Marain
← Work

The estate

A real-time attention index fused from many public signals

Six noisy signals, one decision-ready score

March 10, 2026

search interestencyclopedia trafficresultssocial engagementnews coverageavailabilityone score
Six noisy signals, one decision-ready score, with the weights on the outside where the argument belongs.

Find the thing nobody is counting, make it countable, then build the machine that counts it. This one is the third step, running unattended.

The index tracks digital attention across nearly a hundred entities by fusing six independent public signals: search interest, encyclopedia traffic, results, social engagement, news coverage and availability. They become one configurable score, refreshed every few hours.

How it is built

Collection is fully decoupled from display. A scheduled pipeline fetches every source on a cadence and writes compact data files. A lightweight application loads those into an in-memory analytical database, which is what makes league overviews, head-to-head comparisons and movement alerts instant rather than queued.

Why it is here

Most insight work fails because the signals are noisy and scattered, and the analyst quietly picks the one that agrees with the meeting. The value is in the fusion: many messy public sources turned into one metric a decision-maker can act on, kept current without anyone touching it, with the weighting exposed so the argument is about the weights rather than about the data.

The same argument, elsewhere

Five other pieces on this site make an argument this one is also making. The sentence under each is quoted from that page, which is the only reason to believe the pairing.

MeasurementWhat counts as evidence

InstrumentsBuild the standing thing

MachinesRead at scale, keep it home

All six threads, all thirty-eight pieces →

Have a problem shaped like this?