The agent does the comparing
Start Monday by writing down every claim your brand makes, with the evidence for it in the column next to it. Elin moved house. Her assistant compared eleven suppliers in nine minutes and put two decisions in front of her. Her total attention: about forty seconds. Nobody visited a website, nobody saw an advertisement, nobody was persuaded.
That one example changes what you do all week. The journey no longer begins with the customer sitting down to orient herself. It begins with software acting on a model of her values, and your positioning becomes a claim that gets checked against what you actually publish.
List your claims and their proof
Open a document, write your three biggest brand promises on the left, and put the verifiable fact on the right. Best service, fastest delivery, always available. Each of those needs something an outside system can confirm without asking you. Where the right column stays empty, you have found this quarter's work.
- The promise, in the exact words your marketing uses today.
- The public source that confirms it: your terms, your service page, an independent review platform.
- The owner who keeps that source current.
Most teams finish this in an afternoon and are quietly shocked by how much of the right column is empty.
A claim that nothing outside your own website confirms will lose to a claim that something does.
Publish the facts where a machine can read them
Put your prices, delivery windows, cancellation terms and availability on a page that needs no login and no cookie wall. An agent that compares eleven suppliers in nine minutes reads what it can reach. Everything behind a form, a chat window or a PDF drops out of the comparison before a human ever sees the outcome.
Ask your web team for a list of every page where commercial conditions live. Then ask which of those pages are reachable in one request, without a session. That list is your backlog.
Rebuild the journey from the agent's side
Take your current journey map and draw a second lane underneath it: what the assistant does while your customer does nothing at all. In Elin's case that lane is nine minutes long and contains eleven comparisons. Your existing lane, the one with awareness and consideration in it, is empty for that whole stretch.
The two moments that are still human
Elin spent about forty seconds on two decisions. Design for those forty seconds and nothing else in the front half of the journey. What does the agent need to hand her so she can say yes with confidence? Usually three things: what it costs, what happens when she wants out, and what the switch will do to her week.
Measure the shortlist, not the visits
Change one number on your dashboard this week: count how often you appear in an agent's shortlist. Traffic and time on page describe a journey where a person browses. Elin never browsed.
You can start rough. Run the assistants your customers use, give them the situation Elin had, and record whether you come back. Do it weekly, log it, and watch the line. When you drop out, read what the assistant says about you and check that sentence against the claims document from Monday morning.
What you do on Monday
- Write the claims and the proof in two columns, and hand the empty rows to an owner with a date.
- List every commercial page that requires a login, a form or a cookie wall before it shows a price.
- Add the agent lane to one existing journey map, starting with your highest volume journey.
- Ask three assistants the question your customer would ask, and write down whether you appear.
- Bring both documents to your next steering meeting and ask for one decision: who fixes the empty rows.
The agent lane is easier to draw when you have the shape in front of you. The Agentic Journey Map gives you that lane and the questions per step, so you can fill it in with your own team instead of designing the format first. Bekijk de materialen