Skip to content
growth legendary.
← data
METHODOLOGY PREVIEW

Share of Recommendation: The Methodology

How we baseline and track how often your app is the answer when users ask machines what to install. This brief ships as methodology, panel findings publish once the tracking panel runs at scale.

methodology

This is the original methodology used by Growth Legendary's Share of Recommendation panel, launching Q3 2026. It defines prompt design, scoring, and scope.

01 · Prompt panel design

For each app category, we author a tracked-prompt panel of 30 install-intent prompts ('best budget app for indian users', 'best VPN for streaming', etc.). Prompts are versioned. Panels are refreshed quarterly.

02 · Assistant scope

Panel runs against ChatGPT (default web model), Perplexity (default), Gemini (default), and Copilot (default). Store surfaces monitored: Apple Intelligence app suggestions, Play's Gemini-picked collections, Apple Search AI-influenced results.

03 · Scoring

Each answer is scored on three dimensions: citation share (does your app get named?), position (where in the shortlist?), and sentiment (positive / neutral / caveated). Aggregated monthly by category and geo.

04 · Baseline + delta

Every engagement starts with a baseline panel run before any intervention. Deltas are reported monthly against baseline. We do not report absolute figures without their baseline anchor.

05 · What ships when

This methodology ships now. Baseline results ship per-engagement inside client dashboards. category-level findings ship as public briefs once the panel reaches sufficient scale.

work with the data

Get a baseline for your app.

Request a free app audit