From one product to four
A year after the offsite, Brindlepath bought a small company that made a patient-messaging tool. The logic was sound: Brindlepath’s customers kept asking for a way to reach patients after discharge, and building it from scratch would have taken a year. Overnight, Maya went from leading one product to leading four: the core care-transitions product, an outcomes-reporting module that had grown up inside it, the acquired messaging tool, and the old scheduling product, which still had paying customers despite no longer being part of the strategy.
Each product had its own customers, its own technology, its own team and its own champion on the leadership team. Each wanted more investment. In her first portfolio review, Maya realized that the hardest question was no longer what should this product do next? It was how much should we invest in each product, and what should we stop?
That question is the subject of this chapter. Even if you manage a single product today, the habits here apply to the features and modules inside it, and they will matter the day your company grows a second product.
Products have life cycles
Every product moves through stages, and each stage calls for a different kind of investment and a different measure of success.
- Introduction. The product is finding product-market fit. Revenue is small; learning is the main output. Measure learning velocity and early signs of retention.
- Growth. The product has found fit and is scaling. Invest heavily in reach, onboarding and the features that unlock larger customers. Measure growth rate, net revenue retention and acquisition efficiency.
- Maturity. Growth slows; the product is a reliable revenue and margin engine. Invest in efficiency, reliability, pricing and adjacent expansion. Measure margin, retention and expansion.
- Decline. The market or technology has moved on. Invest only to keep customers safe and satisfied while you plan the exit. Measure cost to serve and migration progress.
The most common portfolio mistake is to manage every product the same way: the same planning cycle, the same metrics, the same expectations. A product in introduction judged on revenue will be killed before it has a chance. A product in maturity judged on growth will be pushed into features nobody needs. A product in decline that keeps its full team will consume resources the growth products badly need.
Mapping the portfolio
A portfolio map makes the investment conversation concrete. There are many versions. The classic Boston Consulting Group growth-share matrix plots products by market growth and relative market share.1 For software portfolios, I find it more useful to plot each product by its revenue growth and its gross margin, with the size of each bubble showing its revenue.
Figure 7.1 shows Brindlepath’s portfolio. The core care-transitions product is large, growing and healthy: the engine. The outcomes module is small, growing fast and highly profitable, because it reuses the core product’s data: a candidate for more investment. The messaging tool, newly acquired, is growing but at a low margin, because it pays a third party for every text message sent: it needs its unit economics fixed before it scales. The scheduling product is shrinking and barely profitable: a product to retire carefully.
None of those conclusions is automatic. The map does not make the decision. It replaces four separate pitches, each delivered with passion by a product’s champion, with one shared picture that the leadership team can argue about productively.
Three horizons
Portfolio thinking is also about time. The three-horizons model, introduced by Mehrdad Baghai, Stephen Coley and David White, sorts work by how far it is from today’s core business.2
- Horizon 1 extends and defends the core business: improvements to the products that pay the bills today.
- Horizon 2 builds emerging businesses that are starting to generate revenue and could become the next core.
- Horizon 3 creates options for the future: research, experiments and early bets that may not pay off for years, if ever.
Companies under pressure tend to pour everything into Horizon 1, because it has the clearest return and the loudest customers. That works until the core business matures, at which point there is nothing ready to replace it. A deliberate allocation across horizons protects the future from the present. A frequently cited guideline, associated with Google in the 2000s, puts roughly 70 percent of effort on the core, 20 percent on adjacent growth and 10 percent on transformational bets. The exact split matters less than making it explicit and reviewing it regularly.
For Brindlepath, Horizon 1 was the core care-transitions product and its record integrations. Horizon 2 was the outcomes module and the messaging tool. Horizon 3 was a small team exploring AI agents that could draft a post-discharge plan for each high-risk patient for a navigator to review, a bet that, if it worked, could change the economics of care coordination entirely.
Platform and products
As a portfolio grows, the same capabilities appear in several products: user management, security, integrations, data pipelines, notifications, reporting, AI services. Building each one separately in every product is expensive and produces inconsistent experiences.
The usual answer is a platform: shared services that several products build on. Brindlepath’s most valuable platform capability turned out to be its record integration layer. Once it worked for the core product, the outcomes module and the messaging tool could use the same patient data without their own integrations, which cut implementation time for customers buying more than one product.
Platform work has a particular trap: it is easy to build a platform that nobody asked for, designed for imagined future products. The healthiest approach treats the platform as a product with internal customers. Extract shared capabilities from real products when two or more need them, rather than building them speculatively in advance.
Cross-selling and bundling
A portfolio creates new commercial options: selling additional products to existing customers, and bundling products together. Both can lift net revenue retention significantly, because selling to a customer who already trusts you is far cheaper than winning a new one.
But a bundle is only as good as the integration between its parts. Customers notice quickly when “the suite” is really three products with three logins, three data models and three support teams. Before you bundle, make sure the products share identity, data and design at least well enough that using them together is easier than using them apart.
Retiring a product well
Sooner or later, every portfolio needs to retire something. Brindlepath’s scheduling product had forty paying customers, a dwindling team and no future in the strategy. Keeping it alive cost more each year in maintenance and security work; shutting it down carelessly would damage trust with customers who also bought the core product.
A responsible retirement has five steps.
- Decide and explain why. Write down the reasons, the timeline and the alternatives customers will have. Share it internally first, so that sales and support can answer questions consistently.
- Give real notice. For business software, twelve months is a reasonable minimum; in healthcare, where changing a workflow requires training and sometimes clinical review, longer may be needed.
- Offer a path. Migration to another of your products, a partner’s product or an export of data in a usable format. Help customers move, especially your best ones.
- Handle data carefully. Regulated data has retention and deletion requirements. In healthcare, records may need to be kept for years after the product is retired. Plan how data will be returned, retained or destroyed, and document it.
- Free the capacity deliberately. Decide in advance where the team and budget will go. Otherwise, the savings from retirement quietly disappear into everything else.
Brindlepath migrated most scheduling customers onto a partner’s scheduling tool connected to the core product, kept a small number on an extended support contract, and moved the scheduling team into the record integration platform, where their knowledge of the health systems’ systems was invaluable.
Running portfolio reviews
A portfolio needs a regular rhythm of decisions, separate from the planning of individual products. Many companies hold a portfolio review each quarter. A good review covers:
- Each product’s stage and its measures of success for that stage, not one common scorecard.
- The portfolio map, updated with the latest numbers.
- The horizon allocation, planned versus actual.
- Shared platform needs and conflicts between products.
- Decisions: where to add investment, where to reduce it, what to stop, and what to explore.
The output is a short list of decisions with owners, not a set of status updates. If a portfolio review ends without any product getting more or less than it had before, ask whether anything was actually reviewed.
Brindlepath in practice: the first portfolio review
Maya’s first formal portfolio review, a quarter after the messaging acquisition, ran for two hours with the CEO, CFO, head of sales and the leaders of each product. She sent the portfolio map in Figure 7.1 two days ahead, with a short page for each product covering its stage, its stage-appropriate measures and a proposed decision.
The core care-transitions product was in growth. Its measures were net revenue retention, time to first value and the North Star. Proposed decision: keep investing, with the standard connector as the priority.
The outcomes module was in introduction moving into growth: small, fast-growing and very profitable. Proposed decision: add two engineers and make it part of the standard package for new customers.
The messaging tool was in growth but with poor unit economics, because the per-message fees it paid a third party were not passed on to customers. Proposed decision: renegotiate the provider contract, add usage-based pricing for high volumes, and hold investment until gross margin passed 65 percent.
The scheduling product was in decline. Proposed decision: announce retirement with twelve months’ notice, migrate customers to a partner, and move the team to the integration platform.
The discussion was lively, and not every proposal survived intact. The CEO pushed back on retiring scheduling so soon; the compromise was fifteen months’ notice. The head of sales wanted the outcomes module offered at a discount; the decision was to include it in the standard package instead. But the meeting ended with a written list of decisions and owners, and the next quarter’s review began by checking them. That habit, more than any framework, is what turned a collection of products into a portfolio.
Common mistakes
One scorecard for every product. Judging a new product on revenue, or a mature one on growth, pushes both in the wrong direction. Measure each product on what matters at its stage.
Funding by volume, not evidence. The product with the most passionate champion often gets the most investment. A shared portfolio map makes the conversation about evidence instead.
Starving the future. Spending everything on Horizon 1 feels responsible until the core matures and nothing is ready to replace it.
Building the platform first. Shared capabilities built speculatively, for products that do not exist yet, are rarely what those products need. Extract them from real use.
Bundling before integrating. A suite that is three logins and three data models disappoints the customers you most wanted to impress.
Never retiring anything. Products in decline quietly consume maintenance, security and support capacity for years. Retiring them well frees investment for growth.
Try this
Draw your own portfolio map. If you have one product, use its major modules or customer segments. Plot each by revenue growth and gross margin, size it by revenue, and label its life-cycle stage. Then write one sentence for each: invest more, hold, fix or retire. Bring the map to your next leadership meeting and ask whether everyone agrees with the sentences. Where they do not, you have found the conversation the portfolio review should have.
How AI changes this
AI becomes a platform capability. Many companies are building shared AI services, such as model access, retrieval over their own data, evaluation tools and safety controls, that every product can use. Treating AI as a platform avoids each team building its own and makes cost, quality and compliance manageable. It is also a natural Horizon 2 or 3 investment for most portfolios.
AI can blur product boundaries. An AI assistant that can act across several products, answering a question by drawing on scheduling, messaging and outcomes data together, can make a collection of products feel like one. That raises the value of integrated data and shared identity, and the cost of a portfolio that is really separate products under one logo.
AI can disrupt your own core. The most uncomfortable portfolio question of this era is whether an AI-native approach could replace your Horizon 1 product. Brindlepath’s Horizon 3 agent experiment could, if it worked, make some of its own core workflow unnecessary. Better to discover that yourself, in a small team you control, than to read about it in a competitor’s launch announcement.
AI makes portfolio analysis faster. Assistants can assemble the portfolio map, reconcile metrics across products, and summarize customer feedback product by product before each review. As always, the judgment about where to invest stays with the leadership team.