Rank journalists, outlets, and creators by their reach and relevance among policy elites. Build a balanced target list and turn it into a client-ready recommendation.
Engagement Lift Index for Target Evaluation. BPI's proprietary measure of whether an account punches above its weight with the people who shape policy. Plenty of accounts reach this audience simply because they reach everyone. The ELITE Index isolates the ones policy elites seek out — where engagement runs well ahead of what the account's ordinary public footprint would predict.
It reads two things at once. Scale is how much of the audience an account reaches. Concentration is how much harder it lands here than with the general public. An account has to earn both: mass reach alone is a general-audience buy, and a high multiple on a tiny base is a whisper. A score in the 70s means this audience engages an account at a rate the public comes nowhere near.
| # | Account | Scale | Conc. | Reach | Reach # | Mult. | Lean | Score |
|---|
Scale and Concentration are each normalised 0–100 on a log scale across the accounts currently in scope, then combined as a geometric mean so an account has to clear a bar on both. Logs rather than percentile ranks, because reach is skewed hard enough that percentiles pin every major outlet at 99 and stop telling you anything. Scores move when the filters move — they describe standing inside the selected field, not an absolute constant. Say that to a client before they ask.
Paste or upload the accounts a client already works with. Names get matched against the audience graph, and you get back what that roster actually reaches — plus the accounts of comparable standing it's missing.
One per line. Handles, URLs and near-misses are fine — matching is fuzzy and every match is shown with its confidence so you can throw out the wrong ones.
Results appear here: match quality, what the roster reaches, where it leans, and the gaps.
The tool ships with the policy-elite audience. Load another one with the same columns and every tab re-scores against it. Your file gets cleaned up on the way in — the tool will attempt to correct for duplicates, non-identical acronyms, and errors, and it will tell you what it changed.
CSV with columns category, categories_all, media_tags, name, gen_pop_reach, all_reach, dem_reach, rep_reach. Everything is parsed in the browser; nothing is uploaded anywhere.
Three questions. You'll get a starting roster built to a sensible shape, which you can then edit like any other list.