Part II: The six domainsChapter summary

An agent looking for an organization starts where its principal’s request leads: a search index, an assistant’s own memory, a directory, a marketplace, a product feed or, increasingly, a registry of other agents and their tools. It reaches the organization’s own site last, if at all, and often only to confirm what it learned elsewhere.

The commonest failure is self-inflicted. Many organizations block agents without meaning to. A security rule written to stop scrapers, a bot-management service on its default settings or a robots.txt file copied years ago turns away the assistant a customer has just sent. The chapter shows how to tell crawlers from agents acting for a person, and how to decide, deliberately, whom to let in.

Assistants and searchDirectories and mapsMarketplaces and feedsAgent registriesYoursite
Figure 4. Where agents look, from the widest sources to your own site. Most of them reach you before your website does.

The second failure is absence. An organization missing from the directories, maps, marketplaces and feeds of its category is invisible to agents that rely on them. Being present is unglamorous work: claimed listings, consistent names and addresses, a catalog feed that is current rather than quarterly.

The third is fragmentation. Agents compare sources. When a hotel’s own site says one thing and three listings say another, the agent either guesses or moves on. Discover therefore ends where Describe begins: once agents can find you, what they find must agree.

Read the whole chapter

This is a summary. The full chapter is in the book, which is coming soon. One email when it is published.