Portfolio management is a craft multiplier

Portfolio management compounds craft in three stages: connect isolated PMs, standardize a shared language for impact through OKRs, then harvest the opportunities no single account can see, like a shared component library a client was quietly paying for several times over. The biggest opportunities live in the space between accounts, and someone has to be explicitly responsible for looking there.

I became a portfolio lead the way most people do: sideways. My early projects were delivery-heavy and complex, and I leaned on best practices and a growing interest in Product Operations to become the person who could steady a project and help more junior PMs find their footing. That turned into a portfolio role, leading the product craft across every account we ran for Google: a team that started at around ten product managers and grew to twenty-five at its peak, each embedded in their own corner of the portfolio, each solving problems the others couldn't see. Not every account had a product manager at first. By the time the portfolio matured, roughly two thirds did.

That last part is the whole story. The default state of a multi-account product team is isolation. Each PM optimizes their own roadmap, reinvents their own artifacts, and fights their own stakeholder battles. Nothing is wrong with any individual account, and yet the group as a whole is leaving most of its value on the table. Portfolio management is the discipline of collecting it. In my experience it happens in three stages, and each one unlocks the next.

First, connection. The unglamorous starting point was getting PMs talking to each other, comparing notes and showing work that could inspire another account. Risks and wins started traveling in both directions, from leadership down and from the delivery teams up. Feature opportunities started appearing that no single account could have found alone, like a sign-in capability on one content property that finally let the client measure whether its content was doing the job it was funded to do.

Second, a shared language for impact. With the connective tissue in place, we could standardize where it mattered. OKRs became the common starting point, which made roadmaps comparable and let us run quarterly intake workshops against goals everyone could align on. Prioritization got sharper and iteration conversations got more honest, because they were grounded in analytics rather than opinion. The road there was not the same on every account. Some had to push hard to get OKRs adopted at all. Others had roadmaps constantly derailed by urgent stakeholder requests because the PM wasn't yet equipped for the tradeoff conversation. That is the real portfolio job: noticing which team needs which help, rather than rolling one process across everything and calling it consistency.

Third, the opportunity nobody could see from inside an account. The realization surfaced in conversations with our tech leads: every team was building the same components from scratch. The same carousel, rebuilt on every account, because the client's teams worked in silos and our structure mirrored theirs. The proposal that followed was simple: a shared library of reusable components with one common backend and front ends themed to each product's brand. The economics of that idea are well documented now. a Forrester Total Economic Impact study commissioned by Knapsack put the estimated ROI of a unified design-to-code platform at 671%, ¹ and Salesforce's own adoption surveys reported a 41% average productivity lift for teams working in Lightning, the experience built on its design system. ² But the number that mattered in the room was blunter: the client was paying several times over for the same work, and no single account could have caught it.

That insight did more for my trajectory than any roadmap I ever shipped. It earned a seat on the account's leadership team, with visibility into where the business was heading and how the product craft could support it, and it changed the kind of work I was trusted with: vision engagements, and redesigns that needed a transformation lens. It also followed me. The themed component architecture became the backbone of a multi-brand consolidation I later led for an industrial client, and the reusability logic matured into a scoring framework in a banking platform vision. Ideas that start as efficiency plays have a habit of ending up as strategy.

The lesson I would hand to any craft lead: the biggest opportunities in a portfolio don't live inside any account. They live in the space between accounts, and someone has to be explicitly responsible for looking there. That is what portfolio management actually is: holding the strategic integrity of the whole so the parts can compound, for the client and for the people practicing the craft.