Measured studies of how agents actually transact with commerce on the open web. Every result is published in full, with the method and the dataset, so it can be re-derived rather than taken on trust.
PlatformDTC Research
A replication at 2.5 times the original sample. Seventy of 147 direct-to-consumer brands serve a live agent-commerce discovery document — 48% of the full list and 65% of those that answered a request at all. The headline survived the wider frame; two supporting numbers did not, and both are corrected here.
Read the study →PlatformDTC Research
Of 58 well-known direct-to-consumer brands, 31 serve a live Universal Commerce Protocol discovery document, and every one answers a Model Context Protocol request with thirteen working checkout tools. An agent can transact with over half the sample today. Most of those brands did not build it.
Read the study →PlatformDTC Research
The usual framing of llms.txt is that it is a cheap bet nobody has taken yet. Measured across 58 direct-to-consumer brands, 73% of reachable hosts already serve one — and 70% of those did not write it. The file most of them publish is not a summary document but a protocol handshake.
Read the study →PlatformDTC Research
Most current advice assumes a store's llms.txt indicates how prepared it is for agents. Requesting /llms.txt and the discovery path on the same 58 hosts shows the two are only loosely related: eight stores accept a full agent checkout while publishing an llms.txt that never mentions it, and four of those publish no llms.txt at all.
Read the study →PlatformDTC Research
Of 147 direct-to-consumer brands, 59 serve a real agents.md, and 57 of those instruct a reading agent to recommend that its user install a cross-store shopping skill operated by the platform — and to prefer that skill over interacting with the store directly. The merchants did not write it.
Read the study →PlatformDTC Research
Twenty-three direct-to-consumer product pages scored for how well a machine reading raw HTML can understand them. Median score 100 out of 100, with effectively no middle of the distribution: pages either emit schema.org Product markup and land at the top, or emit none and sit in a narrow band below.
Read the study →We're open to research collaborations with academic institutions and industry partners who share our vision for safe and beneficial autonomous systems.
Contact Research Team