Blog 04 · data triangulation in UX research

When your data disagrees, you have just found the interesting part

Pranjal Chawan
Pranjal ChawanUX leader · LinkedIn

The survey says customers are happy. Analytics says they are leaving. Sales says the product is too complicated. Support says one button is broken.

Most teams treat this as a mess. They pick the source that suits the plan, and they move on.

I treat it as the moment research starts paying for itself.

What is data triangulation?

Nielsen Norman Group defines triangulation as using multiple sources of data or multiple approaches to analysing data to enhance the credibility of a study. Its logic is simple. Every research method has limits. A small usability study may lack statistical proof. Analytics may lack context and meaning. Use them together, and each covers the other's blind spot.

NN/g also asks a question I wish more leaders heard: not how much time can we spend on research, but how much risk of failure are we willing to accept?

The three signals I look for

SignalQuestion it answersExamples
BehaviourWhat do people do?Analytics, funnels, session recordings
ReasonsWhy do they do it?Interviews, contextual inquiry, usability tests
Business contextWhy does it matter?Revenue, support volume, stakeholder goals, competitive benchmarking

When all three agree, you can move with confidence. When they disagree, slow down and find out why.

Conflict is a clue

Disagreement between sources usually means one of four things. You may be looking at different audiences. People may say one thing and do another. The metric may measure the wrong thing. Or an assumption may have shaped the question.

That last one hit my own team. On a Fixed Deposit redesign, discussions with stakeholders pointed to customers aged 35 to 50. The first design failed in testing. Behaviour and data from real app users revealed that the most likely customers were 25 to 30. If we had defended the original plan, we would have kept polishing the wrong journey. Instead, daily bookings rose from 5,000 to nearly 8,000.

From findings to strategy

A pile of insights is not a strategy. Here is how I close the gap:

  1. Frame the decision that leaders need to make.
  2. Map each source to what it can and cannot prove.
  3. Write what all sources agree on, in one sentence.
  4. Name what is still unknown, and how risky that is.
  5. Recommend one direction, with the evidence that would change your mind.

The fourth step builds more trust than people expect. Saying "we don't know this yet" shows that the rest of your work can be believed.

Match the effort to the decision

A reversible button change needs a few hours of checking. A full product redesign deserves mixed methods and an independent view. That is NN/g's guidance too, and it keeps research fast enough to live in a real product team.

Questions people ask

What is an example of triangulation in UX? Analytics shows a high error rate on a feature, so you check support records and run a usability test to find the cause.

How many research methods do you need? There is no fixed number. Choose methods with different blind spots, and match the effort to the size of the decision.

Got conflicting data?

Good. Send it my way. I would be happy to talk through what it might be telling you at pranjal.productmind@gmail.com or on LinkedIn.

© 2026 Pranjal Chawan. All rights reserved.