Scoring a Buying Committee by Power, Stance, and Coverage

Turning one target account into a working map means naming every person in the decision, then scoring each on three things you can actually observe: their role, their power, and their lean toward you. This guide is the rubric for that scoring, written for founders selling their own product, account executives, SDR leads, and partnerships teams. When you finish, you will have a named roster where every person carries a decision-role, a power score, a stance score tied to a specific piece of public evidence, a confidence level, and a verdict on whether the map is complete enough to work the deal.

Most account mapping content stops at defining the six role archetypes or drops a generic power grid borrowed from project management. Neither tells you how to assign a score from a signal you can see. This guide closes that gap and adds the part that matters most: the coverage test that separates a workable roster from single-threaded risk.

What a scored committee roster actually is

A scored committee roster is a named list of everyone in a decision, where each person carries four things: a decision-role, a power score, a stance score, and a confidence level. It ends with one coverage verdict for the whole account.

The roster is not a contact list and it is not an org chart. It is a judgement, made from evidence, about who moves this specific deal and which way they push. Three properties make it useful mid-deal:

• Every score traces to evidence. A stance of "champion" without a cited signal is a guess. Tie each rating to a profile view, a comment, or something the champion told you.
• It carries confidence, not just a value. A score built on one ambiguous signal is worth less than one built on three. The rubric forces you to say which.
• It resolves to a verdict. The point of scoring is a decision: work the deal as mapped, or fix the coverage gap first.

The stakes are quantitative, not stylistic. A single-threaded opportunity has a 5% chance of winning, but at five engaged people that jumps to 30%, a 6X improvement (UserGems). Despite that, 70% of B2B opportunities still have only one point of contact in the CRM. The roster exists to move you off that 70%.

How big should the roster be

Size the roster to the deal band, not to a fixed number. The research converges on a wide range, so use it as a sanity check on whether your map is plausibly complete rather than as a target.

Buying groups have grown. Harvard Business Review put the average B2B purchase at 6.8 decision-makers in 2017; Forrester's 2024 study now reports 13 stakeholders, with nearly 89% of decisions crossing multiple departments. The table below gives you the bands to check against.

Two readings matter. First, if you are working a seven-figure deal and your roster has four names, you are missing most of the committee. Second, an Influ2 survey of enterprise and mid-market buyers found half had buying groups of only 2 to 4 people, so a large deal band does not license you to invent names. Map who is real, then compare against the band to see where you are probably blind.

The four dimensions you score

Score every person on four dimensions in order: decision-role first, then power, then stance, then confidence. Role tells you what the person does in the decision; power tells you how much their yes or no counts; stance tells you which way they lean; confidence tells you how much to trust the other three.

Decision-role: what they do in the decision

The durable model is Miller Heiman's four buying influences, extended with the roles that kill deals. Tag each person with exactly one primary role.

The load-bearing distinctions here are the ones people get wrong. The economic buyer is the final decision authority who controls budget approval and can override everyone, saying yes when others say no or the reverse. The technical buyer, which usually means procurement, security, or IT, holds veto but not approval: procurement is almost never the real decision maker, it is a veto gate that can stop a deal but rarely champions one. Do not confuse authority to say no with authority to say yes.

Title does not equal role. Job titles often do not reveal the actual buying role, so you discover the role through conversation. An enthusiastic VP is not the decision until you have confirmed sign-off sits with them.

Power: ability to say yes or no

Score power 1 to 5. The published tool, Mendelow's power-interest grid, prescribes quadrants rather than a numeric scale, and a fixed numeric scale is not publicly established, so treat this scale as a working construct you calibrate to your own accounts.

• 5 - Final approval. The one person who can say yes when everyone else says no.
• 4 - Strong veto or heavy influence over the approver.
• 3 - Named influencer whose opinion the approver seeks.
• 2 - User or evaluator with input but no veto.
• 1 - Peripheral; consulted rarely or not at all.

The done condition for this step is a single name at power 5. If you cannot name that person, you have not found final authority yet.

Stance: which way they lean

Score stance minus-2 to plus-2, and require an observable signal for anything off zero.

• +2 Champion - actively selling you internally, will introduce you to the economic buyer.
• +1 Coach - helpful and responsive, shares information, but does not fight for you.
• 0 Neutral - no observable lean either way.
• -1 Skeptic - engaged but unconvinced, or lurking without engaging.
• -2 Blocker - actively opposing, or a known veto owner leaning against.

Reading public signals without being fooled

Read stance from stacked signals filtered to the right person, never from a single reaction in isolation. One signal is interesting, two suggest intent, and three from the same account within 30 days is a buying cycle in progress.

The signals invert by role, which is what trips up most rubrics. The champion engages early and often, liking several posts, viewing your profile more than once, sometimes commenting, because they are building a case internally. The economic buyer shows up late, briefly, from a senior title: one profile view, no likes, often right before a decision. The skeptic lurks without engaging at all. All three are on-track patterns for their roles. Reading the economic buyer's silence as disinterest is a scoring error.

Here is what each signal proves and how it lies.

Two filters are mandatory before you score any engagement. Filter by current employer and ICP persona, because job-seeker engagement from an "Open to Work" contact reads identically to buyer engagement. And exclude active customers via CRM sync, because a comment from a current customer is advocacy, not buying intent.

Finding the right named people to read these signals against is the slow part. Refolk lets you ask in plain English and get the roster back across LinkedIn, GitHub, and the open web, so the hour you would spend hunting titles goes to scoring instead.

Confidence: how much to trust the score

Confidence is the discipline that keeps the roster honest. Label every stance high, medium, or low.

• High - three or more stacked signals, or direct confirmation from a mapped contact.
• Medium - two signals, or one strong signal plus a plausible role fit.
• Low - a single ambiguous signal, such as one like or one profile view.

Downgrade any stance resting on a single ambiguous signal to low. A low-confidence "champion" is not a champion; it is a lead to verify.

The scoring procedure, end to end

Run these eight steps in order. The first pass takes a focused afternoon; steps 7 and 8 are ongoing. Sources disagree on timing of the first step: Miller Heiman maps all decision-makers before the first sales call rather than letting the map develop organically, and a full Blue Sheet takes three to five hours, with new reps needing 90 to 120 days to reach proficiency. Start light and refresh often rather than waiting for a perfect map.

Name | Role | Power (1-5) | Stance (-2..+2) | Confidence | Evidence
J. Rivera | Economic buyer | 5 | 0 | Medium | One profile view, no likes, 2 days before renewal date
M. Chen | Champion | 3 | +2 | High | Introduced me to EB; 4 post likes; commented twice in 30d
P. Osei | Technical buyer (security) | 4 | -1 | Low | Not yet engaged; named by champion as a sign-off gate

The coverage test: workable or single-threaded risk

The coverage test is a pass or fail verdict on the whole roster. It passes when you have three or more active threads spread across budget, technical, and executive functions. It fails, and you flag single-threaded risk, when coverage rests on one person or leaves a function dark.

Multi-threading means building relationships with 3 to 7 decision-makers and influencers so the deal survives personnel changes and committee decisions. For SMB deals two or three contacts may suffice; mid-market and enterprise deals typically require four or more active relationships across budget, technical, and executive functions. Gong's analysis of 1.8 million opportunities found closed deals have twice as many buyer contacts as lost ones, and larger deals above $50K that use multi-threading close 22% faster than single-threaded ones.

The matrix below is how you place each person once role and scores are set, using power against stance.

The verdict is not a feeling. Below three active threads you are, by the win-rate data, in a losing position. Do not declare the map complete because one champion feels strong; a single-threaded deal depends on one person to convince the other five to nine.

Coverage of consensus-builders beats out-pitching a rival, because indecision kills more pipeline than any competitor.

The reason coverage matters more than persuasion is the shape of deal death. Over 40% of B2B deals stall not because a competitor won but because stakeholders fail to align internally. The "no decision" outcome kills more pipeline than any competitor, and 74% of buyer teams show unhealthy conflict during the decision. A roster that covers the consensus-builders is your hedge against that.

Where the scoring goes wrong

The most valuable part of any scoring rubric is knowing how it lies to you. Each failure below is a false positive that feels like progress, with the check that catches it.

• Mistaking a coach for a champion. A helpful, chatty contact you assume will fight for you. Check: a true champion introduces you to the economic buyer. If they are unwilling or unable, they are not a champion. Score them +1, not +2.
• Confusing influence with authority. An enthusiastic VP treated as the decision. Check: ask "Who needs to sign off on this?" and listen for "I'll need to check with," which means you have not reached power 5.
• Reading a single like as stance. One reaction scored as champion. Check: a single like or follow is attention without urgency. Require stacked signals and label the score low confidence until you have them.
• Job-seeker mistaken for buyer. Engagement from an "Open to Work" contact scored as intent. Check: filter by current employer and ICP persona before scoring.
• Current customer mistaken for prospect intent. Advocacy misread as net-new buying. Check: exclude active customers via CRM sync before scoring.
• Declaring the map complete while single-threaded. One strong champion that feels like coverage. Check: run the coverage test and count functions, not enthusiasm.
• Ignoring the unmapped blocker. The roster looks full but a 4pm phone call kills it. Check: explicitly ask the champion who else the economic buyer consults.
• Treating the map as static. Scores decay as the committee shifts. Check: re-qualify each stage; reps who treat their initial assessment as fixed forecast against outdated intelligence.

The late-stage killers and the supply math

Procurement, legal, and security are the recurring late-stage killers, and they enter after the business case is already approved. They are measured on savings and risk reduction, and they re-enter forcefully whenever data flows appear to have drifted from the original decision narrative. Score them as high-power blockers the moment they are named, even if they have not engaged.

Security deserves special attention because it is scarce. In Refolk's index of professional profiles, the US economic-buyer-titled pool and the security-veto pool are wildly uneven.

That is roughly 23 economic-buyer-titled people for every one CISO-class person. The security veto is a thin, hard-to-reach layer, which is exactly why it enters late and why finding that one person early is worth the effort. Ask your champion for the name before the security review appears on the calendar.

Geography changes the math too. The US economic-buyer pool is 11.3x the UK's in Refolk's index, so a playbook that assumes easy alternate paths to budget authority breaks in thinner markets.

In a thin market there are fewer parallel buyers to switch to if your economic buyer leaves, so protecting the single approver you have becomes more important, not less. Finding the CISO, the alternate approver, or the former employee now at one of your customers who can act as a reference is where a plain-English search across the public graph earns its place; Refolk returns those named people so the hunt does not stall your coverage work.

Keeping the map current

Treat the roster as a live document, not a one-time exercise. Re-run every component at each stage gate, because a committee shifts as the deal advances and champion engagement triggers buying-group expansion within 3 to 56 days depending on deal size. A map you scored three weeks ago is intelligence about a committee that no longer exists.

Before you call any pass done, run this checklist.

The rubric's job is to convert what you can see into a decision you can defend to your manager. When the coverage verdict reads single-threaded risk, that is not a failure of the map; it is the map doing its job. Fix the gap, then advance.