In the second week of October, a VP of Customer Success starts planning next year's customer advisory board. The CEO suggests inviting the ten largest accounts by contract value, since their logos will look good online. Eight executives accept, and three of them later send a director in their place. The January session produces polite praise and a list of requests the product team has heard many times before.
Meanwhile, a CSM on the same team spends the year on calls with an operations manager at a mid-sized account. On almost every call, that manager explains how her team exports data each Monday to build a missing report. She has clear views on the reporting screen and the setup steps, and she explains the reason behind each one. Nobody puts her name on the advisory board list, because her account sits in the middle of the revenue ranking.
The two lists rarely overlap, because contract value and the habit of explaining your product's gaps measure different things. Below is a way to recruit board members from your customer calls and run the sessions, plus a simple log that shows members how the board changed the roadmap.
A customer advisory board is a small group of customers who meet with a vendor's leadership to shape product direction. Members usually serve for a year and see plans before the rest of the market does. In return, the vendor hears candid opinions from people who use the product every week.
CS teams often confuse the board with a few nearby programs. User groups are open to any customer and focus on training, while a beta program tests one release. Reference programs ask customers to speak with prospects, as our guide to preparing champions for reference calls explains. Only the advisory board asks customers to help decide what the company builds next, so the member list decides whose problems product hears first.
That choice usually happens in the fall, since most companies hold the first session in January or February. The invitations need to go out by November, so October is when CS leaders pick the voices product will hear.
Most invite lists start with the biggest accounts, and the reasons make sense from inside the company. Each reason, though, favors familiarity over the quality of feedback the board will produce.
An executive sponsor at a large account often sits several levels above the people who use the product daily. That sponsor can discuss budget and strategy, but probably not the export step that eats an hour each week. CS leaders end up with a board that talks in generalities, and product leaves with a thin picture of daily use.
Salesforce's State of the Connected Customer research tracks how strongly customers expect companies to understand their needs. A board session lets a product manager test that understanding directly, by hearing a customer walk through a normal week in their own words.
Every recorded customer call holds evidence of how that customer thinks about your product. Across a year of onboarding sessions, QBRs, escalations, and renewal calls, some customers volunteer opinions again and again. Others answer whatever the CSM asks and then wait politely for the call to end.
Look for the volunteers. They describe the workaround they built, and they ask when a feature will ship because they planned around it. In practice, that habit tells you more about a future board member than company size does. Advisory sessions reward people who can explain their own work in detail, and volunteers do that on every call.
Calls also show who sits on the customer's side of the table. One account may send a director to the QBR and an admin to every support escalation. Another account sends only the executive sponsor to each QBR, which limits whom you can invite.
CS teams that already mine calls for voice of customer intelligence have most of the raw material for a shortlist. The difference is the unit of analysis. Voice of customer work groups feedback by theme, while board recruiting groups it by person. So the question shifts to who gave each piece of feedback and how well they explained it.
Five signals in the call history separate a strong candidate from a customer who is merely happy. A CSM can check each one by reviewing the last two quarters of calls on their book.
Count the moments when a customer raises a product opinion before anyone on your side asks for one. A customer who does this on most calls already behaves like an advisor. Phrases like "what would really help us" often come right before specific feedback, so a search can find them fast.
Strong candidates explain how the product fits into their week, such as which report they pull on Monday. With that detail, a product manager can picture the real use case without a site visit. Customers who only say "it works fine" may be happy, but they will struggle to fill a long session.
A request for a new export format tells product what to build, while the reason tells them what to solve. Look for customers who say why, such as a finance team that needs data shaped for month-end close. Those customers help product find the underlying problem, which often needs a different fix than the one they named.
When an account sends several roles to different calls, you get a real choice of whom to invite. That pattern also shows the product has spread beyond a single champion. A member from such an account can speak for several teams, since they hear from those colleagues every week.
Some customers disagree with your roadmap on calls and still renew the contract every year. That combination is useful, because it means the customer cares enough about the product to argue about it. Leave out customers whose criticism has turned into renewal risk, since a board session cannot fix an open escalation.
If the team records its customer calls, a CS operations lead can build the shortlist in a few weeks. The process has four steps, and the CSMs join at the third one.
The ops lead ranks individual people at this stage, because the best advisor at an account may be an analyst the sponsor has never mentioned. For each name, the ops lead should save two or three quotes, since the invitation and pre-read will need them.
Balance does not mean dropping every large account. A few enterprise sponsors next to hands-on users from mid-market accounts give product both strategic and daily views. The difference is that each sponsor gets a seat because their calls show useful input, and the CS leader can name that input in the invitation.
The two approaches produce very different boards, even when a few accounts appear on both lists. This comparison shows where they split, from the first draft to the renewal conversation a year later.
| Stage | Logo-based board | Evidence-based board |
|---|---|---|
| Starting list | Top accounts by contract value | Customers who give the most specific feedback on calls |
| Who attends | Executive sponsors or their delegates | Daily users plus a few sponsors |
| Invitation | "Join an exclusive group of leaders" | "You told us in March that..." |
| Preparation | Account names and revenue | Quotes and workflow notes from each member's calls |
| Session content | Strategy slides and general praise | Workarounds and the reasons behind requests |
| Role mix | Mostly executives | Admins, team leads, analysts, and executives |
| Follow-up | A thank-you email | A dated log of each input and its outcome |
| Renewal effect | Hard to trace | Members can point to features they shaped |
Product managers care most about the session content and follow-up rows, because those decide whether they leave the room with something to build. For the CS leader, the last row matters more, since it tells each CSM which shipped features to mention on the renewal call.
The invitation is the first place where the call evidence pays off, because the CS leader can quote the customer. Picture an invitation that quotes the customer's March comment about the hour-long weekly export and asks for help fixing it. That note explains exactly why the company picked this person, and a line about joining a group of industry leaders gives no such reason.
Send the invitation from the head of CS or product, and have the CSM give the customer a heads-up first. State the commitment in plain terms, including the number of sessions and the length of the term. Then name what the member gets, such as roadmap previews and time with the product managers who own their requests.
When the invitee sits below the executive sponsor, the CSM should tell the sponsor first. Most sponsors like hearing that someone on their team impressed a vendor, so the courtesy heads off awkward questions. It also lets the sponsor suggest a second person from the same account, who sometimes turns out to be even stronger.
A good session spends most of its time on two or three questions the product team cannot answer alone. Everything else, from the roadmap preview to the lunch, supports those questions.
At the first session, show members what the company did with feedback they gave on earlier calls. If the honest answer is that the work is still in progress, say so and give a date. For later sessions, open with the log from the previous meeting, because proof of listening makes members speak more freely.
Pick decisions the product team is actually weighing, such as which of two workflow changes to build first. A week ahead, send a short pre-read that frames each question with quotes from members' own calls. Members then arrive having thought about the tradeoff, and product hears considered answers instead of first reactions.
The most useful moments often come when one customer explains a workaround and another says they do the same. For that reason, the facilitator should run small-group discussions with one or two vendor staff per table. Product managers should mostly listen during these discussions and save their follow-up questions for the end.
Ask members at the start whether the team can record the session for the product team's notes. A transcript lets product managers quote members accurately later, which matters when they report back on decisions. It also frees everyone on the vendor side from typing notes when they should be listening.
Members stay engaged when they see what happened to their input, and they drift away when sessions feel like surveys. The simplest proof is a running log that ties each piece of board feedback to a product decision.
Each log entry records the member, the date, the quote, the decision, and the current status. The product manager who owns the area updates the entry, and the CS leader shares the log before every session. Include the requests the team declined, with the reason, because a member who hears the reason for a no can plan around it.
Between sessions, the CSM should send a short note whenever a feature a member asked for ships. That note works best when it quotes the member's original words and links to the release. Over time, those notes give the CSM a concrete story for the renewal call about features the customer helped shape.
To judge whether the board works, the CS leader can track two plain measures over the year. The first is how many roadmap items trace back to board input. Just as telling is whether members keep giving detailed feedback on regular calls after they join. If members go quiet on calls after a session, the board has probably started to feel like theater to them.
Rafiki AI is the intelligence layer between your customer conversations and the decisions CS and product make from them. Its conversation intelligence records and transcribes video calls and Aircall or OpenPhone calls in more than 60 languages. That coverage lets the shortlist include customers in every region, even ones whose calls the core team could not follow.
For the first shortlist step, a CS operations lead can ask Gen AI Search a plain-English question across every call. A question like "Which customers described a reporting workaround since April?" points to the calls that answer it. Product can reuse the same search later to hear how members talk about a feature after it ships.
The CS team can also add a custom Smart Call Scoring criterion for whether the customer volunteered product feedback. Rafiki AI then scores every customer call on that criterion alongside the team's existing scorecard. Meanwhile, Smart Call Summary notes product requests in each call recap, and the CS leader can use Gen AI Reports to pull those requests into one report.
Finally, Smart CRM Sync fills custom CRM fields from call content in Salesforce, HubSpot, Zoho, Pipedrive, Freshworks, and Monday.com. A team might add a "latest product feedback" field to each contact record in the CRM. The CSM who reviews the draft list then sees each candidate's most recent opinion without opening a recording.
This week, choose one question to search across the last two quarters of customer calls. A good first question is which customers described a workaround or explained why they needed a feature. Then ask each CSM to add two names from their own book, each with one quote from a recent call.
By the end of October, the CS operations lead should have a ranked draft of about twenty people. During the first week of November, trim that list to your target size and brief each executive sponsor. Send invitations by mid-November for a session in late January, before the product team locks its Q1 plans. That timing means members' input can still change what ships in the spring.
Rafiki AI's conversation intelligence platform helps your CS team find the customers who already give detailed product feedback on calls. Plans start at $19 per seat per month, and you can buy only the seats your CS team needs. You can start a free trial and search last quarter's customer calls this week. If you would like a walkthrough first, book a demo and watch the search run on your own conversations.
Keep the board small enough that every member speaks during each session. In our view, that usually means about eight to twelve people across different accounts and roles. Larger groups tend to split into talkers and listeners, and the listeners often stop attending after the first meeting.
Two to four sessions a year suits most teams, with at least one in person if budget allows. Time the first session before product locks its plans for the next quarter, so member input can still count. Between sessions, short written updates on the feedback log keep members engaged without extra meetings.
Invite the ones whose calls show specific, well-explained feedback, and apply the same test to every seat. A mix of enterprise sponsors and hands-on users lets product hear about budgets and about daily work. If a large account's sponsor rarely joins calls, consider inviting the admin or team lead who does.
Track how many roadmap decisions trace back to board input, using a log that names the member behind each one. Watch whether members keep giving detailed feedback on their regular calls after they join. At renewal, listen for members who mention features they helped shape, since those mentions show the board's work carried into the buying decision.
Start for free — no credit card, no seat minimums, no long contracts. Just better sales intelligence.
