How we measure AI activity

Read by every major AI system

Three independent measures, not three steps of one journey. In the index is the share of listed products an AI crawler has read. Pulled for a live question is the share fetched by an assistant while it was answering someone. Sent a person is each assistant's share of the visitors who arrived from an AI chat. Bots are grouped by the company that operates them, and a company is named only when one of its own agents clears the privacy floor on its own.

Where software gets built

A country is placed by its builders' profile country, and only for builders who chose to show it. Territories fold into the state that administers them. A listing counts toward a country's weight only once it clears the peerPush quality floor, so signup volume alone cannot inflate a market. A country is named, ranked or rated only once enough distinct builders sit behind it and enough distinct builders sit behind the listings the rate is computed over, and it is given product chips only when it ships more of them than the map is willing to show, so a roster can never be read as a count. Everything below those floors renders as an unlabelled dot at a fixed size, carrying no country code at all, so neither bubble area nor page source leaks what the floor withholds.

What people cannot find

Each on-site search is attributed to the themes of the products it actually surfaced, then graded on how close the best match came. A theme's position is its rank among the plotted themes on two axes, how often it is asked for and how well it is answered, never a volume. A rank is relative by construction, so a theme is only called out as badly answered when its own share of weak matches also clears an absolute bar; in a window where nothing does, nothing is called out. Search text is never published: the theme is the unit, and junk or single-character queries are excluded before anything is counted.

What ships together

Across visible use cases on discoverable products we measure how often a pair of use cases travels together in one product, expressed as a share against the rarer of the two. Ribbon and arc width encode that weight; pairs below the privacy floor are never drawn.

Who is building, and market pulse

Who is building reports shares and ratios only: how concentrated quality activity is in the leading markets, how many of the last 30 days something shipped, and how far the highest quality pass rate sits from the busiest market's. Market pulse ranks categories on one of four independent reads at a time, over complete UTC days: peer trust as a depth-weighted rating out of five, and conviction, shipping and discussion each indexed to the leading category at 100. Those three are indexed rather than shared out because a product counts toward every category it belongs to, so the per-category tallies overlap and have no population total to divide by. A metric with too little to compare is dropped from the toggle rather than shown thin.

AI Surface Area (owner dashboard)

On private owner dashboards, AI Surface Area decomposes into four streams: AI crawlers indexing the product, traffic landing from AI chat assistants, programmatic API consumers, and MCP tool calls. Click-through events to external sites are deliberately excluded so the metric reflects discovery exposure, not downstream funnel behavior. Tiles are unweighted: the total is the simple sum of the four streams.

Hidden gems

Products with strong AI signal but low human visibility yet, in the last 30 days. Human visibility is a composite of upvotes, follows, and average rating. The thresholds re-calibrate as the catalog grows; the panel stays empty until the candidate pool is large enough to make percentile cuts meaningful.

Alternative lens (Switcher pressure)

On the per-alternative lens page, switcher pressure is the share of alt-page visitors who click through to a PeerPush challenger product. The page shows the top 3 challengers only; the full graph and verbatim queries are visible only on the verified owner's private dashboard.

Unmet demand

Three complementary signals over the last 30 days. PeerPush search is hybrid (pgvector cosine similarity combined with full-text search via reciprocal rank fusion). Even when the catalog has nothing genuinely close, hybrid search returns weakly-related neighbors - so a zero-result count is rare and not a useful signal. We capture three different angles instead.

• Weak match: queries where the top result's semantic cosine similarity is below 0.55 (configurable). Suggests the catalog has nothing genuinely close. Captured at insert time from the hybrid search path; the simpler full-text fallback path is excluded.
• Low CTR: queries with at least 50 impressions where the click-through rate is below 5%. Either the ranking is wrong, the taglines mislead, or the right product doesn't exist.
• Re-search: queries followed by a different query from the same anonymous visitor within 60 seconds. Strong behavioural “I didn't find it” signal. Within-day actor stability only - cross-day re-search is intentionally invisible.

Every tab applies the k-anonymity floor of 5 distinct visitors before a query surfaces, and every query string passes through the PII redactor.

AI citation freshness (owner dashboard)

On private owner dashboards, we track when each of the tracked AI agent buckets (Claude, ChatGPT, Perplexity, Gemini, Copilot) last cited the product. Severity bands: fresh under 14 days, aging 14-30 days, stale beyond 30 days, never cited otherwise. Buckets the product has not been cited by are surfaced explicitly so the gap is visible. “Cited” means at least one MCP result impression for the product attributed to an agent in the bucket.

AI coverage gap

For each of the top 30 categories (by total events in the last 30 days) and each tracked AI agent bucket (Claude, ChatGPT, Perplexity, Gemini, Copilot), we compute the share of approved + published Products in the category that have been cited at least once by that agent in the last 30 days. Categories with fewer than 10 approved products are excluded so single-product percentages don't skew the matrix. The panel is suppressed until total MCP citation volume crosses the privacy floor.