Data decay

Data decay is the ongoing loss of accuracy in contact and company records, and it runs on its own clock, not on a data provider's refresh schedule. Contact lists decay continuously, not on a schedule: people change jobs, mailboxes get deprecated, role accounts get retired, and an address that verified cleanly two months ago can bounce today. Static lists rot the day a team buys them, which is part of why signal-based sourcing has replaced pull-a-list-and-hope as the default.

Different fields decay at different speeds

Behavioral data, did this contact open three emails this week, view a pricing page twice, decays fast and belongs at the contact-record level. Firmographic data, headcount, funding stage, industry, barely moves month to month and belongs at the company-record level. Conflating the two is a common way a scoring model breaks: stale behavioral fields quietly turn an intent score into a measure of what happened last quarter, not this one, and nobody notices until the model stops predicting anything real.

The same decay shows up inside a CRM, not just in a purchased list. A rep updating last-contact-with-the-economic-buyer from memory, because the field lives in a dropdown, means the input degrades a little more every review cycle, even though the actual fact already exists in an email thread or a calendar. Any scoring framework built to survive contact with real pipeline needs decay logic built in on purpose, not bolted on after a model quietly stops getting used.