Gartner puts the typical buying group for a complex B2B purchase at six to ten people. Those people spend roughly 17% of their total buying time in meetings with suppliers, and that sliver is divided between every vendor in the running. The rest of the decision forms in places you cannot see: internal documents, peer conversations, independent research. By the time a first meeting happens, 81% of buyers already have a preferred vendor and 85% have set their requirements, according to 6sense's 2024 B2B Buyer Experience Report.
Which raises a practical question. If the decision forms before you arrive, what exactly should you know about an account, and how do you keep knowing it as the account changes?
The Knowledge Graph™ is our answer. It is the second engine in Authority OS™, and its job is to hold a living dossier on every target account: who decides, where you fit, what is moving inside the account, and the play that is ready to run when something does. This piece explains how that dossier is built, layer by layer. It is written for readers who want the methodology, not the brochure.
The org chart is a seating plan, not a decision map
Chester Barnard made the point in 1938, in The Functions of the Executive: every formal organisation contains an informal one, and the informal one does much of the actual work. The org chart records reporting lines. It says nothing about who gets consulted before a budget owner commits, whose scepticism can stall a project for a quarter, or which mid-level engineer's verdict on a supplier is treated as final.
So the first layer of the dossier is not the org chart. It is what we call the decision architecture: a working map of the buying group, built from titles, public signals, and what each mapped person has said and done in the open. The system assigns a role hypothesis to each person it maps, economic buyer, technical evaluator, champion, and revises the map as engagement data comes in.
Note the word hypothesis. Nobody can observe from the outside who holds a veto inside another company. What you can do is infer the probable shape of the decision before the first conversation, and correct the map quickly when the account proves you wrong. That is still a considerable head start on walking in with a seating plan and a hunch.
Intelligence has to be relative to you
Around the architecture sits a second layer: the strategic picture. This is where most account research goes wrong, because most account research is symmetric. A general profile of the account, its market, its products. Any competitor could compile the same file, and several probably have.
Intelligence is different from research in one respect: it is relative to the party using it. The Knowledge Graph builds four things per account, each anchored to your offer rather than to the account in the abstract:
A use-case match: where your specific capability meets their specific need, stated concretely enough that a salesperson could open with it.
An approach angle: the single frame most likely to land, given what the account is currently funding and worrying about.
The initiatives they are actively funding, because budget follows initiatives, not sympathy.
A live SWOT, read from your position. Their strength you can attach to, their weakness your offer covers, the threat that makes them receptive this quarter.
A SWOT of the account in general is homework. A SWOT of the account relative to your offer is a weapon. The difference sounds small and is not.
Why the file cannot be static
Friedrich Hayek argued in his 1945 essay The Use of Knowledge in Society that the knowledge that matters most in an economy is "knowledge of the particular circumstances of time and place". It is local, dispersed, and perishable. Prices work, in his account, because they transmit these local changes outward as signals anyone can read.
Accounts transmit signals too. A role change. A board appointment. A funding round. A regulatory filing. A product launch. Each one is a piece of local knowledge surfacing in public, and each one changes the file. The head of quality who held the veto has left. The initiative that had no budget now has one. The acquisition rumour that made everyone cautious has been confirmed or killed.
This is why the Knowledge Graph runs as live deployment rather than a report. The system monitors every mapped person and every account continuously. When an event fires, it becomes a Trigger: a moment of relevance that tells the team exactly when, and with what message, to engage.
An obvious objection: the signals are public, so your competitors can see them too. True, and irrelevant. The asymmetry was never in the signal; it is in what the signal means for your offer specifically, and in having the response built before it fires. Anyone can see a job change on LinkedIn. Very few know, within a day, that this particular change reopens a supplier decision their capability happens to fit, and fewer still have the message for it already drafted. Gartner's 17% figure explains the stakes. You will not learn about these shifts in meetings, because you are barely in any. The account speaks through its signals or it does not speak to you at all.
A dossier that was accurate in March and untouched since is not intelligence. It is archaeology.
The Precision Strike: the play that waits
Knowing when to move is worth little if you then spend two weeks deciding how. Chess grandmasters solved this problem long ago. The novelty that decides a game is rarely invented at the board; it was prepared weeks earlier, against this opponent, for this opening, and played in seconds when the position finally allowed it.
The Knowledge Graph does the equivalent for each account. It builds a Precision Strike: a concrete, account-specific play, constructed in advance and held ready. Each one contains the strategic angle for that account, recommended openers tailored to each named decision-maker, and a step-by-step contact sequence with timing and success indicators.
The openers are not templates with a first name merged in. The system drafts the actual messages, phone openers, LinkedIn messages and emails, by combining three inputs: the strategic intelligence on the account, the Precision Strike angle, and whatever news signal just fired, then matching the result to the recipient's role. A director of business development whose company has just divested an asset gets an email that opens on what that divestment signals about their next move. The draft is generated; the judgement about sending it stays human.
The Triggers tell you when to move. The Precision Strike is what you move with. Together they compress the gap between "something changed" and "we acted on it" from weeks to a day.
A worked example
A real, anonymised account file from live deployment. The client is a Danish surface-treatment specialist. The target is a German high-precision sensor manufacturer, around 450 employees.
The use-case match was specific: the client's PVD coating and electroless nickel plating suit sensor housings and connectors, where fine, uniform layer deposition decides whether the component performs under high-precision requirements. Not "we do surface treatment", but this capability, on these components, for this reason.
The approach angle came from the account's own identity. The target builds instruments that measure surface deviations at micrometre resolution. The argument the system constructed: a company that measures to that standard should demand the same documented precision from its plating supplier. Their quality obsession, turned into the opening frame.
The file tagged three people: the CCO as probable economic buyer, a quality engineer responsible for integrated management systems as technical evaluator, and the managing director. Three roles, three different openers. For the CCO, the message anchored on reclamation risk and making supplier risk measurable through documentation. For the quality engineer, it anchored on qualification batches with verifiable data: layer thickness measurements, contamination testing, SPC records.
The Trigger logic was the sharpest part. The account was in multi-product launch mode, and every new product line opens a component qualification window measured in weeks. That cadence, visible in its news signals, defines exactly when a plating supplier can enter and when the door is shut. The entry point held ready: a zero-commitment qualification batch, offered days after a launch signal fires, not months.
None of this guarantees the deal. What it guarantees is that when the window opens, the account executive is standing in it with a relevant argument, instead of hearing about the launch at a trade fair the following spring.
What the layers add up to
Strip the method to its logic and it is four moves. Map how the decision actually gets made, not how the org chart claims it does. Build intelligence that is relative to your offer, not symmetric research anyone could hold. Keep the file alive through signals, because account knowledge is perishable. And prepare the play before the moment arrives, so the moment is spent acting rather than drafting.
None of these moves is exotic on its own. Good account executives have always kept mental files like this on their best prospects, updated over coffee and trade fairs. The problem was never the idea; it was that a human can hold perhaps five such files with any discipline, and the sixth account is the one that buys from someone else. The Knowledge Graph holds the file for every account on the list, with the same rigour on account forty as on account one.
The result is the line we keep returning to: you never reach out cold. Not because cold outreach became polite, but because the system surfaces exactly when something has changed, and the message you lead with is the one your authority has already established as relevant.
The sales team inherits a qualified basis for the entire sales process: who to contact, in what order, with which argument, and why this week rather than next quarter. Lead qualification usually happens after the first conversation. Here, most of it has happened before anyone picks up the phone.
The org chart tells you who sits where. The Knowledge Graph tells you what to do about it, and when.
Related reading: From Spray-and-pray to Forensic Precision: Why Account Intelligence Wins on how the target list gets built in the first place, and The 3 Signals That Reveal a Buying Window on reading signals before competitors know they exist.
Read the next one before it's archived.
One piece a week, on Authority Engineering and the decisions that shape B2B markets. Written by a human. Worth your inbox.
People at both ends: how The POV Engine works
The POV Engine™ puts people at both ends and technology in between — turning leadership perspective into a position no competitor can copy.
What happens after you sign
See exactly how an Authority OS™ engagement runs at Nevergone — from the Authority Audit through 90 days of activation and the ongoing cadence that follows.