Three visions in one room
Six weeks after Maya Smith joined Brindlepath, the company held its annual planning offsite. By lunch on the first day, it was clear the leadership team did not share a picture of the future.
The CEO wanted Brindlepath to become “the operating system for care coordination,” a platform that every team in a health system would use. The head of sales wanted a product that could win the forty open opportunities in the pipeline, most of which were asking for scheduling improvements and a patient app. The chief medical officer, who had joined from a large health system, wanted the company to focus narrowly on high-risk patients after discharge, where the clinical evidence was strongest and the readmission penalties gave buyers a reason to act.
All three were reasonable. All three were, in their own way, strategies. And because nobody had chosen among them, the product team had been trying to serve all three at once, which is to say, none of them well.
On the second morning, Maya put a single page on the screen. It did not contain a roadmap. It contained a diagnosis of the market, one choice about where Brindlepath would play, one claim about how it would win, and a short list of things the company would stop doing. The argument that followed was the most productive two hours of the offsite.
This chapter is about writing that page, and about the work that makes it true.
Vision, strategy, roadmap: getting the words straight
Few words in business are used as loosely as vision and strategy. Teams waste hours arguing past each other because one person’s strategy is another’s roadmap. A simple cascade helps.
- Vision describes the future you are trying to create for customers, usually three to ten years out. It answers why are we here? It should change rarely.
- Strategy is the set of choices about how you will get there: which customers and problems you will focus on, and how you will win against the alternatives. It answers where will we play and how will we win? It is reviewed every year or so, and changes when the market does.
- Outcomes are the measurable results that tell you the strategy is working: adoption, retention, revenue, the customer’s own results. They answer how will we know? They are set quarterly or half-yearly.
- Roadmap is the sequence of bets the team will make to move those outcomes. It answers what will we try, in what order? It changes as you learn.
- Backlog is the detailed work to deliver the current bets. It changes every week.
The common failure is to skip the middle: a lofty vision on a slide, a detailed backlog in a ticketing tool, and nothing in between that explains why this backlog serves that vision. When the middle is missing, every stakeholder fills it with their own strategy, which is exactly what had happened at Brindlepath.
Writing a vision people can use
A useful product vision has four properties.
It describes the customer’s world, not your product. “Brindlepath is the leading care-coordination platform” is about you. “No patient leaves the hospital without someone who owns their next step” is about the world you want to create. The second is far more useful, because it lets people judge any idea against a simple question: does this help make that true?
It is ambitious but believable. A vision nobody believes is decoration. A vision that is already true is a status report.
It is short enough to remember. If people cannot repeat it after hearing it twice, it will not guide their decisions.
It is concrete enough to say no. “Improve healthcare” excludes nothing. “Every patient has an owner through every handoff” excludes a great deal, including, as Brindlepath discovered, building a better scheduling tool for its own sake.
A good test: show the vision to a new engineer and ask them to name two things it rules out. If they cannot, it is too vague.
Strategy is a set of choices
The best short definition of strategy I know comes from Richard Rumelt, who argues that good strategy has a simple core: a diagnosis of the challenge, a guiding policy for dealing with it, and a set of coherent actions that carry out the policy.1 Bad strategy, by contrast, is a list of goals, a pile of fluffy language, or a refusal to choose.
Applied to product, the three parts look like this:
Diagnosis. What is really going on in the market, and what is the hardest problem we face? Brindlepath’s diagnosis: health systems are penalized for readmissions but have no reliable way to make sure someone owns each patient after discharge. Many vendors sell care-management modules, but few solve the handoff between inpatient and outpatient teams, because it requires integration with records, clear rules about ownership, and workflows that fit outpatient navigators. Brindlepath’s original product had spread itself across scheduling, messaging and reporting, and was not decisively better at any of them.
Guiding policy. What approach will we take? Brindlepath’s: become the best product in the market for owning high-risk patients in the 30 days after discharge, and prove it with outcomes, starting with health systems that face meaningful readmission penalties.
Coherent actions. What will we do, and stop doing, that follows from the policy? Invest in record integration and ownership workflows; build outcome reporting that a chief nursing officer can take to the CFO; stop new investment in the standalone scheduling product; decline deals that need a patient app as a condition of purchase; train the sales team to sell a problem, not a feature list.
Notice that the actions include things the company will stop doing. Michael Porter’s line at the top of this chapter is the most-quoted sentence in strategy for a reason. A strategy that does not make anyone unhappy has probably not chosen anything.
Where to play, how to win
A complementary lens comes from A.G. Lafley and Roger Martin, who frame strategy as an integrated set of answers to a few questions, the most important being where to play and how to win.2
Where to play covers which customers, which problems, which channels and which geographies. How to win covers the advantage that will make those customers choose you: a better outcome, a lower cost, a superior experience, a unique capability or a trusted brand.
The two answers must fit each other. Brindlepath could not win on price against the electronic health record vendor’s bundled module, and it could not win on breadth against large platforms. It could win on depth and outcomes in one specific, painful workflow. That determined where it should play: health systems where that workflow was a priority, not every organization that might buy care-management software.
Positioning: how customers should understand you
Strategy is what you decide. Positioning is how you help the market understand that decision. Even an excellent product fails if customers file it in the wrong mental drawer.
April Dunford’s framework is the most practical I have used. She defines positioning through five components that need to fit together:3
- Competitive alternatives: what customers would do if your product did not exist. Often the status quo.
- Unique attributes: the capabilities you have that the alternatives lack.
- Value: the benefit those attributes enable for customers, in their terms.
- Best-fit customers: the characteristics of the customers who care most about that value.
- Market category: the context you choose to present yourself in, which tells customers what to compare you with.
The order matters. Most teams start with the category and work backward. Dunford recommends starting with the alternatives, because your uniqueness only exists in comparison to something.
Brindlepath’s old positioning put it in the category “care-management software,” where buyers compared it, feature by feature, with large platforms and the record vendor’s module, and usually found it smaller. The new positioning put it in a narrower category, “post-discharge care transitions,” where the alternatives were spreadsheets, phone calls and modules that were not built for the handoff. In that frame, Brindlepath was not the smaller option; it was the specialist.
Positioning is not a tagline. It is the shared understanding that sits behind every tagline, sales deck, website page and product decision. Write it down, test it with customers and prospects, and expect it to change when your market does.
Communicating the vision and strategy
A strategy that lives in the head of the product leader is not a strategy. It is a hunch. The work of making it shared is less glamorous than writing it and far more important.
Write it on one page. If the strategy does not fit on a page, it is not yet clear. The page should contain the vision, the diagnosis, the guiding policy, the three to five most important actions, the outcomes you will measure, and what you will stop doing.
Tell it as a story. People remember stories, not bullet points. “A navigator at one of our customers found two patients last week who would have been missed” carries more weight than “improve post-discharge ownership rates.”
Repeat it until you are sick of it. By the time the product leader is bored of the strategy, most of the company has heard it once. Say it in all-hands meetings, planning sessions, customer calls and onboarding.
Connect every roadmap item back to it. When someone asks why the team is building something, the answer should trace back to the page in one or two steps.
Invite challenge, then decide. A strategy that has survived argument is stronger than one that was handed down. But a strategy that is re-argued every week is not a strategy. Hold a real debate, make the decision, write it down, and revisit it on a schedule rather than on demand.
The “not doing” list
The most powerful page in Maya’s offsite presentation was the last one: what Brindlepath would stop doing. It listed the standalone scheduling roadmap, the patient app, custom reporting for individual customers and two integrations requested by a single prospect.
Each item had a sponsor in the room. Each cut hurt. But the list did something no amount of prioritization could: it freed about a third of the engineering capacity for the work that mattered, and it told the sales team, clearly, which deals to walk away from.
Write your own not-doing list. Share it as widely as the strategy itself. It is the clearest evidence that you have a strategy at all.
Brindlepath in practice: the one-page strategy
Here is the page Maya put on the screen at the offsite, lightly edited.
Vision. No patient leaves the hospital without someone who owns their next step.
Diagnosis. Health systems are penalized for readmissions but lack a reliable way to make sure someone owns each high-risk patient after discharge. Care-management modules exist, but few solve the handoff between inpatient and outpatient teams. We have spread ourselves across scheduling, messaging and reporting, and are not decisively better at any of them.
Where we play. Health systems with meaningful readmission penalties and dedicated outpatient navigator teams, starting in our current regions.
How we win. Depth in one workflow: discharge data within hours, clear ownership, one-step reassignment and outcome reporting, proven with customers’ own numbers.
Guiding policy. Become the best product for owning high-risk patients in the 30 days after discharge, and prove it with outcomes.
Actions. Build the ownership queue and the standard record connector. Build outcome reporting for finance teams. Retrain sales to sell the problem. Move existing customers off custom integrations.
Outcomes. High-risk patients with a confirmed owner within 24 hours; renewal rate; time to first value.
Not doing. New investment in standalone scheduling; a patient app as a condition of sale; custom reports for individual customers; integrations requested by a single prospect.
The page was revised twice in the following year, both times on schedule. The not-doing list grew by one item and lost one: the patient-messaging need returned, but through an acquisition rather than a build, as Chapter 7 describes.
Common mistakes
Mistaking goals for strategy. “Grow revenue 40 percent” is a goal. A strategy explains how, given the market’s realities, you will achieve it, and what you will not do along the way.
Writing a vision about yourself. Visions that describe the company (“the leading platform”) inspire nobody and guide no decisions. Describe the customer’s world.
Refusing to choose. A strategy that tries to serve every segment, match every competitor and say yes to every request is not a strategy. It is a list of hopes.
Positioning against the wrong alternative. Teams compare themselves to the competitor they admire, while customers compare them to the spreadsheet they already have. Start from what customers would really do without you.
Strategy by slide. A strategy presented once at an offsite and never repeated does not exist for most of the company. Repetition is not redundancy; it is how strategy becomes real.
Changing strategy every quarter. Constant change looks agile and produces teams that never finish anything. Review strategy on a schedule, and change it when evidence demands it.
Try this
Write your product’s strategy on a single page, using the template in the Appendix: vision, diagnosis, where you play, how you win, guiding policy, three to five actions, outcomes and a not-doing list. Then run two tests. First, give it to a new engineer and ask them to name two ideas it rules out. Second, ask a sales leader to describe, from the page alone, the customer you should walk away from. If either test fails, the strategy is not yet clear enough to guide decisions. Rewrite it until both pass.
How AI changes this
AI is reshaping product strategy in three ways that every product leader needs to think through.
Features commoditize faster. When a capability can be built in days with a foundation model, a feature on its own is rarely a lasting advantage. Competitors, including the large platforms, can match a clever summarization or drafting feature within a quarter. Durable advantage moves to the things that are hard to copy: deep integration into workflows, proprietary data and feedback loops, trusted relationships, distribution, and the accumulated understanding of a specific customer’s problem. When you write your “how to win,” be honest about which of these you actually have.
New competitive alternatives appear. For many products, a general-purpose AI assistant is now a competitive alternative. A navigator might paste a discharge summary into a chatbot instead of buying specialized software. Your positioning has to explain why your product is better than that, and in regulated settings the answer often includes trust: privacy, accuracy, auditability and integration that a general tool cannot offer.
AI is a great sparring partner for strategy. Ask a model to attack your diagnosis, list the strategies a competitor might pursue, or argue for the opposite choice. Ask it to rewrite your vision in plain language, or to find the actions in your plan that do not fit your guiding policy. It will not make the choice for you, but it will help you see where your thinking is thin.
A warning: AI also makes it easy to produce strategy documents that sound impressive and choose nothing. Fluent prose about “leveraging synergies across the care continuum” is now free. A real strategy still requires a person willing to say no.