What it is and what problem it solves
Almost every segmentation exercise ends the same way: a list of segments ranked by size, and the biggest one goes first. It is one of the most expensive mistakes in B2B. The biggest segment is usually also the most contested, the slowest to open and the one that demands the most proof.
The matrix scores each segment on three axes. Volume: how much business there is. Winnability: how much of that business you can win with the resources, channel and proof you have today. Value: the margin, recurrence and cost to serve each customer leaves behind. The output is not a ranking. It is a position per segment with its reason written down, which leadership can debate and sign.
When to use it and when not to
It is used in the diagnosis phase, after the market analysis and before defining the ideal customer profile and the decision-maker map. It makes sense when there are at least four candidate segments and resources to attack one or two with real intensity.
It adds nothing if the market is a single obvious segment. It is not a tool for prioritizing named accounts either: that belongs to the account plan. And if there is no minimum sales data, you run a round of interviews first and score the matrix with hypotheses flagged as such, to be corrected as soon as data arrives.
How to build it
- Define workable segments. A segment exists if you can name it, count it and reach it through a channel. "Hotels" isn't one if the chain and the independent buy differently. Deliverable: four to ten segments with their membership criteria.
- Estimate volume. From your own sales, public records and a read on the channel. Work in ranges, not falsely precise numbers. Deliverable: relative volume per segment with its source.
- Score winnability. Using the template's sub-criteria, each backed by evidence. Deliverable: a 1-to-5 score per sub-criterion.
- Estimate value. Margin, recurrence and cost to serve. Deliverable: relative value per segment.
- Position and decide. Each segment gets a position and a justification. Deliverable: the signed matrix, which feeds the activation sequence.
Anatomy of the deliverable
The matrix has two tables. The first is the decision:
| Segment | Volume (1-5) | Winnability (1-5) | Value (1-5) | Position | Why |
|---|---|---|---|---|---|
| Segment A | 5 | 2 | 3 | Second wave | The decision-maker buys through a distributor we don't have today |
| Segment B | 3 | 4 | 4 | First wave | Direct access to the owner and proof available |
| Segment C | 2 | 4 | 2 | Watch list | Easy to win, thin margin |
| Segment D | 4 | 1 | 4 | Out this period | Multi-year contracts with the competitor |
The second breaks down winnability, which is where complacency hides:
| Sub-criterion | Question | Score 1 | Score 5 |
|---|---|---|---|
| Access to the decision-maker | Do we reach the person who decides? | No contact, no channel | Direct relationship or own channel |
| Channel | Is there a channel already serving this segment? | It has to be built | Up and running |
| Proof | What do we use to demonstrate the promise? | Nothing verifiable | Comparable customers and our own data |
| Switching cost | What does it cost the customer to leave their supplier? | Contracts and technical dependencies | Open purchase on every order |
| Offer fit | Does the current offer work without rebuilding it? | Needs a new product | Fits as is |
House rule: any score above 3 requires named evidence. And the axes are never averaged: a 5-1-5 and a 3-3-5 produce similar averages and opposite decisions.
The classic framework: GE-McKinsey, and why we change one axis
The GE-McKinsey matrix dates from the 1970s, when General Electric asked McKinsey for a method to decide which business units to invest in. It crosses two composite axes, the attractiveness of the market or industry and the competitive strength of the unit, in a nine-cell grid that leads to three decisions: invest and grow, hold selectively, or harvest and divest. Its advantage over the BCG matrix is that each axis combines several weighted factors instead of one.
We keep its central logic: separate what the market offers from what you can do in it. We change three things:
- The unit of analysis. We move from the business unit to the customer segment, which is where a mid-sized company makes its commercial decisions.
- Competitive strength becomes winnability. In practice, strength gets scored optimistically. Winnability rests on observable facts: access, channel, proof, switching cost.
- Value stands on its own. A large market with poor margin is not attractive to a mid-sized company, however attractive it looks in the abstract. Splitting it out stops volume from hiding margin.
Where the original falls short: the factor weights are subjective and the picture is static. We don't remove judgment; we make it visible, because every score carries its reason and gets reviewed at the period committee.
Common mistakes
- Scoring winnability with wishes. "We have a good relationship with the industry," with no names. It happens because the person scoring is the person selling, and salespeople are optimists by trade.
- Confusing winnable with priority. A segment that is very easy to win and thin on margin consumes the team and doesn't move the result.
- Not writing down the "out." If nobody signs off on which segments are not being pursued, the sales team keeps chasing them because somebody knows somebody.
- Segmenting by industry and stopping there. Industry is a database label. What defines a segment is how it buys.
An illustrative scenario
Suppose a Spanish manufacturer of industrial maintenance equipment is looking at the Dominican Republic. It has four candidates: free-zone companies, chain hotels in the tourist region, family-owned food manufacturers and construction firms.
In this hypothetical case, the free zones score high on volume and low on winnability, because the purchase is decided at headquarters, outside the country. Chain hotels buy through a central purchasing office with approved suppliers. Family-owned food manufacturers offer direct access to the person who decides and shorter cycles. The matrix puts them in the first wave and leaves the free zones for a second wave worked through the parent companies. How to approach that market is covered in go-to-market in the Dominican Republic.