Closing the feedback loop

There's a famous study in consumer research where participants were shown identical pairs of jeans and asked to choose their favourite. They overwhelmingly picked one pair and gave confident, detailed explanations for why: the stitching, the colour, the fabric quality. The jeans were identical.

The reasons they gave had nothing to do with how they chose. This is the say/do gap, and it's everywhere in product feedback. Ask customers what they want and they'll tell you one thing.

Watch how they use your product, where they struggle and what they skip, and you'll learn something completely different. The best feedback strategies combine asking with watching.

Listening to customers is passive. Collecting feedback takes deliberate effort.

That means having dedicated channels: regular customer discovery calls, open feedback calendars where customers can book time directly with product, and customer success teams trained to capture insights systematically.

Every signal should be recorded in a central system and tagged against themes or roadmap items. When feedback is scattered across Slack messages, email threads and meeting notes, the useful parts get lost in the noise.

The most valuable feedback often comes from watching what customers do. Session recordings, usage analytics, drop-off funnels and support ticket patterns show you where the product is failing in ways no survey will.

When a customer says "the product is fine" but their usage data shows they abandon a key workflow 40% of the time, the data is telling you the truth.

Build your feedback system around both qualitative input (what people tell you) and quantitative observation (what people do). The most useful insights are usually in the gap between the two.

Not all feedback is equally useful. "Make it better" is low quality. "When I try to do X, I can't because Y, and that costs me Z" is high quality.

The quality of feedback you receive depends entirely on the questions you ask and the environment you create. Teach your customer-facing teams to dig deeper by asking "what are you trying to achieve?" instead of "what do you want us to build?"

When customers describe their needs as problems rather than solutions, the feedback becomes far more useful for product decisions.

Feedback becomes actionable when it's aggregated and connected to product decisions. Individual requests are noise. Patterns across multiple customers are signal.

When the same problem comes up from five different sources in different words, that's a roadmap item, and it earns its place on the evidence.

The discipline is in reading and reviewing every piece of feedback, linking it to existing themes, and letting the weight of evidence guide prioritisation, however loud any single voice is.

The most neglected step: telling customers what you did with their feedback. "You told us X was a problem. We built Y. Here's how it works."

It's good manners, and it also encourages the behaviour you want. Customers who see their input reflected in the product give better feedback next time and become more invested in the relationship.

The reverse is also true: customers who feel their feedback disappears into a void eventually stop giving it, and you lose the signal entirely. Closing the loop turns a one-off conversation into an ongoing cycle of feedback and improvement.