50 shades of … blue

50 shades of blue - a color chart and paint brush

How you feel or think about a design does not matter. It’s the results it delivers that truly matters. That’s the general view that Marissa Meyer took when she helped Google make an extra $200m in revenue in a single year by experimenting which shade of blue created the best outcome for Google Ads.

It might be an old story from 2014, but it’s still relevant today. A lot of time is spent in design where people debate how things look, how they feel and what others think and feel about them. But, the amount of time spent looking at performance data to figure out what is actually better still seems to be a generally missing component in many development teams and organizations.

It might be hard to put the right A/B testing infrastructure in place, but the returns can be huge. And it allows teams to use data to determine what the right implementation solution can be.

Debates about design and feelings are best solved by assessing the outcome data. It might leave designers in an uncomfortable place, but the upside is worth it.


How might you use data and experiments to inform design debates you, your team and stakeholders have?