The four actions framework: redesigning your business model, not just your product

The four actions framework asks four questions about your industry: what to eliminate, what to reduce, what to raise, and what to create. Most teams answer those four questions on a feature list instead of a business model. Eliminate the fax option. Reduce the onboarding steps. Raise the support response time. Create a mobile app. None of that touches the business model, and none of it produces what Kim and Mauborgne actually meant by eliminate, reduce, raise, create. Their four questions were built to challenge an entire industry’s model, not a product backlog.

This guide blends the four actions framework with the nine blocks of the Business Model Canvas, the combination Alexander Osterwalder and Yves Pigneur describe in Business Model Generation. You get the four questions to ask about each block, three places to start, two real examples of what happens when the questions actually ripple through a model, and the point where the framework runs out of road.

What the four actions framework actually asks

The four actions framework comes from Kim and Mauborgne’s Blue Ocean Strategy. It asks four questions about the factors an industry competes on and treats as given:

1. Which factors that the industry has long competed on should you eliminate?
2. Which factors should you reduce well below the industry standard?
3. Which factors should you raise well above the industry standard?
4. Which factors should you create that the industry has never offered?

Eliminate and reduce lower your cost base. Raise and create lift the value a customer experiences. Ask all four questions in the same sitting and you get what Kim and Mauborgne call value innovation: value goes up, cost goes down, in the same move. That rejects the trade-off most strategy courses teach as a law of competition: differentiate or compete on cost, pick one. It is not a law. It is an assumption your industry has never bothered to question.

Kim and Mauborgne also pair the four questions with a strategy canvas, a graph that plots your factors against how much competitors invest in each one. That is a genuinely useful, separate tool and deserves its own treatment. The canvas I am blending the four actions with here is the nine-block Business Model Canvas, not the strategy canvas, because the business model canvas shows you something the strategy canvas cannot: what a value-side change costs you on the other side of your operation.

Most teams stop at the four questions. Four sticky-note columns, a facilitator, a flipchart, a good afternoon. It is a fine ideation exercise. On its own, it is not a Blue Ocean strategy, and it is not business model innovation.

Why an ERRC grid on its own stalls

I see this pattern regularly: a team fills in an ERRC grid, eliminates three cost items, creates two flashy features, and calls it strategy. Then nothing changes, because nobody asked what breaks somewhere else in the model.

An ERRC grid is four boxes on a wall. It tells you what to eliminate, reduce, raise, and create. It does not tell you what happens to your operations, your partnerships, or your margin once you actually do it. In 100+ sessions helping teams test business ideas, I have watched a group get genuinely excited about a “create” idea, then discover two weeks later that the idea needs a resource, a partnership, or an activity nobody costed at the time. The excitement dies in the finance review. Not because the idea was bad. Because nobody traced the ripple before committing to it.

This is not a knock on Kim and Mauborgne’s thinking. The four questions are sharp. The problem is scope: a standalone ERRC grid analyzes your value proposition in isolation from the rest of your business model. And the rest of your business model is exactly where the bill for “raise” and “create” gets paid. That gap between designing a new answer and tracing what it costs elsewhere is the same gap that stalls most business model innovation efforts, not just ERRC exercises.

The canvas makes the trade-offs visible

The Business Model Canvas has a right-hand side and a left-hand side, and that split is the entire reason blending it with the four actions framework works. The right side, Customer Segments, Value Propositions, Channels, Customer Relationships, Revenue Streams, is where value and emotion live. The left side, Key Resources, Key Activities, Key Partnerships, Cost Structure, is where efficiency and logic live.

Eliminate or reduce something on the value side, and something on the cost side moves with it, usually down. Raise or create something on the value side, and something on the cost side almost always has to grow to support it: a new resource, a new activity, a new partnership, a different cost driver entirely. Run the four questions against a single block and stop there, and you have designed half a business model. Run them against a block and then ask what happens across the rest of the canvas, and you have designed a system.

I see this constantly with manufacturing clients who consider a service layer on top of a product. Raising the value proposition, uptime guarantees, remote monitoring, faster turnaround, sounds free until you trace it. It usually is not. It needs a data platform, a monitoring team, a different partnership with a components supplier. The four actions framework asks whether that trade is worth it. The canvas is what lets you actually see the trade.

Three places to start

Kim and Mauborgne’s four questions can start from any block, but three starting points make the exercise tractable: the value proposition, the customer segment, or the cost side. Each is a version of what I use to structure ideation more broadly: business model innovation tends to originate from an offer, a customer, a resource, or a revenue change, and the most transformative shifts touch more than one at once.

Starting from the value proposition

Begin by asking the four questions about what you offer. Which less-valued features or services could you eliminate or reduce? What could you enhance or create instead, to produce a genuinely different customer experience? Then work sideways: what does the change cost, and how does it affect the customer side of the model, your channels, your relationships, your revenue streams? A value proposition change that does not ripple to at least one of those three has not actually changed anything a customer will notice.

Starting from the customer segment

Begin with who you serve instead of what you offer. Which new customer segments could you focus on, and which could you reduce or drop? What jobs do those new segments actually want done, not the jobs your current segment cares about? How do they expect to be reached, and what kind of relationship do they expect? Then price it: what does serving this new segment cost you in resources, activities, and partnerships you do not currently have?

Starting from cost structure

Begin with your highest-cost infrastructure. Which activities, resources, or partnerships cost the most? What happens if you reduce or eliminate the expensive ones? How would you replace the value they created, using something less costly? And on the other side: what value would a planned new investment actually create, before you approve it? This is the perspective I use most often with cost-heavy operations, because it forces the question everyone skips: are we protecting a cost center, or a source of value the customer would miss?

All three starting points end at the same place: a change on one side of the canvas, traced all the way to what it does on the other.

Two examples that show the ripple effect

Business Model Generation applies this blend to two well-known cases. Neither is mine, both are worth walking through because they show the ripple, not just the ERRC grid.

Cirque du Soleil. Traditional circus competes on a set of factors the industry has never questioned: star performers, animal shows, aisle concession sales, multiple simultaneous show arenas. Cirque du Soleil ran the four actions against that list.

Eliminating animals and star performers cut cost immediately: no animal care, no celebrity contracts, fewer touring logistics. Creating the theme and artistic elements changed Key Activities and Cost Structure in the other direction, more design and choreography work, but that new value proposition combined circus, theater, and opera into something adult audiences would pay significantly more to see. The eliminate side funded the create side. That is the ripple, not just the grid.

Nintendo’s Wii. The console industry competed on processing power, graphics fidelity, and game realism, three factors that mattered mainly to a narrow segment of dedicated gamers. Nintendo eliminated investment in top-tier chip development and console subsidies, and reduced raw console performance below what Sony and Microsoft shipped. It created motion control, the Wii Remote, and shifted its target customer segment from hardcore gamers to a much broader, more casual audience.

The customer segment change came first. Once Nintendo decided casual players were the target, the rest followed: off-the-shelf components instead of custom chips lowered cost, motion control created a new kind of value proposition, and the console sold at a profit per unit instead of at a subsidized loss. Start from the customer segment, and the cost and value changes follow logically. That is what “three places to start” means in practice, not three unrelated exercises.

For more worked examples of canvases built this way, see Business Model Canvas examples.

Running the four questions on your own canvas

The order matters more than most teams assume. Run it like this:

1. Pick one block. Not the whole canvas at once.
2. Ask the four questions against that block only: eliminate, reduce, raise, create.
3. Before writing anything into a grid, trace what each answer does to the rest of the canvas. What does eliminating this cost you in revenue, and what does it save you in resources? What does creating this require that you do not currently have?
4. Only once you have the ripple mapped, fill in the ERRC grid as a summary, not as the working document.
5. Repeat with a second starting block if the first round touched fewer than three canvas blocks. A change that stays contained to one block rarely produces real business model innovation.

Step 3 is where most sessions rush. Slow down and check four specific blocks every time, regardless of where you started: Key Resources (does this answer require something you do not own), Key Activities (does it require doing something you do not currently do, or stopping something you currently rely on), Key Partnerships (does it need a partner you do not have yet, or make an existing partner redundant), and Cost Structure (does the net effect actually lower cost, or did you only account for the savings and skip the new cost it introduces). An eliminate answer that looks like pure savings almost always has an offsetting cost hiding in one of those four blocks. Find it before the grid goes on the wall, not after budget approval.

What goes wrong when you reverse the order: teams fill in the ERRC grid first, treat it as finished strategic thinking, and only discover the ripple effects during implementation, when they are expensive to reverse. The grid is the output of the analysis. It should never be the analysis itself.

Where the framework stops

Here is the honest limit. The four actions framework, blended with the canvas, is excellent at generating options and showing you their consequences. It does not tell you whether customers actually want what you decided to create, or whether they will actually stop caring about what you eliminated. Every eliminate, reduce, raise, and create answer is still a hypothesis dressed up as a decision.

I have sat with teams who ran a sharp four actions session, produced a genuinely differentiated model on paper, and then built it without checking a single assumption with a real customer. Two of those assumptions turned out to be wrong, and both were exactly the ones the team was most confident about. Confidence in a workshop is not evidence.

There is a second limit worth naming. The framework assumes you already know which industry you are questioning and roughly who is in it. It says nothing about whether that industry frame is even the right one, or whether a completely different customer segment would make the whole exercise irrelevant. That earlier question, where in your business model the real opportunity sits, belongs to a broader diagnostic than four questions can carry on their own.

Two bridges from here. First, before you commit resources to a “raise” or “create” answer, use the Value Proposition Canvas to check whether the value you think you are adding maps to a job, pain, or gain your customer actually has. The four actions framework tells you what to change. The Value Proposition Canvas tells you whether the change lands. Second, treat every eliminate, reduce, raise, and create decision as a testable claim, not a conclusion. Testing Business Ideas covers how to turn those claims into experiments before you redesign the rest of your operation around them. The four actions framework generates the options. It was never built to validate them.