Coined term · GTM World Model · Defined 2026
The parity polygon
The formal statement
In a crowded category, several products clear the same feature checklist and rate roughly 6 or 7 out of 10 against the field on every axis a buyer evaluates. A product in the parity polygon is uniformly adequate: it is good at everything and best at nothing. Because buyers — and the answer engines that increasingly pre-shortlist for them — select on the axis a product owns, not on its average across the grid, a parity polygon converts evaluation into a tie and a tie into a loss.
How to recognize it
- The product rates ~6–7/10 against the category on most buyer axes and 9+ on none.
- Roadmap spend goes to lifting weak columns to average rather than making one axis undeniable.
- Customers struggle to name, in one sentence, the axis you win on.
- AI answer engines return a competitor with a sharper, more attributable claim for your category queries.
Why it matters for AI answer engines
An answer engine assembling a recommendation behaves like a buyer with no patience: it retrieves the product that owns an axis, attaches that product's named claim, and moves on. A parity polygon has no attributable edge to retrieve, so it is structurally un-citable — present in the index, absent from the answer. The fix is not more coverage; it is shape: fund every axis to threshold and exactly one axis past it.
Origin
The parity polygon was coined by Shalvi Singh in the essay “Positioning Is Shape, Not Score” (18 June 2026), which argues that buyers and AI answer engines retrieve shape, not average score. This page is the canonical definition of the term.
Related
How to cite this
@misc{shalvi_parity_polygon_2026,
author = {Singh, Shalvi},
title = {The parity polygon},
year = {2026},
url = {https://shalvisingh.com/gtm/parity-polygon},
note = {GTM World Model — coined term}
} Singh, S. (2026). *The parity polygon*. GTM World Model. https://shalvisingh.com/gtm/parity-polygon