The main product for the company that I am working is an ad server. So many sprints are tech and back end focused with small tweaks to the UI. The design iterations are so slow that it may take months for a user to notice slight changes on the dashboard of the product.

Another issue that our users are publishers and advertisers (ad ops) with specific knowledge about ad servers, so it is almost impossible to use user testing platforms as our user's tech skills are very specific and also very hard to get them on board to run user tests.

Few of my colleagues proposed to do user tests for the entire journey by adding additional tasks to the user tests. I feel like this is very painful as the journey is very long and what we want to test are small style updates and features.

What your suggestions ? How often should I run user tests ?

2 Answers 2


Depends on what "slight" changes means since the word is subjective and could have a huge impact to the end user.

Brendon Attebury wrote a great article about dogfooding (internal testing) when you only have a little time and need qualitative responses quickly that I would recommend reading. Read Article ↗

I'd also create a guideline of usability for UI as it pertains to UX for the company to follow while developing, share it with stakeholders for approval (if you need to) and share it with everyone else when it's in a good place. This could help you avoid reinventing the style wheel every time which is costly. Check out Google Materials' "Material Motion" as an example.

Raise awareness with the people who suggest adding new features in order to test is adding to an already long issue. Also suggest a positive flow you'd be happy with. In order to do so, I might suggest an audit or testing the existing flow with no new features to see if there's anything that can be simplified or stripped down. Through testing the existing flow, a new feature need could emerge based on what users really want or have an issue with. Also, adding new features could be feature creep if time has not been allotted for the work. Check out one of my favorite UX stories, Jared Spool's $300 million dollar button. Read Article ↗

As long as your conversations are always in the context of keeping the company's budget in mind, working smarter and helping the user get through flows as quickly as possible, your coworkers should be relatively responsive to trying your plan out.

Hope this helps! Interested to read and learn from other responses.


When you say "Adding tasks to the user journey", does that mean user testing for the entire flow of the application? I'm not sure if there is a need to repeat again and again these kind of extensive tests. They are extremely important during the exploratory phase and when modeling a new product, but there is no need to run them again and again for each minor change. I assume that there is a pipe holding future features. Why not simply gather a bulk of such changes, generate a prototype and then devise a test that will let the user pass through all these changes at once? You don't need the entire user journey - just to test behavior with the future changes.

  • There is a roadmap for major features and updates, but some of the demands are very spontaneous ( eg : add specific filters, add more information or additional technical features that require UI tweaks ) and they come from our important clients. Some of these updates are very irrelevant for the rest of our users. I know this is horrible from UX angle but our business dictates to answer demands of these high-end clients. Nov 2, 2016 at 9:31

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.