I tracked my time for a month when I was a PM at a mid-stage startup. The results were embarrassing.
40% — Meetings. Standups, sprint planning, backlog grooming, stakeholder syncs, design reviews, engineering check-ins. Most of these could have been async messages.
25% — Writing and updating tickets. Acceptance criteria, edge cases, answering questions in comments. JIRA was my most-used app by a wide margin.
15% — Status communication. Roadmap updates, progress reports, Slack threads explaining what's shipping and why. Keeping everyone "aligned."
12% — Reactive work. Escalations, urgent bugs, sales requests, that thing the CEO saw and wants an opinion on by EOD.
8% — Actually understanding customers and making product decisions.
Eight percent. The one thing that determines whether you build something people want — and I was spending a half-day a week on it. Everything else was overhead disguised as process.
The process trap
Here's how it happens. You join a team. The team has a process. Sprint planning every other Monday. Backlog grooming Wednesdays. Standup daily. Design sync Thursday. Stakeholder update Friday.
Each meeting is individually reasonable. Together they consume your calendar. But no one questions the aggregate because each piece seems necessary. "We need alignment." "Engineering needs context." "Leadership wants visibility."
Nobody asks: visibility into what? If the PM isn't spending time understanding customers, there's nothing meaningful to be visible about. You're just reporting activity, not insight.
Where PM time should go
The highest-leverage PM activities, in order:
1. Customer contact (target: 30% of time)
Interviews. Shadowing sessions. Reading support tickets. Sitting with sales on calls. Analyzing usage data. Every hour here directly improves the quality of every downstream decision.
The customer interview synthesis framework makes this time productive rather than just "talking to users." Structure turns conversations into evidence.
2. Analysis and prioritization (target: 25% of time)
Synthesizing what you learned into themes. Assessing severity. Mapping evidence to potential features. Running the prioritization framework against your options. This is the thinking work. It's where insight becomes strategy.
3. Defining the work (target: 20% of time)
Writing specs, yes — but specs that encode customer context, not just acceptance criteria. The difference between "add a search bar" and "users with 50+ documents can't find what they need — here's the evidence, here's the target metric, here are three approaches with trade-offs."
4. Alignment (target: 15% of time)
Some meetings are necessary. Stakeholder trust matters. But most alignment should be async: a weekly written update, a shared evidence board, a clear roadmap rationale doc. If people keep asking "why are we building this?" the problem isn't insufficient meetings — it's insufficient evidence communication.
5. Everything else (target: 10% of time)
Admin. Standups that could be Slack messages. Ticket hygiene. The stuff that doesn't compound.
Why the inversion persists
Three reasons PMs spend 80% of their time on low-leverage work:
It feels productive. Clearing a JIRA backlog is satisfying. Writing twenty tickets feels like output. Customer interviews feel ambiguous and slow. The high-leverage work is harder and less immediately rewarding.
It's expected. Organizations measure PMs on delivery, not discovery. "How many features shipped?" is a common performance question. "How well do you understand the customer?" is not. The incentives are backward.
It's safer. Talking to customers might surface uncomfortable truths. Maybe the thing you've been building for two months doesn't solve a real problem. Maybe the roadmap needs to change. Process work lets you stay busy without confronting those questions.
How to reclaim your time
Audit your calendar
Block one week. After every 30-minute block, write one word: customer, analysis, alignment, tickets, reactive. At the end of the week, tally the categories. The numbers will surprise you.
Cut one meeting
Find the recurring meeting with the lowest information density. The one where you sit through 45 minutes to deliver or receive 3 minutes of relevant content. Cancel it. Replace it with a shared doc. See if anyone complains. They usually don't.
Block customer time first
Put customer work on the calendar before anything else. Two mornings a week for interviews, ticket reviews, or usage analysis. Protect them the way you'd protect a meeting with your CEO. If customer understanding is the highest-leverage activity, treat it like one.
Automate the synthesis
The reason customer research feels slow is that the synthesis step is painful. You talk to twelve people, take pages of notes, and then... what? The notes sit in a doc. The insights live in your head. The team hears a secondhand summary.
Tools like Mimir automate the extraction-to-insight pipeline — turning raw conversations into structured themes with evidence. Whether you use tooling or a spreadsheet, the goal is the same: make it fast enough that you actually do it regularly.
The compounding effect
A PM who spends 30% of their time understanding customers makes better decisions in the other 70%. Their specs are more precise because they know the actual workflow. Their prioritization is more defensible because it's grounded in evidence. Their stakeholder conversations are shorter because the reasoning is clear.
A PM who spends 8% on customer understanding makes mediocre decisions for the remaining 92% — and needs more meetings to compensate for the lack of conviction.
The math is simple. The hard part is saying no to the process that fills your calendar. But your job isn't to run the process. Your job is to figure out what to build. Everything else is overhead.