Retrieval probability is a coverage problem (do you have a passage that can win facet i at all). Selection strength is a quality and authority problem (given you were retrieved, does your audition chunk beat the others). Expected citation = the average of retrieval x selection across the fan-out.
Your estimate updates as you type.
The formula that replaces rank tracking
Rank tracking answers a question the engine no longer asks. Expected citation decomposes into two probabilities you can influence: retrieval (coverage) and selection (quality plus authority). Most teams spent a decade optimising selection for a handful of queries while leaving retrieval at zero for most of the fan-out. This estimator shows why covering one more facet usually beats polishing an existing one.
How to use it
- Enter your topic's facet count and how many you cover.
- Set an average selection strength.
- See expected citation rate and the lift from covering one more facet.
Frequently asked questions
What is expected citation?
The metric that replaces rank tracking under fan-out: the average, across a topic's sub-queries, of retrieval probability (do you have a passage that can win the facet) times selection probability (does your chunk beat the others). It captures coverage, not just position.
Why does coverage beat rank here?
Because each facet is a separate contest. Ranking first on one facet buys a small edge; answering a second facet at all roughly doubles your chances of being cited. The estimator shows that covering one more facet usually outperforms polishing an existing page.
Is the Expected-Citation Estimator free to use?
Yes. The Expected-Citation Estimator is completely free, with no sign-up, no usage limits and no watermark on the output.
Is my data private?
Yes. The Expected-Citation Estimator runs entirely in your browser. Nothing you paste or enter is uploaded, stored or sent to any server.
Related tools and reading
Keep going with the playbooks and tools behind this one.