You Shouldn't Find Out Where You Land in the Magic Quadrant From the Magic Quadrant

Every analyst-relations team lives the same annual ritual: months of briefings, a questionnaire the size of an RFP, then silence — and then a PDF that fixes your market position for a year. The placement lands as a verdict. But the evidence the analysts scored didn't appear the week the report shipped. It accumulated all year, in public: releases, hiring, pricing changes, customer reviews, the competitors' own briefing decks.

Which means the placement is, in principle, forecastable — if you hold the evidence in a structure that can be scored the way the analysts score it.

Six frameworks, one graph

PYRAMYD reads analyst placements into the product graph across six frameworks — Gartner Magic Quadrant[1], Forrester Wave[2], IDC MarketScape[3], Everest PEAK Matrix, Gartner Critical Capabilities, and ISG Provider Lens. You upload the editions your company is licensed for; they stay in your tenant. Each vendor's dot stops being a picture and becomes data: axes, segment, and — where the report publishes them — criterion-level scores, joined to the companies and products already in the graph.

Every placement records how it was measured: read from the figure, computed from published sub-scores, or estimated from the segment — so a downstream consumer always knows whether it is looking at a measurement or an inference. Evidence that does not exist is never invented.

The grain law

The subtle failure mode in scoring yourself against analyst criteria is mixing grains. A Wave's Current Offering criteria score the product; its Strategy and Market Presence criteria score the vendor.[2] Product evidence — features, reviews, release cadence — can support the first family and says nothing about the second. So the engine enforces what we call the grain law: only product-axis criteria are ever scored from product evidence. Vendor-axis criteria are surfaced by name for a human to reason about, and never folded into the score.

Projecting the next edition

With placements as data and live evidence beside them, the forecast becomes a two-sided synthesis — the same engine that predicts competitor moves. For each vendor in your market it combines the current position, its momentum in our signals (release cadence, hiring, review trajectory, expansion), and the report's own narrative, and produces the next edition's movement: a predicted segment and signed deltas on each axis, with the drivers named.

We predict the movement, not the level — the forecast applies to the position in the edition you actually ingested, so it survives contact with your real entitlements.

And because the engine is two-sided, it does not stop at prediction. For your own company it produces prescriptions: which evaluation dimensions to improve, in what order, to move the needle before the analysts score it — each tied to the criteria it affects and the evidence that would have to change.

What this is not

• Not analyst redistribution. You upload what you are licensed for; it stays in your tenant; nothing is republished.
• Not a certainty. Forecasts ship with confidence attached, are logged with the indicators that would confirm them, and are scored against the next edition when it arrives.
• Not a black box. Every projected move names its drivers, and every driver traces to graph evidence with a source URL.

The question this changes

Analyst relations has always been played in hindsight: the report drops, you diagnose, you brief harder next cycle. A forecastable placement turns it into the same game the rest of PYRAMYD plays — see the move coming, price it, and act while acting still changes the outcome.

See the six frameworks and the crosswalk →