The Buying Committee Keeps Growing. Here's Who's Actually In It.

Ask five sources how many people sit on a B2B buying committee and you'll get five different numbers, each stated with the same confidence. UnboundB2B says 10 to 11, with 52% including someone VP-level or higher. Jolly Marketer says 10 to 13, with deal cycles running 11.5 to 12 months. And a number you'll see everywhere online is 11, usually attributed to Gartner in a sentence with no link attached.

I went looking for the primary document behind that 11. Gartner has published real research on buying-team behavior, and its own site is where the citation supposedly leads. But every path to it dead-ends at a paywall or a blocked page, so I can't verify what the underlying study says versus what got repeated through a hundred SEO blogs until it hardened into fact. That's not a reason to throw the number out; the commonly cited range, 10 to 13 stakeholders, is close enough to plan around. It is a reason to stop quoting a specific headcount like you personally confirmed it.

Here's what the field cites, side by side.

What nobody disputes is the direction. Every practitioner guide agrees the committee has grown. But almost none explain why, or what a seller with a smaller team is supposed to do about it. That's the real story.

Why the committee grew while your sales team didn't

The stat everyone skips is the one on the seller's side of the table. ICONIQ Growth's 2026 "Leaner, Smarter, Flatter" report, based on a survey of more than 150 B2B software GTM leaders, found that teams with high AI adoption run 20 to 30% leaner than low-adoption teams and still produce roughly twice the net-new ARR per GTM FTE: $640,000 versus $370,000.

Read those two facts together and the committee-size number stops being trivia. Buying groups expanded through the same years GTM headcount, at the teams doing this well, shrank. Nobody wrote a memo connecting the two. But the mechanism isn't complicated. Deals still route through the legal reviewer, the procurement officer, the executive sponsor, and the person who quietly controls whether the technical evaluation ever gets scheduled. All of that stayed the same size or grew. What changed is how many humans on the selling side are available to cover it.

Some of this is risk aversion baked in after a few bad budget years: nobody wants to be the single name on a purchase that goes wrong. Some of it is flatter, more matrixed org charts, where a decision that used to sit with one VP now touches three peers who all technically have a say. And remote, distributed teams add their own layer. It's harder to informally read the room when half the committee is in a different city and you've never been in a conference room with them. None of that shows up in a stakeholder-count stat. All of it is why the stat keeps climbing.

The seats, and what each one blocks on

Every deal has a version of these roles, whether or not the org chart uses these titles. What matters for forecasting isn't the label. It's what each seat blocks on, and the one signal that tells you they're engaged instead of just cc'd.

The Champion. Internal advocate, usually no budget authority of their own. Blocks on: they can only sell what you armed them to sell internally; if you never gave them a business case built for someone else's inbox, the deal stalls in a room you were never invited to. The full definition, and the field test for telling a real champion from a friendly coach, lives in champion vs. coach vs. economic buyer.

The Economic Buyer. Controls the budget, wants ROI and payback math, not a feature tour. Blocks on: if nobody with signing authority has heard your case in their own words, the deal is a coin flip no matter how warm the champion feels. Same piece above covers the tell that separates this seat from someone who just approves invoices.

The Technical Buyer. Evaluates architecture, security posture, and integration risk. Blocks on: one unresolved security or data-residency question, and the deal freezes regardless of how much the business side wants it. The signal that they're engaged, not just cc'd: they're asking about your roadmap, not just your current feature set.

The End User. Runs the workflow your product touches every day. Blocks on: quiet veto power. They rarely kill a deal in a meeting; they kill it by not adopting the thing after signature, which shows up as churn six months later instead of a lost opportunity now. Engagement looks like specific complaints about the current tool, not polite enthusiasm about yours.

Procurement. Owns commercial terms. Blocks on: they will reopen pricing and contract language regardless of how excited every other seat is, because that's the job description. Treat their late-stage involvement as normal, not as a red flag that the deal is falling apart.

Legal and Compliance. Owns contract and data risk. Blocks on: they move last, slowest, and with the least patience for a rushed signature. A data processing addendum nobody flagged in month two becomes a six-week delay in month five. Engagement here looks boring on purpose: specific redlines, not a general blessing.

The Executive Sponsor. Rarely in the weeds, decides tie-breaks when the committee is split. Blocks on: without a named sponsor, a truly divided room has no mechanism to reach a decision, and it just sits. The signal they're engaged: they've asked a business-outcome question, not a product question.

The Influencer. No formal authority, disproportionate informal weight. Often a peer the economic buyer trusts, or a former user of a competing tool. Blocks on: an influencer who's quietly against you can slow every other conversation without ever showing up on a call. Watch who the champion mentions checking with before committing to anything.

The Gatekeeper. Controls access, usually an EA or a chief of staff. Blocks on: they can stall you by omission, simply not forwarding the meeting request. And nobody above them will ever know you asked. Quiet, not loud. Treat them as a real stakeholder, not an obstacle to route around.

The Blocker. Actively resists, usually out of loyalty to the incumbent tool or fear of the change itself. Blocks on: they don't need formal authority to kill momentum. But a well-placed "have we really thought this through" in the wrong meeting does the job. The signal to watch: specific, recurring objections aimed at the same feature or risk, not vague hesitation.

The Initiator, or Coach. Flagged the original problem, sometimes coaches you on the internal politics. Blocks on: almost nothing directly, but losing their trust dries up your visibility into everything happening in rooms you're not in. Coach is a distinct job from champion, and conflating them is one of the more common forecasting mistakes; see champion vs. coach vs. economic buyer for the full split.

How to find these people in a specific account

You don't find every name by guessing titles off an org chart template. You find them by watching what the account does: who's posting about the problem you solve, who shows up in a hiring listing for a role your product touches, whose name is attached to a case study for the incumbent tool you're about to replace.

That's signal-based account research, and it's the difference between a persona list and a mapped account. Run an enrichment waterfall against the target, we use Clay with FullEnrich as one of the fallback sources, and you get named people attached to real titles instead of a template that assumes a VP of RevOps exists at a forty-person company that never hired one.

No credible source publishes a number for how much faster this makes multithreading. So I'm not going to invent one just to have a stat in this section. No made-up multiplier, no matter how good it would look in a headline. What's defensible is the shape of the problem: fewer sellers are covering more of the committee than five years ago, and a spreadsheet of assumed titles doesn't survive contact with an org chart that's flatter, more matrixed, and half remote.

This is close to the core of what GTM engineering is: the system that turns account signal into a mapped, named committee before your first call, instead of hoping your rep guesses right. It's also the argument against AI SDRs that automate personalization without fixing the mapping problem underneath it, and the case for running ABM against the full committee instead of one warm contact. The exact mechanism connecting a signal to a champion to a pipeline stage is its own piece; see how GTM engineering feeds the enterprise motion for that bridge.

What happens when you only cover three of the seats

Say you've got a warm champion and a technical buyer who likes your architecture. That's two seats, maybe three if the end user is friendly. The deal doesn't die in a dramatic no. It goes quiet in procurement. Or legal flags a data-residency clause nobody prepped for. Or the executive sponsor who was supposed to break the tie gets reassigned in a reorg you never saw coming, because you never had a name for that seat to begin with.

Single-threaded deals rarely fail at the stage you're watching. They fail at the seat you didn't map, two or three stages later, when it's too expensive to go back and build the relationship you skipped. There's no clean industry number for how often that specific pattern plays out. Every version of that stat traces back to a marketing blog with no methodology attached. So I'm not handing you one either.

What's worth building instead is a short, repeatable way to check your own coverage before you're in that spot: role, what they block on, current engagement level, next signal to watch. We built the full version of that, including a kill-risk score, in multithreading enterprise deals. Worth reading before your next forecast call, not after the deal goes quiet.

Here's the part that should bother you more than the exact headcount. The committee grew through the same stretch that GTM teams, at the ones doing this well, got leaner. Not a coincidence. And not a problem you fix by hiring three more reps to split the multithreading work. It's the argument for building a system instead of stacking more people on the same guesswork.

We don't know if the real number is 10, 11, or 13. Not the point. We know it isn't shrinking. And we know most sales teams are already lean enough that they can't afford to guess at half the room.