There’s method in the user research and testing madness

There was a particular tweet (pictured below) that went viral last year, and we had a good laugh about it at DigitalBridge. Amazon’s algorithms had clearly seen Jac buy a toilet seat and then retargeted her a couple too many times about the other toilet seats she might need.

Tweet from @GirlFromBlupo

It’s an amusing story. However, the inability of Amazon’s algorithms to differentiate between Jac’s desire and necessity trips up many retailers each year. Retailers need to know why people make a purchase to sell more to that person or more of that product to other consumers. This might be smaller items like a toilet seat or larger items like a full bathroom suite.

At DigitalBridge we’re keen to understand why people buy, and whether there are patterns in buying behaviour. This information feeds into our product design decisions, and helps us iterate our technology to create a best in class product.

In our last blog we dived into why user research and testing is so critical, but this time we want to focus on how to we conduct our research and testing. We have four main levers to acquire the information we need:

C U S T O M E R    I N T E R V I E W S

We invite people who’ve renovated their bathroom into DigitalBridge HQ to have an informal chat. It’s important to us to speak to people who’ve gone through the process and explore the journey, rather than asking hypothetical questions. From this we gather insights and find patterns in the problems people have during their renovation process. This is key, as it means we can start to identify common problems and ensure our tool alleviates these frictions.

U S A B I L I T Y   T E S T I N G

Once we have a solution, it’s time to build but we still need to make sure our users can actually use our product. We do this by chatting to users and observing them using our product. From this we gather insights, change our designs and start the process again until we get our product just right for the majority of our users.

A / B   T E S T I N G

Working in a Build, Measure, Learn environment is key to getting features online quickly. We prefer to test two (or sometimes more) designs or features against each other to check our UI and feature designs are in the correct place or if they will be missed completely by the user. Even if the two tests you put against each other don’t work, this is still a useful learning, as you can try new places and run the test again.


It seems obvious, but surveys are a key method for understanding people, which can be too often overlooked. We run surveys and polls to gain statistically significant data. This sort of quantitative information allows us to validate the trends and insights that we discover from our qualitative research on a much larger scale.

That’s all folks! Stay tuned for our next blog, when we’ll be looking at the Jobs To Be Done framework and why we think understanding the ‘job’ a bathroom is doing is so important to creating a successful product.

In the meantime, if you’re a retailer and would like to find out more about DigitalBridge technology drop us a line on

Search Pivot