How to Choose the Axes for a Positioning Map
Choose two axes that pass three tests: each must be buyer-relevant (a dimension customers actually weigh when choosing), the pair must be independent (not two labels for the same thing), and each must be discriminating (it spreads competitors out). Fail any one and the map flatters you while telling you nothing.
Quick Answer: Good positioning-map axes are buyer-relevant, independent, and discriminating. Start from your best-fit customer's real decision criteria, avoid correlated pairs like price versus quality that collapse onto a diagonal, and keep only axes that actually separate competitors. If an axis fails a test, replace it.
The axes are the highest-leverage decision on a perceptual map: they decide what counts as "open space" before you plot a single competitor. Choose them from what buyers care about and honest gaps appear on their own. Choose them to make yourself look good and you produce a flattering but useless map — an empty corner you drew, not one the market left open.
Prerequisites: know your best-fit customer's decision criteria
Before you can choose axes, you have to know who your best-fit customer is and what they actually weigh when they choose — because the axes are simply those criteria drawn as lines. Skip this and you are guessing at dimensions instead of measuring them.
April Dunford's Obviously Awesome frames positioning as context: you are defined relative to the alternatives your best-fit customer would otherwise choose, and the value you deliver that those alternatives don't. That reframing hands you the raw material for axes — the dimensions on which your best-fit customer compares their real alternatives are exactly the dimensions worth putting on a map.
Pull candidate criteria from evidence, not a whiteboard brainstorm:
- Win/loss interviews — the reasons buyers give for picking you or a rival name the criteria that actually moved the deal.
- Sales-call questions and objections — what prospects probe before buying is what they weigh.
- Review mining — the attributes customers repeat, in their own words, across public reviews.
- Support tickets and churn notes — where the value you promised did or didn't land.
Notice these are the same inputs that feed the plot itself. Choosing axes and building the map step by step are two halves of one research exercise, not separate stages.
The three axis tests: relevance, independence, and discrimination
Run every candidate axis through three tests, in order — relevance, independence, discrimination. Relevance and discrimination judge each axis on its own; independence judges the pair together. An axis has to pass all three to earn its place on the chart.
Test 1: Relevance — does the buyer actually decide on this?
An axis is relevant only if it is a criterion your best-fit customer genuinely weighs when choosing. Labels like "innovative," "best-in-class," or "market-leading" fail on sight: buyers don't rate you on them and don't buy because of them. Relevant axes are the concrete trade-offs behind a real decision — how fast you reach value, how much control you keep, how much hand-holding you need.
Test 2: Independence — are the two axes measuring different things?
The two axes must vary independently, or they collapse into a single diagonal and waste half the chart. This is the classic price-versus-quality trap: because buyers perceive cheap as lower quality and premium as higher, nearly every brand slides onto the diagonal, leaving two corners structurally empty. That "high quality, low price" gap is a mirage created by correlation, not an opening the market left for you.
Test 3: Discrimination — does the axis spread competitors out?
A discriminating axis is one competitors score differently on; if everyone lands at the same value, the axis is flat and tells you nothing. "Has a mobile app" might be relevant and independent, but if every rival ticks it, it separates no one. Keep the axes where the field genuinely spreads out — that spread is what turns the map from a decoration into a diagnosis.
Bad axis pairs and why they mislead
The most common bad axis pairs fail in one of two ways: they correlate, so competitors bunch on a diagonal, or they flatter, so you land alone in a corner no buyer is actually shopping in. Both produce empty space that looks like opportunity and isn't.
The table below contrasts axis pairs that quietly sabotage a map with pairs that earn their place, and names the specific reason for each verdict.
| Axis pair | Verdict | Why it lands there |
|---|---|---|
| Price vs. quality | Misleads | Correlated — competitors collapse onto a diagonal, faking two empty corners |
| "Innovative" vs. "trusted" | Misleads | Vanity attributes buyers can't rate and don't decide on |
| Feature count vs. price | Risky | "More features" isn't a buyer value, and feature count often tracks price |
| Ease of setup vs. depth of control | Works | Independent, buyer-relevant, and a real trade-off that spreads the field |
| Speed to first value vs. breadth of use cases | Works | Two dimensions buyers genuinely weigh against each other |
Takeaway: A good axis pair captures a tension buyers actually live with, where no answer is universally "more is better." When you catch yourself choosing axes because they park you alone in an empty quadrant, stop — you are drawing the flattering-but-useless map, not reading the market. The honest test is whether a buyer would nod at both dimensions, not whether you like where you land. Axes that pass all three tests do something more valuable than locate you: they surface where you genuinely differ, which is the bridge from a position to a defensible edge — the subject of positioning versus differentiation.
Iterate: draw a second map when the first pair fails
Axis selection is empirical, not a one-shot guess. If your first pair bunches competitors together or buries your edge, swap an axis and redraw — expect to sketch two or three candidate maps before one both spreads the field and puts your intended position in genuinely open space.
Treat each candidate as a hypothesis. A pair that passes relevance and independence on paper can still fail discrimination once you plot real competitors and watch them cluster. Because that spread only appears against real placements, the dots have to come from customer perception rather than memory — confirming which dimensions buyers actually weigh, before a launch rides on the map, is the kind of cheap validation Edmired is built for.
When a map underperforms, change one thing at a time:
- Replace a flat axis with a criterion that actually separated winners from losers in your win/loss notes.
- Break a correlated pair by swapping one axis for an unrelated dimension buyers also weigh.
- Sharpen a vague axis from a fuzzy label like "quality" to the specific thing buyers mean by it, such as "reliability under load."
One caution: iterate to find axes buyers recognize, not to engineer a flattering empty corner. If you keep redrawing until you happen to stand alone, you have fit the axes to your story instead of the market. Trust the map only when both dimensions would survive a customer reading them back to you.
Key Takeaways
- Axes are the highest-leverage choice on a positioning map. They define what "open space" means, so a lazy pair manufactures fake gaps and a flattering pair hides the truth.
- Relevance comes first. An axis belongs on the map only if it is a criterion your best-fit customer actually weighs when choosing — not a vanity label like "innovative."
- Independence prevents the diagonal. Correlated pairs like price versus quality collapse competitors onto one line and fake two empty corners; keep axes that vary separately.
- Discrimination earns the axis its place. If every competitor scores the same on a dimension, it is flat — drop it for one that spreads the field.
- Start from evidence, not a brainstorm. Win/loss interviews, sales objections, and review mining surface the real decision criteria that make good axes.
- Obviously Awesome sets the frame. Positioning is relative to your best-fit customer's alternatives and the value you add, so the dimensions they compare on are the dimensions worth mapping.
- Iterate honestly. Redraw with new axes until competitors spread, but never torture the axes just to land yourself alone in an empty corner.
Frequently Asked Questions
How many axes should a positioning map have?
Two. A positioning map's whole value is being readable at a glance, and two axes create the four quadrants people can reason about instantly. If you genuinely need to compare competitors across many attributes at once, switch tools to a Blue Ocean strategy canvas, which plots numerous value factors as a curve instead of forcing everything onto two lines.
Why shouldn't I use price and quality as my two axes?
Because they correlate. Buyers tend to perceive cheap as lower quality and premium as higher, so nearly every brand slides onto a diagonal and the off-diagonal corners sit empty for structural reasons, not market ones. The tempting "high quality, low price" gap is usually a mirage. Use two dimensions that vary independently instead.
How do I know if my axes are good?
Test each candidate against three questions: does the buyer actually decide on this (relevance), do the two axes measure different things (independence), and does the axis spread competitors apart (discrimination)? An axis that fails any one wastes half the chart. The fastest real-world check is whether your best-fit customer would recognize both dimensions as things they weigh.