How the desk decides what is worth watching

The watching infrastructure cannot cover everything that exists. The number of sources is large and growing; the number of potentially relevant signals across them is larger still. The desk's decision about what to watch, and how often, is one of the more consequential parts of the work.

The first criterion is plausibility of relevance. A source where information about the client is likely to appear (a regulatory filing system that covers their industry, a press of record that covers their region, a database that indexes the kind of records their structures generate) is watched. A source where information is unlikely to appear is not.

The second criterion is speed of update. A source that updates daily is watched daily. A source that updates monthly is watched monthly. A source that updates rarely is watched on a slower cadence, with the understanding that significant changes will not be missed because the watching window is wide enough.

The third criterion is signal-to-noise. A source where most updates are noise (most of what appears does not matter to any one client) is watched in summary form, with the watching set to flag a much narrower category of update that is likely to be material. A source where most updates are signal is watched in fuller detail.

The fourth criterion is leading versus lagging. A source that gives early warning of something that will later appear in more visible records is watched ahead of the lagging ones. The earlier a movement can be identified, the more time there is to consider a response.

The desk reviews the watching configuration regularly. The world changes; the sources change; the categories of relevance shift. The watching plan from a year ago is not the watching plan of today, and the work of keeping the configuration current is part of keeping itself.