What to Say at a Customer Advisory Board (Hint: Less Than You Think)

Most customer advisory boards are product demos with better snacks.

The vendor books a nice room. The product team brings a 60-slide deck. The roadmap slide has quarters on it. Customers nod politely for two hours, ask one or two questions near the end, and everyone goes home feeling vaguely good and slightly less informed than they hoped.

We just ran a CAB that went the other way. We ran out of time, not because the deck was too long, but because the conversation was too good. Customers argued with each other (in a good way). They told us things that weren't in any support ticket. A few of them raised their hands for pilots before we'd finished the session.

None of that was luck. It came from four deliberate choices about what to say, and more importantly, what not to say. Here's what we did, and what I'd do again.

1. Open with a vision they can see themselves in

The first 20 minutes set the tone for everything after. Most teams spend them on the company: the mission, the funding, the logo slide, the platform diagram with 14 boxes. Your customers don't care about any of that yet. They care about whether you understand their day.

So we opened with their words, not ours. The very first content slide put two real customer quotes on the screen. Not testimonials. "Honest feedback." One of them said, in effect, that they'd been long-time users and knew the product was due for a refresh. That's an uncomfortable way to start a meeting with your best customers. It's also the fastest way to earn the right to talk about the future, because it tells the room you were listening.

Then we gave them a vision built around them, not around our architecture. Three ideas carried the whole thing:

• A belief, stated plainly. Ours was that AI agents should make people better at the work only people can do, not replace them. Every later slide had to pass that test.
• A before and after they'd recognize. Traditional software records the work. What we're building gets the work done. Dashboards you interpret become suggested next actions. Generic queues become escalations that arrive with full context.
• An analogy that sticks. We described the shift as going from a paper map to Waze. A map gives you all the information and leaves the route to you. Waze tells you the next turn, reroutes when things change, and you can always ignore it. You're still driving.

That last line did more work than any feature slide. It answered the question every operator in the room was silently asking about AI: am I still in control?

One more thing we said out loud:

User experience isn't how the product looks. It's how much of your day it gives back.

If you only have time to fix one sentence in your CAB deck, make it the one that tells customers what's in it for them.

2. Give half the room back to them

Here's the uncomfortable math.

If you present for 90% of a CAB, you've paid to fly your best customers in so they can watch a webinar in person.

We spent close to half our time on structured exercises. Not "any questions?" at the end of a section. Actual activities, on the walls and tables, with sticky notes, dot stickers and fake money. It sounds a little like a team offsite. It works far better than a slide that says "Discussion."

We ran three:

Your first 15 minutes. Everyone wrote down the questions they need answered at the start of their day to know where to focus, one per sticky note, plus where they get the answer and how long it takes. Yellow for what they check today, pink for what they wish they could check. The rule that made it work: people count as systems. "I text my site manager" is a valid answer, and it's usually the most valuable one, because it shows you exactly where the software is failing them.

The autonomy dial. We put a long strip on the wall with four zones: AI does it silently, AI does it and tells me, AI suggests and I approve, always a human. Then we listed real tasks, like sending a tour reminder, approving an applicant who clearly qualifies, waiving a late fee. Each person placed a dot per task in their company's colour. The magic wasn't the average. It was the spread. When one customer puts a task at "silent" and another puts it at "always a human," you just ask them to talk to each other. That conversation is worth more than a month of discovery calls.

Buy what matters most. Everyone got $100 in fake money to spend across the capabilities we'd just shown, plus a blank card for anything we missed. First on their own, then as a table agreeing on their top two. Ranking tells you order. Money tells you intensity. And the table negotiation surfaces the "why" without you having to ask for it.

A few things made the exercises land:

• Ground rules up front. Tell us how it works today, not how it's supposed to. Disagree out loud. No fixes yet (hold your wish list, there's a place for it later).
• Clear roles. One person facilitated and kept time. One person captured themes live on a flip chart where everyone could see their own words. I framed each exercise and stayed out of the way.
• Placement matters. We ran the first two exercises before the demos, so we could point back at the wall during them: "Most of you put income approval at 'suggest and approve.' Here's how this handles it." We ran the money exercise after the demos, so people were spending on things they'd actually seen.

3. Show the roadmap by problem, not by quarter

The classic CAB roadmap slide is a calendar. Q1, Q2, Q3, with features dropped into boxes. It feels like transparency. In practice it's a list of promises you'll probably break, presented to the people most likely to remember.

We did two things differently.

First, we organized everything around problems, not features. Before showing a single screen, we put six customer problems on one slide, each written the way a customer would say it. Leads go cold before they sign. Every vacant day is rent you don't get back. Balances sit unpaid while staff spend hours chasing them. Then every section of the demo opened with one of those problems, in their words, before we showed what we're doing about it.

We said this shift out loud too: we'll walk through what we're building by the problem it solves, not the features it has. That sentence changes how a room listens. People stop evaluating your feature list against a competitor's and start asking whether you've understood their problem. That's a conversation you can win.

It also gave us an honest way to handle gaps. Two of our six problems don't have much built against them yet. Instead of hiding them, we gave them their own section called "Where we want your help." Customers lean in when you admit you haven't figured something out and ask them to shape it. The room had some of its best discussion right there.

Second, we only committed to now, next and later. Every capability carried one of three labels: live today, in development, or on the horizon. No quarters. No dates on the slide.

This isn't about dodging accountability. It's about being honest about what you actually know. You know what's live. You have reasonable confidence in what's being built. Anything further out is a direction, not a commitment. A calendar roadmap pretends those three levels of certainty are the same. Now, next, later doesn't.

The discipline goes both ways, though. If something is labeled "in development," every word about it has to say coming or being built, never available. And when a customer asks "when?" (they will), the right answer is a range and a reason, not a date you'll regret.

4. Close the loop before you leave the room

A CAB that ends on "thanks for coming" leaves its value on the table. We ended on two things.

A specific ask. We were picking a small group of customers to pilot two new capabilities, and we said exactly what that meant: a named champion, data in reasonable shape, and a short check-in every couple of weeks. In return, early access, a direct line to the product team, and a hand in setting the guardrails. The exercises had already told us who was ready. The dots on the autonomy dial made great conversation starters at the break.

A promise we can keep. Everything on the walls gets photographed. A "what we heard" summary goes to every attendee within two weeks. And at the next session, we show them what changed because of this one. That last promise is the one that gets people to come back, and to bring their honest opinions with them.

The short version

If you're planning a CAB, here's the checklist I'd steal:

• Open with your customers' words, even the uncomfortable ones.
• State one belief and one analogy that answer "what's in it for me?"
• Spend close to half the time on structured exercises, not open Q&A.
• Give exercises ground rules, a facilitator and a visible note-taker.
• Organize the roadmap by customer problem, written in their language.
• Commit to now, next and later. Leave the quarters off the slide.
• Give unsolved problems their own section and ask for help.
• End with one clear ask and one promise you'll keep.

The best compliment a CAB can get isn't "great presentation." It's running out of time because nobody wanted to stop talking.

So, dear product people: next time you're building your advisory board deck, count the slides. Then cut half of them and hand your customers a sticky note.