- CORE QUESTION
- How do we create, deliver, and capture value sustainably?
- PRIMARY OUTPUT
- A scalable business model, unit economics, and a scaling strategy
- LEAD
- Leadership / Strategy
- KEY COLLABORATORS
- Finance, Marketing, Sales, Operations, Product
Once value and product begin to take shape, treat what you own as a dynamic portfolio of business models, not a collection of products or business units, and manage, evolve, and rebalance it as deliberately as a fund manages capital. The question shifts from "what businesses do we own?" to "how do we continuously evolve them?"
Discover
What businesses do we actually have? Get every model onto one page each, and all of them onto a single portfolio map.
Tools and casesBusiness Model Canvas Business Model Portfolio Canvas Case Study
Open step 1 →2Position
Plot every business by current strength and future market attractiveness, into four strategic zones: Scale, Invest, Transform, Harvest.
Tools and casesBusiness Model Portfolio Canvas Case Study
Open step 2 →3Explore
Identify the next business models through a structured chain: market shifts to customer problems to jobs to new propositions to model options to experiments
Tools and casesAssumption Tracker Case Study
Open step 3 →4Decide
Allocate investment explicitly across horizons: roughly 70 percent core, 20 percent adjacent, 10 percent new, adapted to your situation.
Tools and casesPortfolio allocation Case Study
Open step 4 →5Manage
Allocate investment explicitly across horizons: roughly 70 percent core, 20 percent adjacent, 10 percent new, adapted to your situation.
Tools and casesThe portfolio rhythm Case Study
Open step 5 →Business Model Workflow.
From inventory to allocation: a five-step cycle for continuously managing, evolving, and rebalancing your portfolio of business models, with AI support mapped to each step.
Discover the Portfolio
Activity
Make the whole portfolio visible: capture every business the organization runs, its customer, problem, value proposition, revenue model, channels, capabilities, costs, and owner. One canvas per model, every model on a single map.
Tools
- Business Model Canvas
- Business Model Portfolio Canvas
AI Support
- LLM extraction into canvas blocks
- AI clustering into value spaces
- Sidekick canvas summaries
Build the portfolio map as a live board in Miro, with one linked Business Model canvas per business, so discovery is done once and then maintained, never redone from scratch.
Position the Portfolio
Activity
Assess every business against a common yardstick: how strong is it today, and how attractive is its market tomorrow? Plot the portfolio into four zones, Scale, Invest, Transform, Harvest, and read the gaps.
Tools
- Business Model Portfolio Canvas
AI Support
- AI market scans for attractiveness
- Automated strength scoring
- Bear-case Sidekick prompts
Keep the matrix live on the shared board, every score linked to the evidence behind it, so positions move the moment the data does instead of once a year.
Explore Next Models
Activity
Fill the gaps the matrix exposed by running a structured chain: market shifts to customer problems to Jobs-to-be-Done to new value propositions to business model options to experiments that test the riskiest assumptions first.
Tools
- Assumption mapping
AI Support
- LLM trend synthesis
- AI clustering of customer problems
- AI-drafted experiment designs
Run the whole chain left to right on a single board, so every model option stays traceable to the market shift that produced it and no idea arrives orphaned.
Decide the Allocation
Activity
Make investment choices explicit: allocate across horizons, roughly 70% core, 20% adjacent, 10% new, adapted to your situation, with an owner, milestones, tranches, and an end condition on every funded line.
Tools
- Ambition matrix
AI Support
- AI scenario runs on allocations
- Evidence roll-ups per bet
- Sidekick board summaries
Publish the allocation board to the whole organization, with owners, tranches, and milestones visible, so the portfolio choices can discipline everyday behaviour instead of living quietly in a deck.
Manage the Rhythm
Activity
Turn the portfolio into a quarterly rhythm: sense, update, assess, explore, decide, execute, learn. Run it in a day, off one map, one matrix, one allocation, and end every cycle with real movement.
Tools
- Portfolio review
AI Support
- Sensing Flows into the board
- AI deltas since last quarter
- Auto-refreshed performance data
Keep the board as the single source of truth the quarterly cycle runs on, so every quarter starts where the last one ended, with nothing rediscovered and nothing lost.
A portfolio of businesses, not a collection of products.
Ask most organizations what they own and they will hand you a product list or an org chart. Neither answers the question that matters.
Underneath the products and the units sit business models, distinct ways of creating, delivering, and capturing value, and it is at that level that companies quietly thrive or quietly age. The good way to manage them is to treat the whole set as a dynamic portfolio of businesses: living positions to be watched, rebalanced, grown, and sometimes ended, with the same discipline a fund manager brings to capital.
That is what this workflow does, in five steps that form a single loop.
Five steps, one loop.
It starts by making the portfolio visible. Discover captures every business the organization actually runs, each in the same blocks, its customer, its job, its proposition, its revenue model, its channels, capabilities, costs, market, and owner, and brings them onto one map. Position then plots each model by two measures that refuse to be confused with each other: how strong the business is today, and how attractive its market will be tomorrow. Four zones fall out of that matrix, Scale, Invest, Transform, and Harvest, and each implies a different move. The insight that changes behaviour is that the portfolio, not the individual business, is the unit of strategy.
The map will always show a gap, and Explore exists to fill it deliberately. Not with a brainstorm, but with a chain: market shifts point to changing customer problems, problems sharpen into Jobs-to-be-Done, jobs suggest new value propositions, propositions imply business model options, and options end as experiments on their riskiest assumptions. Decide then converts everything into the only language an organization fully believes, explicit allocation: roughly seventy percent to the core, twenty to adjacent moves, ten to genuinely new models, adapted to the situation but always chosen on purpose, across horizons, rather than funding everything as if it carried the same risk.
And because every one of those answers expires, Manage turns the whole thing into a quarterly rhythm: sense what changed, update the map, assess the positions, explore the next options, decide the adjustments, execute with owners and milestones, and learn by feeding the results back in.
The shift that matters.
Read the workflow as one movement and it runs from the market inward: shifts become customer insight, insight becomes jobs, jobs become models, models become portfolio positions, positions become investment choices, choices become experiments, and experiments become the next businesses. The portfolio evolves, continuously, on purpose.
Which is the real point of the five panels that follow. The question this workflow retires is "what businesses do we own?", a question any spreadsheet can answer. The question it installs is harder and worth everything: "how do we continuously manage, evolve, and rebalance our portfolio of business models?" The organizations that can answer it do not just have good businesses. They have a way of always having them.
The business model portfolio canvas, seeing every bet at once.
Why one canvas is not enough.
Step 1 of Business Model Innovation produces a Business Model Canvas: a single page describing how the business creates, delivers, and captures value today. That is necessary, but it quietly assumes something rarely true, that the organization is running exactly one business model. Almost none are. There is usually a core model that reliably pays the bills, and alongside it one or more emerging models being explored, tested, or half-committed to, often without anyone in leadership saying so out loud, or agreeing on how each should be judged.
Treating them all the same way is a common and expensive mistake. A proven, cash-generating model and an early, unproven experiment need different funding, different metrics, and different patience from leadership. Judge them by the same standard and one of two bad things happens. Either the new bet gets starved of the room it needs to prove itself, measured too early against a maturity it has not had time to earn. Or the core business gets quietly bled to fund an exciting experiment that has not yet shown it deserves the investment.
The trouble is that this rarely happens through a bad decision anyone would defend out loud. It happens through drift. Nobody stands up and says the experiment should be held to the core's standards, or that the core can afford to lose a little funding this quarter. The comparisons just happen quietly, in budget meetings and hallway conversations, and by the time the damage shows up, the reasoning behind it has long since evaporated. The Business Model Portfolio Canvas exists to make that distinction visible, and deliberate, before either failure sets in.
What the portfolio canvas actually plots.
Instead of mapping one model, the portfolio canvas plots every model the organization is running or considering across two dimensions that decide how each should be managed: how mature and proven it is, and how well it fits the organization's current strategy and capabilities. Maturity asks a concrete question. How much real evidence exists that this model works, in paying customers, repeatable revenue, and a cost structure tested against reality rather than projected on a spreadsheet? Strategic fit asks a question of judgment. How well does this model use the brand, relationships, channels, and capabilities the organization already has, versus requiring it to become, in effect, a different company?
Plotted this way, three broad zones emerge, and each calls for a different way of managing what falls inside it.
Models that are both mature and a strong strategic fit sit in a protect-and-scale zone. These are the proven earners that fund everything else, and leadership should be actively defending them from underinvestment, not assuming they will keep running themselves.
Models that are partly proven with a reasonably good fit sit in a nurture zone. They are worth real investment and a real growth plan, but should not yet be held to the core business's standards.
And models that are unproven, exploratory, or only loosely connected to current capabilities sit in an explore zone. These are small, deliberate bets, funded and measured as such. They are expected to fail more often than they succeed, and when they do, they should never threaten the resourcing of the core.
The zones are not permanent addresses. The whole point of plotting the portfolio is to watch models move across it over time: an explore bet that earns its evidence graduates into the nurture zone, a nurtured model that proves itself becomes part of the protected core, and a once-dominant core model can slide toward the edge as the market shifts beneath it.
A worked example, one company running three models.
The restaurant inventory company carried through the earlier Customer Value Creation chapters offers a clean illustration. On the surface it looks like one business. Plotted on the portfolio canvas, it is clearly three, and each one needs to be managed as if the other two did not exist.
Its core model is the subscription SaaS product itself: independent restaurants paying monthly for the usage-tracking and ordering tool validated across those earlier chapters. It is proven. Real customers, real recurring revenue, a cost structure the team actually understands from operating it. It sits firmly in protect-and-scale, and the canvas's job here is discipline more than discovery: making sure this model's growth is properly funded, rather than treated as a foregone conclusion while attention drifts toward newer ideas. Mature businesses are rarely killed by competitors. They are neglected by their own leadership, who assume the thing that pays for everything will keep paying without attention.
Alongside it, the same customer relationships suggest an adjacent model: a supplier marketplace, connecting the restaurants already using the inventory tool directly with local suppliers, monetized through a small transaction fee. It reuses the core model's customer base and trust, a real strategic-fit advantage. But the marketplace mechanics, recruiting suppliers, managing two-sided trust, processing transactions, are new territory the team has not operated before. It is partly proven at best, and belongs in nurture: worth serious investment given the fit, but not yet judged by the SaaS product's mature unit economics. Hold it to the core's numbers in year one and you would kill it before it had a chance to work.
Further out sits a transformational option: aggregating anonymized purchasing and waste data across the restaurant network, and selling regional trend reports to food suppliers and distributors. This is an entirely different customer, suppliers rather than restaurants, an entirely different value proposition, market intelligence rather than operational savings, and a business the team has no experience running. It is unproven and an uncertain strategic fit, exactly the profile of a model that belongs in explore: funded as a small, contained bet, evaluated on what it teaches rather than on revenue it has not had a chance to generate, and not held to the same bar as the core business.
Now imagine the same company without the canvas. All three models land in the same budget meeting, described in the same slides, judged against the same instinct: which will make money soonest? The marketplace and the data business look thrilling next to a subscription tool everyone already takes for granted, so they attract the funding and the best people while the core quietly starves. A year later the exciting bets have burned cash without proving themselves, and the one model that actually worked has stalled from neglect. Nobody chose that outcome. It simply happened, because three fundamentally different things were forced to compete on a single scale.
Notice what the canvas does instead. It stops the three models from competing for the same attention on the same terms. Each is funded, measured, and protected according to what it actually is: the core defended, the marketplace nurtured, the data business explored. That is the whole point. Not to pick a winner today, but to give three very different bets the three very different kinds of room they each need.