Continuous Discovery Has an Infrastructure Problem, Not a Willpower Problem

Depending on which 2026 survey you trust, somewhere between a quarter and two-thirds of product managers actually talk to customers every week.

The surveys disagree, and the disagreement is worth naming rather than papering over. One State of PM 2026 report puts weekly customer contact at just 28%; another pegs continuous-discovery practice closer to 62%. Pick whichever you like – either way, a large share of teams aren't sustaining the one practice nearly everyone in product agrees they should. The usual explanation is discipline: PMs know they should talk to users, they just don't make the time. I think that explanation is comfortable and mostly wrong.

The blockers aren't about willpower

When you look at why teams don't sustain discovery, the reasons aren't moral failings. They're missing plumbing.

The single biggest blocker in the ideaplan data is the lack of a recruitment pipeline – 44%. Not "PMs don't care." There's just no reliable way to get a qualified customer in front of them this week. The other dominant blocker, from the makeitnice survey, is time pressure: 57% say delivery commitments leave too little room for discovery. Again, structural – the system is loaded such that discovery is the thing that gives.

Tell a PM with no recruitment pipeline and a full delivery quota to "just be more disciplined about talking to users" and you've diagnosed the problem as a character flaw when it's a logistics one. They could have all the discipline in the world. Without someone to talk to and an hour to do it in, the practice still doesn't happen.

The tell: who actually sustains it

The most useful number cuts straight to the cause. Teams with dedicated research-ops support are 3x more likely to maintain weekly customer contact.

Read that as the diagnosis it is. The thing that separates teams who sustain discovery from teams who don't isn't grit – it's infrastructure. Research ops is exactly the plumbing the blockers point to: the recruitment pipeline, the scheduling, the incentives, the place insights get stored so they compound instead of evaporating. Where that infrastructure exists, discovery sticks. Where it doesn't, it depends on individual heroics, and heroics don't scale or survive a busy quarter.

The payoff is real enough to justify building it. Continuous-discovery teams are 2.4x more likely to hit their OKRs. And it's already where the market is putting its money – customer research and insight synthesis is the top product investment priority for 2026, named by 39.3% of teams. The recognition is there. What's mostly missing is treating it as a system to build rather than a habit to nag people about.

Build the system, not the resolve

If discovery is an infrastructure problem, then the fixes are infrastructure, and most of them are unglamorous on purpose.

Build the recruitment pipeline first. The 44% blocker is the one to kill early. A standing way to get qualified customers in front of the team on a predictable cadence removes the friction that willpower keeps losing to. This is plumbing, and plumbing is fundable.

Make room in the load, don't just exhort. If delivery commitments crowd out discovery, the answer is to protect the time in the plan – not to ask people to find it in margins that don't exist. Discovery that isn't on the roadmap loses to everything that is.

Give insights a home. Conversations that aren't captured and reused are conversations you pay for once and benefit from once. A shared, searchable place for what you learn is what turns scattered interviews into compounding understanding – the difference between doing discovery and having it.

You don't get continuous discovery by convincing people to want it more. Almost everyone already wants it. You get it by building the system that makes it the path of least resistance, and then it mostly takes care of itself. The teams that sustain it didn't out-discipline everyone else. They just built the plumbing and let it do the work.