A positioning map plots competitors on two axes so relative position becomes visible at a glance. It is the most useful output of a competitive analysis and the easiest to draw badly.

Badly drawn, it produces the diagram everyone has seen: four quadrants, competitors scattered around, and your own company alone in the top right. That map has told you nothing. It was drawn to reach that conclusion.

The axes are the whole exercise

Everything depends on the two dimensions you choose, and there are only two rules.

Rule one: the axes must be things buyers actually decide on. Not things you find interesting. If no buyer has ever chosen a vendor based on an axis, plotting it produces a true picture of an irrelevant space.

The test: can you recall a real deal where this dimension decided the outcome? If not, discard it.

Rule two: the axes must be independent. Plotting “price” against “value for money” produces a diagonal line, because the two are versions of the same thing. Correlated axes waste a dimension — you have drawn a two-dimensional chart carrying one dimension of information.

A quick check: if knowing a company's position on axis A lets you predict its position on axis B, pick a different B.

Axes that usually work

Not universal, but reliably productive starting points:

  • Breadth vs depth. Suites against specialists. Almost always a real decision axis, and almost always contested.
  • Self-serve vs high-touch. How the product is bought and adopted. Determines cost structure and therefore who can profitably serve whom.
  • Time-to-value vs ceiling. Fast and limited against slow and powerful. Buyers trade explicitly on this.
  • Buyer seniority. Bought by a practitioner or by an executive. Predicts nearly everything about how a deal runs.
  • Configurability vs opinionation. Flexible-but-you-assemble-it against constrained-but-it-works.

Axes that usually disappoint: company size, funding raised, year founded, feature count. They describe companies rather than choices, and they encourage the reasoning error the inversion test is designed to catch.

Placing companies honestly

Two disciplines separate a real map from a flattering one.

Place from evidence, not impression. Before plotting, write down what would put a company at each end of each axis, then place against that definition. This is slower and it prevents the map from confirming what you already believed.

Place yourself last, and let someone else check it. Self-placement is where bias enters. If your company lands in the ideal corner, you have almost certainly chosen your own strengths as the axes.

A useful sanity test: would your competitor draw this map the same way? If they would draw it completely differently, at least one of you has picked axes that describe a preference rather than a market.

Reading the result

Three things are worth extracting.

Clusters. Companies bunched together are competing on the same terms, which means they compete on price and execution. If you are inside a cluster, differentiation is your problem regardless of what else the analysis says.

Empty space. The most interesting part of any map, and the most misread. Empty space is either an opportunity or a graveyard, and the map alone cannot tell you which. Before treating a gap as an opening, ask why nobody occupies it. Sometimes the answer is that nobody has tried. More often it is that the economics do not work, the segment cannot pay, or someone tried and failed quietly.

Trajectories. Static maps hide the most important information. Plot where each competitor was two years ago alongside where they are now, and the arrows often matter more than the positions — a company moving toward you is a different problem from one sitting still nearby. Hiring patterns and product releases are the leading indicators here; see building a watchlist.

Draw more than one

The single-map instinct is the biggest limitation of the format. Two axes cannot hold a market.

Three maps on different axis pairs will show different clusters and different gaps, and where they disagree is informative. A competitor who looks adjacent on one map and distant on another is competing with you in one segment and not another — which is exactly the kind of nuance a single map flattens away.

What it is not for

A positioning map shows relative position. It does not show market size, growth, profitability, or who is winning. A crowded quadrant may be crowded because that is where the money is.

Pair it with something that carries magnitude, or you will conclude that the empty corner is valuable purely because it is empty.

Our Competitive Intelligence report builds positioning maps on axes derived from how the category actually buys, with placements evidenced and trajectories included rather than a single static snapshot.