Telemetry moat
A telemetry moat is the advantage a company builds from the behavioral data it collects as a by-product of serving its customers. A model can be copied and a feature rebuilt in a weekend, but that data cannot be bought from public sources, and it grows with every customer.
Worked example
In December 2025 OpenAI took equity in Thrive Holdings. It paid with research staff and platform access, not cash, and in the deal for Shield, Thrive's IT-services rollup, it got the right to train its models on the platform's IT-services data. Read on the surface, Shield is a margin play on IT providers. Read as telemetry, it is a window into every small business that runs its IT through those providers: every ticket, incident, new user and installed application.
Metropolis shows the same pattern outside software. It bought SP+, the largest parking operator in North America: about 4,000 locations and $4 billion a year in payments through its camera checkout. The parking business pays for the sensors; the record of urban movement is the asset.
How it differs from
- A data flywheel. A flywheel says more users improve the product. A telemetry moat says the data is valuable to buyers beyond your customers, even if the product never improves.
- Telemetry in IT operations. Engineers use "telemetry" for logs, metrics and traces from systems. Here it means the economic behavior of the customers behind the systems.
- A data set you can license. Anyone can license a data set. Telemetry comes from serving the customer, so you can only get it by being their provider.
Questions
When does telemetry become a moat? Past a density threshold. One IT provider is noise; 200 across a national market are a view of the small-business economy that no software vendor holds.
Can the AI labs collect it themselves? Partly. Their agents inside companies see the workflow. They do not see what the business earns, owes and pays, and they cannot backfill years of history.
Do acquirers say this is why they buy? Rarely. They tell investors about margin expansion. The sign is that the most active buyers chase coverage, more firms in more regions, rather than cash flow per firm.
Origin
Set out by Viktoria Izdebska in The Prediction Economy, chapter IV (2026), and The Prediction Economy (3/4): Everyone is watching the wrong number.