Customer research is a weekly discipline

If you only speak to customers occasionally, you are not staying close to the truth. You are checking in on it from time to time. Strong product teams keep a weekly pulse on their customers, use a stable quarterly research spine to compare patterns over time, and treat single interviews as clues until repetition turns them into evidence.

One of the most common weaknesses I see in product teams is not a lack of intelligence. It is a lack of contact.

They care about the customer in theory. They discuss customer problems in planning. They reference support tickets, analytics and stakeholder anecdotes. But they are not actually talking to customers often enough to maintain a real, current feel for the problem space.

That is a mistake.

Teresa Torres defines continuous discovery as at a minimum weekly touchpoints with customers by the team that's building the product where they conduct small research activities in pursuit of a desired product outcome. That is a useful minimum because it turns customer understanding from a project phase into an operating rhythm.

The point is not to conduct giant studies all the time. It is to avoid going stale.

“If you only speak to customers occasionally, you are checking in on the truth from time to time. You are not staying close to it.”

Occasional research is usually too late

Many companies still treat research as something they do at the beginning of a project, or when a major initiative needs validating, or when things have already gone wrong.

That cadence is usually too slow.

By the time the team realises it has drifted away from the customer, it has often already spent weeks or months compounding the error through product decisions, design choices, delivery effort and stakeholder confidence. A weekly rhythm prevents that drift from becoming expensive.

NN/g's guide to UX research methods makes the broader point well: most projects benefit from multiple methods used over time, rather than relying on one familiar research motion or one upfront phase. That is exactly why weekly cadence matters. It keeps research from becoming ceremonial.

Use a stable quarterly spine

The mistake some teams make, once they adopt a more regular cadence, is treating every week like a brand-new research question.

That creates noise.

A much better approach is to use a stable script for the quarter. Keep the core questions steady enough that patterns can accumulate, while allowing room for topical follow-ups as the product evolves. This gives the team something precious: comparability over time.

A quarterly research spine should help you answer a few enduring questions:

• What is getting harder for customers right now?
• What are they trying to achieve?
• What workarounds have they invented?
• Where are they losing time, confidence or trust?
• What did they try recently that disappointed them?
• What still feels confusing or too manual?
• What would make this journey materially easier?

That does not replace more specific usability tasks or concept tests. It gives them context.

Without a stable spine, research often becomes a collection of interesting conversations. With a stable spine, it becomes a learning system.

“Use a stable quarterly script so patterns can accumulate instead of dissolving into interesting anecdotes.”

Mix discovery interviews with weekly usability testing

A good weekly cadence should not rely on one method alone.

NN/g's framework is useful here because it reminds teams that different methods answer different questions. Interviews help you understand experiences, needs and language. Usability testing helps you see whether a design or workflow actually works in practice. Behavioural and attitudinal methods are complementary, not interchangeable.

So in practice, a weekly cadence should usually include both:

• Exploratory interviews to stay close to problems, workflows, frustrations and unmet needs
• Usability testing to see whether the current or proposed solution is clear, usable and effective

That combination matters.

If you only interview, you may understand the problem but fail to see where your solution still breaks. If you only usability-test, you may optimise an interface without fully understanding whether the underlying problem is still the right one.

The best teams keep both muscles active.

One person is a clue, not a conclusion

This is where many teams need more discipline.

A single customer saying something does not automatically make it strategy. One interview can be revealing, provocative or directionally important. It is still one data point.

NN/g's 2025 guidance on analysing qualitative usability data is very helpful here. It explicitly says that a data point should not be taken at face value and should be judged in context, using factors such as consistency and repetition. In other words, you are not looking merely for something interesting. You are looking for something that repeats cleanly enough to deserve action.

That is why I like working in rolling batches of 5–7 for weekly qualitative work.

For usability testing, NN/g's long-standing guidance remains that 5 participants per user group will usually uncover the majority of the most important issues, and that the better investment is to run repeated small studies rather than one giant one.

For open-ended interviews, the logic is slightly different. NN/g notes that five interviews are often not enough for broad interview-based studies because the variability in people's experiences is higher; in customer-needs work, saturation may require more like 20–30 interviews over time.

Those two ideas fit together quite neatly:

• Use 5–7 as the practical weekly operating batch for rapid qualitative work
• Treat each batch as directional, not definitive
• Let patterns build across the quarter before declaring you "know" the customer
• Expect broader need-finding to require more cumulative interviews than a usability check

That is a much more serious way to work than canonising the loudest anecdote.

“One person is a clue, not a conclusion.”

Research gets stronger when it compounds

Another reason a weekly cadence matters is that research strength does not come only from sample size. It comes from diversity and accumulation.

NN/g argues that usability findings drawn from a broad base of diverse studies are more credible than conclusions drawn from many users in a single narrow study. That is a useful correction to the endless obsession with one-off sample sizes.

It means that a team which runs:

• Five usability sessions this week
• Five more after the next design change
• Five customer interviews the following week
• Another round with a different segment later in the month

…is often building stronger practical evidence than a team that runs one big study and then goes dark for a quarter.

The compound effect matters.

Weekly research does not just keep the team informed. It keeps the team calibrated.

The team building the product should hear customers directly

This is another important point in Torres's definition: the weekly touchpoint should be by the team that is building the product.

That matters because second-hand research always loses something.

A summary may preserve the conclusion and lose the texture. A slide may preserve the quote and lose the hesitation. A dashboard may preserve the trend and lose the context.

When product managers, designers and engineers hear customers directly, the product conversation changes. It becomes less abstract, less political and less dependent on internal storytelling. This is one of the reasons continuous discovery works so well when the trio is actually involved: it keeps customer truth from being over-processed on its way into the product.

A weekly rhythm that actually works

The biggest objection people raise to weekly research is usually practical: we do not have a full research team; we do not have time; we cannot recruit at that pace.

That is exactly why the rhythm needs to stay small.

A workable weekly cadence looks something like this:

This is sustainable. More importantly, it is strategically safer. Because the alternative is usually not “perfect research later”. It is building on stale assumptions now.

The goal is not more research. It is less guessing.

That is the key point.

Weekly research is not about turning the company into a research bureaucracy. It is about reducing the amount of guessing in everyday product decisions. It gives the team a current feel for the customer, a running check on usability, and enough repeated contact that genuine patterns start to separate themselves from noise.

A single interview can be useful. A repeated cadence becomes intelligent. A quarter of repeated cadence becomes an asset.

That is when research stops being a task and becomes part of how the company thinks.

And that is the standard more teams should aim for: not customer research as an occasional event, but customer research as a weekly discipline.

“Five users is often enough to uncover most usability issues. It is not enough to declare that you understand all customer needs.”

“The goal is not more research. It is less guessing.”