Plumb measures what AI does to your content — before any session — and turns it into refresh, write-new, and retire decisions ranked by recovered revenue.
Plumb’s value scales linearly with what one well-placed article is worth in your business model. For affiliate and B2B vertical-trade publishers, the math is straightforward. For general programmatic publishers, it is not.
Plumb surfaces a finance query cluster where your citability score is 41/100 and competitors hold 70% of Perplexity citations. Refresh the existing piece + commission two adjacents.
Plumb surfaces an emerging cluster trending +340% in 4 weeks across Perplexity and ChatGPT, with only 3 citation voices — two stale. Commission a 2,800-word authoritative piece before the field fills in.
Civic and regional news doesn’t compete on AI answer surfaces. The commissioning workflow doesn’t apply. Plumb generates CPM protection value (agentic traffic detection) but not revenue restoration.
Plumb joins first-party crawl telemetry, GSC AIO data, sampled citation probes, and structural citability scoring into a ranked weekly work order.
“Your article on student loan refinancing is being retrieval-crawled by PerplexityBot 6×/month and hasn’t been visited in 3 weeks. GSC shows AI Overviews rendering on this cluster; your CTR dropped from 11% to 2%. Bankrate and NerdWallet hold 70% citation share. Citability score: 41/100 — 480-word intro before the answer block, no Q&A structure. Commission: refresh with a 150-word answer block in the intro. Est. recovery: 150 AI-referred sessions/month.”
No aggregation window, no sampling. Every AI crawler, as it happens. Retrieval vs. training classification on each hit. When a blocked bot gets through, you see it this second.
See the live viewProfound and Peec serve brand buyers. Chartbeat serves traffic analytics. Plumb serves publisher editorial operations — an uncontested segment in the current landscape.
| Capability | Profound | Peec | Chartbeat | Plumb |
|---|---|---|---|---|
| Publisher editorial workflow | — | — | — | |
| First-party crawl log analytics | — | — | ||
| GSC AIO dimension (cannibalization) | — | — | — | |
| AI citation panel | — | |||
| Structural citability scoring | — | — | — | |
| Commission engine (write/refresh/retire) | — | — | — | |
| Closed-loop outcome tracking | — | — | — | soon |
| MCP server (read + write) | read-only, gated | read-only | — | 13 tools |
| Brand sentiment / share-of-voice | — | — | ||
| Real-user conversation panel | — | — | — |
Three confidence tiers on every metric — measured (first-party edge), sampled (citation probe), inferred (probabilistic model) — so you can defend the number in a planning meeting.
One design partner with real Cloudflare logs, real affiliate or B2B newsletter economics, and 12 weeks to validate whether the commissioning workflow generates traceable revenue. Walk-away rights either side.