The Three Pillars of Product Operations
Structure a product ops function around data insights, customer insights, and process to free PMs for strategic work
- Difficulty
- Moderate
- Time to result
- ~months to results
- Steps
- 3
- Confidence
- 95%
Product operations is best understood as three distinct capability areas rather than a catch-all support role. Business and data insights covers the quantitative side (engagement, revenue, dashboards); customer and market insights covers the qualitative side (research repositories, participant panels, sales/support feedback); and process and practices covers the operating model (roadmap templates, governance, go-to-market coordination). The goal across all three is to increase the speed and quality of decision-making so PMs spend the majority of their time on strategy rather than harvesting data off the side of their desk.
Origin
Developed by Melissa Perri and Denise Tilles in their book 'Product Operations: How Successful Companies Build Better Products at Scale.' Perri first named the concept in her 2018 book Escaping the Build Trap after building an early version at Athenahealth. She credits Blake Samic (Uber, Stripe, OpenAI) as one of the earliest practitioners pushing the discipline.
Core principles
- 01Product ops informs decisions but never takes decision-making rights away from the PM
- 02Each pillar serves a different consumer: PMs, product leaders, and executives all benefit
- 03The right pillar to invest in first depends on where your biggest fires are
- 04It liberates rather than adds overhead — it removes busywork, it is not more process for its own sake
How to run it
- 1
Build the business and data insights pillar
Stand up repeatable dashboards, ad hoc reporting, and product-lens metrics so PMs and leaders can make strategic decisions without writing SQL or fighting bureaucracy for data. Connect to existing data science/BI teams rather than reinventing them, and add a product lens (e.g. ARR by customer segment, adoption by product line) instead of overall company metrics.
Pro tip Put anything you look at repeatedly — board slides, retention cuts, OKR tracking — into a standing dashboard so it never becomes obsolete right after you build it manually.
Watch out This person is usually not a product expert and should not be inventing the questions to answer — the PM or leader supplies the questions; the analyst pulls the cuts.
- 2
Build the customer and market insights pillar
Aggregate all qualitative research into a shared findings database, build an opt-in participant panel of customers willing to do interviews/alphas/betas, and pull qualitative signal out of Salesforce, support tickets, and scattered Google Docs into one queryable place so insight isn't stuck in silos and research isn't duplicated.
Pro tip Maintain an opt-in customer panel who already know the cadence and what a beta entails — it makes PM outreach instant instead of a 50-hour scheduling effort.
Watch out This is the 'squishy middle' pillar people forget. The ops person streamlines and systematizes research — they do NOT replace the user researcher or take research away from the PM.
- 3
Build the process and practices pillar
Standardize the product operating model: consistent roadmap templates, cross-functional roadmap review cadence, portfolio roll-up tooling, OKR tracking, and go-to-market coordination. This gives leaders transparency into what teams are doing and how it ties to strategy.
Pro tip Larger enterprises and transformation companies should usually start here, because they lack the operating-model infrastructure to run good product at all.
Watch out Do not staff this with an Agile coach who has never been a PM — they revert to optimizing stand-ups and Scrum instead of product-management processes.
In the wild
With 360+ PMs and 5,000 developers, the CEO was digging through Jira trying to find what teams were working on. Perri realized he actually wanted a portfolio roadmap tying big feature pushes back to strategy. Product ops trained people to write substantial Jira epics and rolled them up into a portfolio view.
→ Executives finally got transparency into R&D allocation and roadmaps against strategic goals like moving upmarket into hospitals.
Common mistakes
Treating product ops as more overhead
The role is meant to remove busywork and free PMs, not add another process layer or another approval gate.
Forgetting the customer/market insights pillar
Teams default to data or process and neglect the qualitative middle pillar, leaving huge amounts of customer signal trapped in silos.
Is it for you?
Best for
Product leaders (CPO/VP Product) at scaling companies with many product teams who see PMs burning 20-30% of their time harvesting data instead of doing strategic work
Not ideal for
Very small startups where a single PM can still run their own queries and research without it becoming a bottleneck
From the transcript
“The way we think about it is is kind of structured around the three pillars that we talk about in the book. So, business and…”
“And finally, the third pillar is is process and practices.”
“Product operations does not take away decision-making rights from the product manager. It's there to inform them.”
“how do we monitor the health of our product? So, like as a chief product officer, AR is interesting to me, but it's not as…”
From the episode
The ultimate guide to product operations
Melissa Perri and Denise Tilles