Google Analytics

A/Bsolutely fabulous testing - a pointer

Sticky notes showing A/B test hypothesis

I'm posting a link to Tim Leighton-Boyce's recent post on the Inside GOV.UK blog. It's an excellent example of engaging the whole agile team with data, and working closely with user researchers to validate that behaviours observed in usability testing are demonstrated at …

Improving your average: using R to beef up Average Time on Page

Comparing the calculated median and the Google Analytics mean average

Google Analytics’s Average Time on Page can be a hugely important metric but also a very misleading one. As a mean average, it can be influenced by a skewed distribution, a problem that you can’t diagnose in the web interface. So I found a way to break down the data into individual visit lengths.