How UX debt messes with your metrics

Do poor test results indicate bad solutions, or is UX debt making sound solutions appear unsuccessful?

You’re testing a new feature, but results come back and your key metrics have hardly moved — they may have even dropped. “This solution didn’t work? I guess we’ll roll it back and work on something else.” Sound familiar?

At DoorDash we meticulously test every new feature on our Consumer, Merchant, and Dasher products. When metrics come back lower than expected it’s tempting to write off the feature as a failure and move on. Despite a seemingly poor performance, the concept behind many solutions is actually sound. Execution that doesn’t address UX debt is often responsible for poor results.

Focusing on Discoverability, Clarity, and Viability will prevent UX Debt from making good solutions appear unsuccessful.

Discoverability: Users can’t find the new feature

The right feature in the wrong place isn’t likely to succeed.

Your new feature won’t perform well if it’s haphazardly dropped into a mess of other features. Results will be poor if users can’t find the feature, not because it’s a bad idea or can’t be successful.

Prevent UX debt by planning ahead.

Products quickly become less effective without a plan for how individual features will eventually create a streamlined system. Thinking several steps ahead will prevent your product from becoming a Frankenstein’s monster. Concept how new features will elegantly layer into existing features in v2 and v3. Define discrete build stages that all move towards the final vision.

Adding Filters to the DoorDash Consumer app. The MVP adds Filters near Sorts, although other elements are distracting and will hinder results. We’ve planned for how Filters will be increasingly elevated as other features evolve.

Clarity: Users don’t understand the new feature

Customers can’t embrace unintuitive features.

Metrics plummet when customers don’t understand how a feature works, don’t believe a feature will do what they want, or get stuck and don’t know what to do next. Users look for visual cues and familiar patterns, and may overlook features that don’t match their pre-existing mental model.

Prevent UX debt by leaning into conventions.

Creating clarity is not just good for customers, but can have a significant impact on your business. We saved $6 million at DoorDash by intuitively grouping and labeling form fields in a single application flow.

Lean into usability conventions to prevent customers from needing to relearn functionality. Add established industry standards to your MVP to ensure it meets expectations. Clear labels, instructions, descriptions, and error states make features intuitive and successful. These details can be easily overlooked when trying to quickly get a new feature out the door.

One of these fields uses conventions to create clarity around functionality, and one does not.

Viability: Users don’t trust the new feature

“It’s Apple Maps bad.” Years later, the first version of Apple Maps is still a cautionary tale.

MVPs that aren’t minimally viable erode trust.

Minimal viable products that are too narrowly-scoped won’t meet a customer’s needs. This breaks trust and creates a negative first impression that’s difficult to rewrite.

We test features in isolation, but customers experience products and brands as a whole. Lost trust builds up over time. If shown too many incomplete experiences customers may begin to doubt the reliability and quality of all future releases. Future tests won’t do as well if customers already believe new features are unlikely to meet their needs.

Prevent UX debt by only shipping complete experiences.

An MVP can be simple, but should always solve the customer’s need. Customer feedback is the best way to know what is immediately necessary and what can wait for later. A good MVP will also follow fundamental usability standards. It should include primary use cases and edge cases, have strong visual design, and use clear copywriting. This will create a good first impression, build trust in future iterations, and produce better test results.

Each iteration should meet customer needs. Customers who get a bad first impression in earlier stages won’t stick around for the stage when everything finally comes together. Image from Quora.

Move past poor test results.

When you launch a new feature and metrics barely move it can be surprising and disappointing. Don’t worry. Poor results often indicate that the execution needs work, not always that the concept is wrong. This is especially true when UX debt hasn’t been addressed before launch.

Metrics will improve as you reduce UX debt, so you should continue moving forward with confidence. Poor results from a quick execution shouldn’t turn you away from a great solution that just needs some polish.


Ryan Scott is a Senior Product Designer at DoorDash. He’s worked for Salesforce.com and Palantir Technologies, and has launched products at TechCrunch Disrupt, AWS re:Invent, and Dreamforce.

If you enjoyed this story, please recommend and share to help others find it! Feel free to leave a comment below.