly necessary to meet those goals e.g. the checkout process on an ecommerce website.

Then I'd work out what sub-journeys most users would have to take in order to meet the site's user goals.

Once I'd done that, I'd look through the analytics to work out which of the necessary journeys were broken for actual users. Then I'd work out which of the "most users need to take" journeys were broken.

Then I'd sketch out solutions to the problems in the broken journeys, and go into the normal iterative design -> test -> deploy -> analyse -> redesign process.

There's little point in mapping out literally everything. When you only have the time and resource to fix 10 things the point of this initial investigation stage is to work out which 10 things to focus on fixing.

If it turns out that after fixing the first 10 things, you've got resource to fix another 10 things, start the whole process from the beginning, and fix the new 10 most broken things.

Stephen Krug's Rocket Surgery Made Easy is a good book that touches on this problem. It's worth the 3 - 4 hours it takes to read the whole thing.


Focus on what matters

This is a travel website and your Destinations pages are only 10% of traffic? That seems suspicious. Cross-referencing analytics and user testing you should be able to identify why that is fairly quickly. You'll have to ask yourself a lot of questions about how visitors move specifically to, from, or around that area.

Are you landing too many visitors on the root page rather than the destination they were searching for?
Are you driving visitors straight to a phone number or form before they've seen your inventory?
Is there another step in the journey that derails them before they find your Destinations?
And a whole lot more …

Armed with that understanding, you'll know if Destinations even matter (as I suspect they do).

Assuming Destinations matter …

When you find a critical "sub-path" like browsing the things or services you sell, it's worth digging deeper to find the hidden plot lines. Dive into Destinations with a microscope:

  • Slice your analytics in every reasonable way to identify patterns
  • Draft what you think the customer journey(s) looks like
  • Test your existing product with target users
  • Reevaluate your journey maps
  • Kick off a design sprint to create something big

Mapping to drive rapid prototyping

Developing a perfect map is not the holy grail. You just need a good map and the right people to help you find the opportunities in it.

Rapid prototyping process

If you follow the GV sprint format, you'll leverage your understanding of the customer's journey (always a critical baseline) to quickly zero in on what you think the right solution is.

The key is rapidly prototyping something you think will hit your OKRs. Testing that assumption with real live users in a very short span of time will tell you what you need to do next.


A Heuristic map of the flow would be a good idea, at a high level.

Combining it with your data can help you create 'How Might We' statements and pin them on the map to see if you're targeting the user problem and correcting the journey.

It would be feasible to start moving into the sketching phase of the design process and conduct usability tests on the new, potential design going forward, but if you have bandwidth and time, I don't think it would hurt.


In my opinion I would like you to map out user journeys and then move on to sketching, Because you will encounter many new/old things that can be included into the journey which otherwise would be difficult in incorporating in already made wireframes.

In my perspective 10% is a considerable amount, If this is not in your case still go with this approach but you may want to pay less attention as per your requirement.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.