Boulders, Pebbles, and Sand
What I finally understood about product portfolios
read time 3 minutes
About four years ago, someone on the product team asked our VP of Product for e-commerce at Walmart a question during all-hands: “How should I think about my roadmap this year? What does good actually look like?”
The answer came back as an analogy: a few boulders, a handful of pebbles, and a bit of sand.
It sounded interesting, but abstract. Over time, it became one of the most practical ways I’ve learned to think about product work. A portfolio isn’t about doing more, it’s about carrying the right weight.
You can’t carry many boulders, one or two is usually enough. These are the big bets that drive step change in a product or how users interact with it. For me, those boulders looked like building a generative AI search experience, an algo-driven new arrivals page, complete-the-look features in fashion, and now hybrid agentic AI workflows in enterprise software.
Pebbles show up differently. They aren’t breaking changes, but they still matter. They improve the product in ways users feel every day. They rarely headline a roadmap or keynote, but they reduce friction, sharpen experiences, and still move metrics and revenue in meaningful ways.
Sand is the foundation. Quality work, tech debt, fixing small bugs before they turn into large ones. This work doesn’t always map cleanly to business impact, but without it, nothing else holds. It keeps systems healthy enough for bigger bets to exist at all.
How I started seeing it
On search and discovery, the boulder was a new generative AI search experience. Around it were things like intent classifiers that needed to work correctly. Beneath that, we had to improve how the system handled synonyms and units of measurement like “oz” versus “ounces,” while also maintaining the core search experience and retiring old ranking signals so everything actually felt better to use.
On personalization, the boulder was a fashion shopping experience that understood size and fit, a multi-quarter, multi-team effort unique to Walmart. Alongside it were complete-the-look features and other enhancements on item pages. Then came dozens of smaller UI improvements that made the experience feel more premium. Together, it all compounded. Even when one piece slipped, the overall momentum stayed intact.
Now at Adobe, I’m working on enterprise agents and hybrid AI experiences. At the same time, I still care about and work on core platform problems, migrations, quality issues, and the smaller work that keeps a large system usable and extensible.
Why this matters
Different people care about different rocks.
Executives want to understand and evangelize the boulders. Managers care about how everything connects. Meanwhile, teams feel the sand most acutely because they live with quality issues and technical debt every day.
If you only have boulders, you’re fragile. If you only have pebbles, you’re forgettable. If you’re drowning in sand, you’re busy but invisible.
The job is learning how to balance all three, knowing which one needs the most attention when, and telling the story at the right altitude, at the right time.
This feels especially relevant now. A lot of attention is on flashy boulders, often big AI bets. There’s real demand and value there, but it doesn’t mean the other work stops mattering. Pebbles and sand are still where users feel consistency, trust, and everyday delight (especially in large product spaces like enterprise platforms or e-commerce), while the bigger vision takes shape.
I’ve seen boulders fall off, and when all you have are boulders, you neither delight users nor change their experience in a lasting way.
Every PM ends up carrying weight, that’s part of the job. The ones who grow are intentional about what they carry, when they carry it, and why.
— Akash