Most ABM programmes are not orchestrated. They are executed in parallel the hope that the account cannot tell. That hope is the strategy.
The vocabulary of ABM campaign orchestration has already been agreed. Every platform now offers ‘account tiering’, ‘intent signals’, and ‘omnichannel activation’, so the language of orchestration has become universal while the practice of it has not. What most teams call orchestration is a set of campaigns running at the same time, aimed at the same logo, coordinated only by the fact that they share a target list. The account experiences four or five unconnected touches a week and reads them as noise. The team experiences a dashboard that reports activity and calls it progress.
The account is the unit nobody owns
The fracture between marketing and sales sits at the account level. The account is the unit of orchestration, and it is also the unit nobody owns. Ownership sits with the campaign, the channel, the SDR sequence, the regional team, the opportunity. Each is measured on its own output. None is measured on whether the account moved through the buyer journey in a meaningful way that will impact pipeline.
Three motions, not one
Three operating motions exist, and they are not interchangeable. One-to-one covers a small number of top-tier accounts, typically between one and ten, with dedicated practitioners, bespoke plans, and high-touch offline engagement. One-to-few covers clusters of roughly ten to one hundred accounts, grouped by vertical, geography, or shared business challenge. One-to-many covers larger segments of a hundred accounts and above, using intent and ideal customer profile data across scaled targeting and automated nurture. Each motion carries its own cost structure and its own definition of a good week.
The conundrum is not that teams cannot personalise. It is that personalisation, omnichannel delivery, and data-driven timing are three separate disciplines, and running all three at once against a single account produces more coordination overhead than most marketing functions were built to absorb. The tools promise alignment. The operating model decides whether it happens.
The mechanism is not alignment. It is suppression.
An account receives a signal. A visit to a pricing page, a download, an intent surge against a keyword the platform has weighted. That signal enters the unified scoring engine, where first-party behavioural data, second-party publisher signals, and third-party dark-funnel intent are combined into a single weighted score. The score crosses a threshold. The account progresses to consideration or decision stage. Marketing air cover activates. Sales activation fires. The account is now in motion.
The mechanism breaks here, and it breaks predictably. The scoring engine does not know what the account has already seen. It knows what the account has done. A buying committee of six to ten contacts across finance, procurement, operations, and
IT will generate behavioural signals from several of them at once. Each signal is treated as an account-level event. Each event triggers a touch. The account receives deal acceleration ads, stage-based email nurture, and automated multi-touch cadences simultaneously, because each system is doing exactly what it was configured to do.
This is not personalisation. It is accumulation.
The account as control point
The account-centric operating model treats the account as the central control point for orchestration, suppression, and reporting. The word doing the work in that sentence is suppression. Orchestration without suppression is noise generation at scale. The mechanism that makes omni-channel ABM function is the same mechanism that makes it fail when omitted: an account must be removed from a motion before it is added to another. The buying committee member who has already received three touches this week from three different systems is not being nurtured. They are being managed into fatigue by a stack that has no shared memory of what it has already said.
The unified scoring engine reads intent. It does not read saturation. The account progresses on the strength of its signals, not on the tolerance of the people inside it. High-intent handoff triggers push the account into the CRM and alert the assigned representative with the intent keywords and visited URLs attached. What the alert does not carry is the count of prior touches across every channel. The representative opens a cadence against a committee that marketing has already contacted four times this month, because the two systems were integrated on data and not on restraint.
The fracture is structural. Three disciplines, one account, no shared ceiling on contact frequency. The tools promised coordination. What they delivered was concurrent execution.
The source material does not specify a contact frequency cap, a suppression rule across channels, or a saturation model. What it does name is the operating logic that would prevent the problem: the account as the central control point for orchestration, suppression, and reporting, with an individual contact’s action treated as an account signal rather than an isolated opportunity.
That distinction carries the consequence. If the account is the unit of control, a contact who has been touched four times is not a lead to pursue. They are a signal that the account is saturated, and the orchestration layer’s job is to suppress the next touch rather than schedule it. The suppression rule is the part most stacks do not build, because suppression looks like lost activity. It reads as a gap in the funnel report. It is the opposite. It is the only mechanism that converts concurrent execution into coordinated execution.
What changes when the account runs the model
Adopt the account-as-control-point model and three things change. Reporting changes: pipeline is measured at the account, not the contact, so a committee that has gone quiet shows as a stalled account rather than four cold leads. The handoff changes: the alert that pushes an account into the CRM carries the full touch history, not just the intent keywords, so the representative opens a cadence on informed ground. The content changes: modular blocks are swapped by account tier and persona role, which means the same account can receive a CFO narrative and a procurement narrative without the two colliding.
Where to start depends on where you are
Broadly, three states of affairs exist when embarking on this change and each has one action that fits it.
If sales and marketing are operating in silos, with separate systems and separate teams – e.g. Demand Generation, Inside Sales, Account Executives – the action is to build a single shared account list before building anything else. Not a scoring model. Not an orchestration layer. One list, agreed by all functions, ranked by the same criteria, owned by the same person. The account cannot be a control point until teams agree on which accounts to target and why. Everything downstream depends on that agreement, and no technology purchase can substitute for it.
If sales and marketing are aligned and systems are integrated but performance metrics remain siloed – e.g. marketing measures meetings or SALs, Sales measures pipeline value and stage – the action is to replace the channel scorecard with an account scorecard. The integration already carries the data. What it does not carry is a shared definition of whether the account moved in a meaningful way. Measure weighted pipeline at the account, report it once, and retire the separate marketing and sales dashboards that describe the same account in two different languages. The suppression logic described above becomes possible only when both functions read from the same number.
If the operation is tracked and integrated end to end, with shared metrics and a shared account view, the action is to monitor saturation. The stack already has the touch history. Account scoring rules need to evolve to accommodate anonymous account engagement, known contact engagement and distinguish relevant buying group behaviour from merely a high total account score. This determines how and when to signal that a meaningful threshold has been reached.
This is the job of the next generation of AI-enabled orchestration platforms