Most teams collect both analytics and research, but they tend to live in different rooms. Analytics dashboards sit with the growth team. Research notes sit in a document nobody opens after the kickoff. The result is a gap: a number nobody can explain, or a story nobody can size.
What analytics is good at
Analytics is excellent at the "where" and the "how many". It shows where in a journey people leave, where they click on things that do nothing, where they bounce straight back, and how a change moves a measurable outcome. These are facts a team can act on.
What analytics misses
The reason. Analytics will never tell you that the customer left because the headline used a word they did not recognise, because the next step sat below a confusing legal notice, or because they were tired and the form felt long. It also tends to hide vulnerable customers inside an average. The people most likely to be harmed are often the people whose behaviour rarely shows up as a clean trend line.
Use the friction finder to decide where to look
The friction finder is one of the ways to set up a UXDuty audit. It uses UX signals from Microsoft Clarity or Google Analytics, with thresholds you set, to uncover the journeys where customers are struggling, with a reason against each step. Rage clicks, dead clicks and quick backs are the kind of signals it reads.
Treat the result as a shortlist of moments worth asking customers about, not as an answer. It tells you where friction might be. It cannot tell you what the customer was thinking.
Then speak to customers
For each journey on the shortlist, find the people who went through it: recent applicants, claimants, or customers who contacted support from those pages. Then:
- Ask five of them to talk you through what they did, and watch them do it.
- Read the support tickets and chat transcripts that start on those pages.
- Ask what they expected to happen, where they hesitated, and what nearly made them stop.
- Include customers in vulnerable circumstances on purpose, because they are the ones an average hides.
This is where you catch the moment somebody says "I gave up because I was not sure that was the right link." No scan can find that, and it is often the most useful thing you will learn.
What the audit adds
The same audit reads the pages on those journeys against the Consumer Duty checks and gives you findings to resolve: a buried complaints route, an exit fee that is not shown beside the rate, wording a customer will not follow. Fixing them may well ease some of the friction.
Your team verifies each finding, fixes the page, and UXDuty checks it again before the finding is marked resolved. Each rerun adds a point to the trend chart, so you can show the findings falling. But the findings are a head start, not the whole picture. Speaking to real customers will uncover a lot more than the audit can.
How they work together
Analytics finds the cliff. The audit gives you things to fix. Customers tell you what both of them missed. Then you go back to the numbers to check whether the fix moved them.
For Consumer Duty in particular, no single source is enough. Numbers do not prove harm, and quotes do not scale. You need the size of the problem and a believable account of what it cost a real person.
Back to resources