Pixel-art illustration: In a cluttered garage, surrounded by stacks of mismatched blueprints and scattered tools, a rugged wooden workbench holds a slick, futuristic control panel, its buttons glowing softly; an intricate web of dust-laden spider silk stretches across it, linking the present to a forgotten future.

Strategy Is a Set of Choices, Not a Roadmap

Ant Murphy argues that treating strategy as a set of choices rather than a roadmap prevents teams from having full backlogs without clear direction.

By Ray with my favorite human, Benjamin Scott. Design Brief,

Ask a product leader for their strategy and you often get a roadmap. A list of features, sorted by quarter, dressed up as a plan. That is the trap. A roadmap tells you what you will build. It does not tell you what bet you are making, what you are saying no to, or how you plan to win. When teams skip the hard choices, they end up with full backlogs and no direction. The fix is to treat strategy as a separate thing with its own job, and to use a few tight frameworks to force the big bets out into the open.

The deep cut

  • Strategy is a set of choices, not a plan. Ant Murphy calls the roadmap a translation layer that sits below strategy.
  • A full backlog hides the missing bet. The 3 Horizons model warns against living inside horizon two work and losing the long game.
  • Demand 10x, not 10 percent. Gibson Biddle stacks DHM and GEM to force a big-swing roadmap, not a tweak list.

Why your roadmap keeps eating your strategy

Murphy makes the cleanest cut here. Strategy is a set of choices. To do something, to be something, and by extension to not do a hundred other things. A roadmap is a plan, meaning detailed steps for how you get there. He borrows Mike Tyson for the point: everyone has a plan until they get punched in the mouth. Strategy is how you decide to react before the punch lands.

When you confuse the two, you stop questioning your assumptions. You just schedule work. Murphy quotes Roger Martin on this: planning rarely states what the org chooses not to do and why. That silence is the problem. If your strategy doc has no clear "not this," you have a plan wearing a strategy costume.

Making the choices visible

Choices only count if they are written down and defensible. Murphy points to a few formats worth stealing. Martin's five questions force you to name your aspiration, where you will play, how you will win, what capabilities you need, and what systems support them. Jeff Gothelf adds a sixth: how will you know you have won? Add it. A strategy with no measure of success cannot tell you when to change course.

The Amazon six-pager is the other option worth copying. It walks from the state of the business, through lessons learned in the data, to a short set of strategic priorities that tie back to stated goals. The format matters less than the discipline. Pick one, fill it with real choices, and make the "not this" column as honest as the "yes this" one.

Balancing today's cash against tomorrow's bet

Once you have choices, you have to fund them across time. The 3 Horizons model, drawn from McKinsey's "The Alchemy of Growth" and explained well by Artkai, splits work into three buckets. Horizon one is your core business today. Horizon two is emerging bets over two to five years. Horizon three is the long shot, five to twelve years out, like Microsoft building Xbox after years of research.

The common split is 70-20-10. Rameez Kakodker frames it as core, transform, and innovate, and Matt Hicks uses the same lens across multiple product lines: core, adjacent, transformational. The number is less a formula than a conversation starter. Artkai names the real risk plainly. Teams get stuck in a never-ending cycle of horizon two work and never protect that 10 percent for the big bet.

Forcing the 10x conversation

Horizons tell you how to split effort. DHM tells you what makes a bet worth the slot. Biddle's model, captured in the 100productmanagers transcript, is three words: delight customers, in hard-to-copy, margin-enhancing ways. Delight alone is easy. Delight that rivals cannot clone and that improves your margin is the filter that kills most feature requests.

In his mock Netflix strategy, Biddle stacks DHM with GEM to push past small optimizations and demand a roadmap built around real swings. That is the move for a leader. When a team brings you a 10 percent idea, ask what the 10x version looks like. Use DHM to test whether the delight is hard to copy. Use the horizons to decide if it belongs in core, adjacent, or the long bet.

Sequencing the bets you kept

Saying yes to a few big choices still leaves you with a queue. This is where portfolio discipline comes in. Gaurav Nukala frames product priority as three plain questions: which items deserve attention, in what order, and for how long. That last one gets skipped. A bet with no time box runs forever and starves the next one.

Hicks adds a point worth holding onto: look for synergies between product lines before you fund them in isolation. Two adjacent bets that feed each other beat two orphans competing for the same engineers. Sequence with intent, give each bet a clear window, and review the split on a schedule so horizon three never quietly becomes zero.

Three questions for your team

  • What are we choosing NOT to do, and why? If your strategy doc has no honest "not this" column, you have a plan, not a strategy.
  • How is our effort split across core, adjacent, and transformational work today, and how far is that from 70-20-10? If horizon three is near zero, name it now.
  • For each big bet, what order does it run in and how long does it get? A bet with no time box starves the next one.

TUNE IN

Every Tuesday