# Agent readiness: [quarter, year]

- **What's changing.** Customers are starting to ask AI assistants to research, compare and buy for them. Those assistants visit [your site] the way a customer would, but read it differently. If one can't get in, can't read our prices or can't complete a purchase, it recommends someone else, and the lost customer never shows in our reports.
- **Where we stand.** AgentScore [score] out of 100 ([band]), from [last quarter's score] last quarter. Main issues: [two or three findings, in plain words]. AI agents [can / can't yet] complete [our most important journey, such as checkout] today.
- **What we're doing.** [Fix 1]: [owner], done by [date]. [Fix 2]: [owner], done by [date]. [Fix 3]: [owner], done by [date].
- **How we'll know it's working.** Each quarter we report the trend in AgentScore by category, the pass rate of our key journeys tested with AI agents, agent traffic from our logs (verified and impostors), and visits and orders from AI assistants. We'll present AI referral revenue as a floor, not a total.

**The ask.** [Time, tools or decisions for this stage, such as developer time for the fixes above], reviewed on [date]. One accountable owner: [name, usually the head of e-commerce or digital]. A decision from legal and security on AI training crawlers and verified agents by [date].

---

**Notes for the presenter**

- *Are we letting AI companies take our content?* No. Training crawlers and the assistants that fetch pages for a customer use different names, so we can block one and allow the other. That's a policy choice we make once, with legal.
- *Does this open us up to bots?* Not if it's done properly. We let in the agents we want, checked by verified identity, and keep blocking the rest.
- *What happens if we do nothing?* Probably nothing we'll notice, which is the problem. Some customers' assistants will fail on our site and choose a competitor.
- If you quote any outside figure, cite the source on the slide. If you can't, describe the trend in words instead.
