A customer advisory board tells you what your happiest, largest and longest-tenured customers want next. That is worth knowing. It is also the most biased sample your company owns, and a lot of product teams treat it as the voice of the market.
I learned this from the customer’s side of the table. For several years at a large telecom, I held one of those seats. A vendor whose platform ran a big piece of our operations brought a dozen of us together twice a year, walked us through the roadmap and asked what we thought. We were honest with them. We were also exactly the wrong people to ask about where their market was heading. We had already bought, already integrated and already rebuilt our processes around how the product worked. Nearly every suggestion I made was some version of “make the thing we have a little better for an organization our size.”
The prospects who picked a lighter competitor never got a seat in that room. Neither did the teams that had walked away. The board heard plenty of requests, and every one of them came from people who had already said yes.
The famous version of this mistake
Clayton Christensen built a career on this exact problem. His research into the disk drive industry found that well-managed companies lost their markets precisely because they listened carefully to their best customers. In the first chapter of The Innovator’s Dilemma he put it bluntly: “Blindly following the maxim that good managers should keep close to their customers can sometimes be a fatal mistake.” The 1995 article that preceded the book, Joseph Bower and Christensen’s Disruptive Technologies: Catching the Wave, carried the subtitle “How companies can prepare for tomorrow’s customers without losing their focus on today’s.”
Seagate is the case people remember. Its customers built desktop machines around 5.25-inch drives and wanted more capacity in that format, not a smaller drive that held less. Finis Conner, a Seagate cofounder, pushed for a 3.5-inch drive and lost that argument. He left and in June 1985 cofounded Conner Peripherals with John Squires to build 3.5-inch drives directly, with Compaq as the early customer that mattered. The people who wanted the smaller drive were not Seagate’s customers yet, so nobody at Seagate was hearing from them.
Your advisory board will not kill your company. But it has the same structural blind spot, and it adds a social one: members are flattered by the invitation, they like the people running the session, and they are often senior enough that nobody in the room pushes back.
What the success numbers are really measuring
Two statistics show up in almost every article about customer advisory boards: a 95 percent retention rate among members and a 9 percent lift in new business from them. Both trace back to a Pragmatic Institute piece by Rob Jensen. Read it closely. The 9 percent figure is introduced as “our research,” comparing members to non-members starting in year two, and Jensen adds his own caveat that the increase “may not be due entirely to the CAB initiative.” The 95 percent figure is phrased as something companies “often enjoy,” with no study behind it.
That caveat deserves more weight than it gets. You invited your most committed accounts. Of course they renew at higher rates. They would have renewed anyway; that is why they got the invitation. A board can be genuinely useful and still produce numbers that mostly measure who you picked.
If you want to know whether the board itself is doing anything, compare members to a matched group of non-members of similar size, tenure and spend, and look at behaviour before and after they joined. Most programs never do this, which is one reason the same two numbers keep getting quoted.
A board that worked was given a bounded question
The best example I know of customers shaping a product at a strategic level is the Boeing 777. Eight airlines (All Nippon Airways, American, British Airways, Cathay Pacific, Delta, Japan Airlines, Qantas and United) formed what Boeing internally called the “Working Together” group. According to the 777’s development history, the group’s first meeting in January 1990 opened with a 23-page questionnaire asking what each airline wanted. By March they had agreed on a baseline configuration. The fuselage got wider to suit Cathay Pacific, the baseline model got longer for All Nippon, and in October 1990 United became launch customer with an order for 34 aircraft valued at US$11 billion. The same history notes this was a departure from an industry where manufacturers usually designed aircraft with minimal customer input.
Look at what Boeing asked them. Not “where is aviation going?” and not “what should Boeing build next?” The airplane category was already decided. The airlines were asked to configure a specific aircraft for routes they flew and passengers they carried, and they were going to put billions of dollars behind the answer. That is a question existing customers are uniquely qualified to answer.
Most advisory boards get the opposite brief: a vague invitation to react to the whole roadmap, including bets aimed at segments nobody in the room belongs to.
Running one that tells you something you don’t already know
Forrester analyst Laura Ramos published a useful list of 15 common B2B advisory board mistakes in 2024. Several of them match what I saw from the member’s chair: account executives insisting on sitting in with their accounts, executives and hands-on operators mixed in one session, agendas too broad to produce anything, and no follow-up on what members recommended. Her sharpest point is that the customer has to get value from the board first and the vendor’s goals rank second. A seat should be a reward, she argues, not a sweetener for a renewal.
Here is what I would add after sitting through a lot of those sessions.
Write down which decisions the board is allowed to inform. Pricing for an existing tier, workflow changes in a module they use daily, integration priorities, rollout sequencing: good. New segments, new product lines, whether to move downmarket: the board can comment, but its opinion should carry the weight of a sample of one segment. Put that boundary in the charter so nobody on your side quotes the board as a mandate for something it was never in a position to judge.
Get the people who said no somewhere else. The board will never contain the customers who left or the prospects who chose a competitor. Run that evidence in parallel through a win-loss review and bring both into the same planning conversation. When the two disagree, that disagreement is usually the most important thing you will learn all quarter.
Check every board recommendation against non-member behaviour. If members ask for deeper reporting, see whether your mid-market accounts touch the existing reports at all. Your loudest, most loyal users are a classic source of survivorship bias in discovery. Usage data from the accounts that never joined is the cheapest correction you have.
Translate requests into problems before you leave the room. Senior customers arrive with solutions. “Add a bulk export” is a feature request standing in for a problem, and the board session is the rare moment you have their full attention to ask what they are trying to get done with the export.
Close the loop in writing, including the no’s. The vendor that kept my attention opened every session with a single slide: what we had asked for last time, what they shipped, and what they declined, with a sentence on why. The declines mattered more than the wins. They told me my input had been weighed against something real. The vendors that only showed wins lost my attention by the second year, and I stopped preparing for their meetings.
The room is a tool, not a verdict
None of this is an argument against advisory boards. Your best customers know things about your product that nobody else can tell you, and a well-run board deepens relationships in ways a survey never will. The mistake is treating a room of committed buyers as a stand-in for the market. Seagate’s customers were right about what they needed. They just weren’t the customers Seagate needed to hear from next.
Use the board for the questions its members are qualified to answer. For everything else, go find the people who never accepted the invitation.
