Customer Advisory Board

For a long time, what a large customer wanted arrived through one person. An account manager came back from a visit saying the client needed approval workflows, and the product team had to judge whether that was the company's position or one administrator's opinion.

The familiar scene goes like this. Sales says a named account won't renew without a particular feature, and the room goes quiet. Nobody has grounds to argue, so the feature lands in next quarter's plan, and six months later the only person at that account using it is the administrator who asked. The reverse happens too: someone sorting through churn reasons notices three departed accounts left for the same cause, long after the contracts ended.

A customer advisory board moves both of those. Opinions arrive before the decision instead of after it, and they arrive in a room with several customers in it rather than through a single channel. When four of six companies name the same gap, that is a market signal rather than a support ticket. When one names it, that is one company's internal situation. Putting them in the same room is what separates the two.

One quarterly session runs through five stages.

1. Choose the members. Eight to twelve companies is typical. Start with the largest accounts, but don't fill the room with the ones who like you. The customers with complaints say the most useful things, so mix industries, company sizes, and depth of usage.
2. Invite them and agree on the commitment. Four sessions a year, roughly two hours each, scheduled in advance. Attendance holds only when the people coming get something for the time. Seeing the product roadmap before it is public, and hearing how similar companies use the product, are usually what that something is.
3. Send the agenda two weeks ahead. Narrow it to two or three topics. Handed out on the day, it produces reactions; sent early, it gives members time to collect opinions inside their own companies.
4. Run the session. Open by reporting what was adopted from last time and what was set aside. Then share the roadmap, and spend most of the remaining time on discussion. A session heavy on presentation turns into a briefing, and briefings lose their audience by the next round.
5. Reply within two weeks. Write the feedback up as a list, mark each item adopted, under review, or declined, and give a reason for anything declined. Skip this step and the whole meeting loses its point, because silence reads as proof that the two hours changed nothing.

Stages 1 and 5 are where this practice breaks most often. Invite only the comfortable accounts and you spend two hours collecting compliments. Delay the follow-up and attendance halves next quarter. Hold a second session with nothing adopted from the first and the room visibly empties.

From the customer side, one or two people per company attend, and the useful profile is someone who uses the product and has a say in decisions about it. Send only daily users and the discussion fills with requests about screen details. Send only an executive with the budget and nobody knows the product well enough to go deep. A department head running the work usually sits between the two.

Three roles attend from your side. The product lead presents the roadmap and takes the questions. A customer success manager facilitates and takes notes. The chief executive appearing at least once signals that the session carries weight. Account managers may sit in, but they don't bring up contracts. The moment this meeting sounds like a sales call, people stop saying what they think.

Timing splits into two questions. One is frequency: quarterly is the norm, dropping to twice a year when the preparation load is too much. The other is whether to run one at all, and the arithmetic only works when there are few customers and each pays a lot. Ten accounts at $38,000 a year put $380,000 of revenue in one room, which justifies the two hours easily. Thousands of users paying $8 a month cannot be handled the same way; there you collect voice of customer (VOC) — the practice of gathering customer opinions and complaints and feeding them into the product — through surveys and usage data instead.

"Four of the six companies at this session put access controls first."

Not one customer's request but a shared one, which is grounds for reordering next quarter's plan.

"Let's put that on the advisory board agenda. We don't know enough to decide it internally."

A single feature request nobody can settle, so it goes to the customers directly.

"Can we take the advisory board members out of the at-risk group?"

A question about whether the thicker relationship should change the renewal forecast — the exercise of judging in advance which customers will continue their contracts.

Pagedeck started running a board of nine enterprise accounts last year. The most common request at the first session was an access log, an item sitting around twentieth on the internal list at the time. Each company had filed it separately as a support ticket, so it had never been visible as one demand.

Sofia says her conversations with those nine accounts changed afterward. She used to visit two months before renewal to ask what was going badly; now she has heard it every quarter already, and spends that meeting on next year instead. Grace opening each session in person started for the same reason.

Some read the board as hospitality — a nice room, a good meal, and the largest customers kept happy. The picture comes easily, but the people attending are spending their own working hours. If all they take away is lunch and a handshake, they send a deputy the second time. What keeps the practice alive is members seeing their own words show up in the product.

The other misreading is that everything raised has to get built. Asking for opinions feels like it creates an obligation to honor them. In practice the board is still a slice of the customer base, and what a large account needs is often an obstacle for the hundreds of smaller ones. Take the transcript straight into the plan and the product starts to resemble a few companies' internal systems.