Classification Value

Origin & evolution

The value inherits its lineage from positioning🎪PositioningGo-To-MarketProduct positioning statementView reference → and perceptual mapping. Once Al Ries and Jack Trout argued in *Positioning* (1981) that brands occupy coordinates in the customer's mind, the practical question became which coordinate, and a coordinate is a set of values, one per axis.

The statistical tradition formalised the placement. Perceptual maps built with multidimensional scaling, the psychometric technique developed by Torgerson, Kruskal, and others from the 1950s onward, position brands by deriving their values from customer similarity data rather than assigning them by assertion. Distance on the resulting map carries meaning: brands sharing similar values cluster, and the field reads that clustering as competitive substitutability.

Practice settled on a clean separation between the axis and the value. The axis defines the spectrum once; the values are the buckets along it that get applied many times. This matters because the buckets need🫀NeedUserA user need, pain, desire, or constraintView reference → to be mutually exclusive and collectively exhaustive to be useful. An axis of "company size" whose values are "small", "big", and "enterprise" is broken, because "big" and "enterprise" overlap, and a competitor could honestly land in either.

How it works in practice

A competitive team has settled a company-size axis with three values: SMB (under 50 staff), mid-market (50 to 1,000), and enterprise (over 1,000). They tag eight competitors. Six land in enterprise, two in mid-market, none in SMB.

That distribution is the finding. The SMB value sits empty across every rival, which says the lower end of the market is unserved by the current field. The team also tags their target persona as an anti-fit for the enterprise value: the buyer they want is a 20-person studio that enterprise tooling actively repels with procurement overhead. The values turn a vague sense of "we're more downmarket" into an exact, defensible claim about who sits where.

Classification value vs. its neighbours

• Classification axis is the dimension; the value is one point on it. "Deployment model" is the axis; "self-serve" and "sales-assisted" are its values. A value is meaningless without naming its axis, the way a coordinate needs to say which axis it measures.
• Competitor is an entity you tag with a value; the value is the label, the competitor is the thing labelled. One competitor carries several values, one per axis it has been classified on.
• Persona can be tagged as an anti-fit for a value, marking who the position deliberately excludes. That negative tag is as informative as a positive one, because it names the buyer you are choosing not to serve.

In the graph

In the Unified Product Graph, a classification value lives in the market intelligence region as the discrete label node. A classification_axis_includes_classification_valueClassification AxisincludesClassification Valuehierarchy edge binds each value to its parent dimension; competitor_classified_as_classification_valueCompetitorclassified asClassification Valuesemantic tags rivals; and persona_anti_fit_for_classification_valuePersonais anti fit forClassification Valuesemantic records deliberate exclusions. Modelling the value as its own node, joined upward to its axis and outward to competitors and personas, lets one shared bucket like "enterprise" do segmentation work across the whole graph, and makes an empty value (a bucket nobody occupies) visible as a queryable market gap.