Defining the ICP, the Buying Group, and the Anti-Target in Demand Generation
Written by Robin Caller, Leadscale CEO on March 6, 2026
What Defines the Targeting Layer of a Demand Generation System
Only 10% of a vendor’s total addressable market is actively in-market at any given time (6sense 2025 Buyer Experience Report). Spend against the other 90% without a qualification mechanism and you are not generating demand. You are generating noise. That is why the targeting layer requires three constructs — an ideal customer profile, a buying group map, and a formal anti-target — defined together before any campaign launches.
An ideal customer profile is the account-level definition of which organisations should receive investment. It answers one question: where do we direct resources? But an ICP on its own is incomplete.
It tells you which buildings to walk into. It does not tell you which people to speak to once you are inside, and it certainly does not tell you which buildings to walk past without stopping.
The ICP defines where to invest. The buying group defines who to engage within those accounts. The anti-target defines where not to spend.
These are not three separate exercises run by three different teams at three different times. They are three views of the same targeting decision, and they must be defined together — before any content is commissioned, before any media budget gets committed.
Most organisations get the first one roughly right. They treat the second as a sales problem. And they ignore the third entirely. That sequencing failure, more than any technology gap or attribution challenge, explains a significant proportion of wasted spend in B2B demand generation today.
What Defines an Ideal Customer Profile in Demand Generation?
An ideal customer profile defines the account-level characteristics of organisations most likely to buy, succeed with the product, and generate long-term commercial value. Account-level. Not person-level. That distinction matters more than most marketing teams realise, and getting it wrong cascades into every downstream decision — content strategy, campaign targeting, lead scoring, sales prioritisation, the lot.
The ICP answers a single question: which organisations should receive investment? Everything that follows inherits the quality of that answer. A weak ICP does not simply reduce conversion rates. It actively damages pipeline economics.
Across 48 client engagements, Leadscale consistently sees conversion effort costing five to ten times the initial acquisition spend, once SDR time, nurturing sequences, data enrichment, and sales cycles are factored in. Precision at the account level is not a refinement. It is an economic imperative.
What an ICP Contains
A functional ICP operates across four dimensions.
Firmographic fit covers the observable characteristics: industry vertical, company size, revenue band, geography, growth trajectory. These are the filters most teams start with. They are also where most teams stop. Necessary, but nowhere near sufficient.
Technographic fit examines the technology environment. What platforms does the account run? What stack decisions have they already committed to? What integration constraints exist?
An account running a competing platform is not an ICP match; it is a conversion project with fundamentally different economics. Those economics need to be acknowledged rather than wished away.
Behavioural fit looks at engagement patterns and buying signals: content consumption, event attendance, website visits, third-party intent data. This is where intent signals earn their keep — but only when layered on top of firmographic and technographic qualification. Intent without fit is a distraction.
Economic fit assesses whether the account can generate value that justifies the cost of acquisition and servicing. Deal size potential, contract duration, expansion likelihood, servicing complexity. These variables separate an account worth pursuing from one that will consume resources without producing margin.
Cross-Product ICP Scoring: Where Most Exercises Go Wrong
Here is where the methodology breaks down in practice. ICPs are built by a single product team, for a single product line, in isolation from every other team running the same exercise.
Consider a major UK telecommunications provider with consumer pay-as-you-go, consumer pay-monthly, business mobile, broadband, and IoT divisions. Each product team builds its own ICP. Each defines its own ideal customer. None of them examine the overlap.
The smarter approach: get every marketing team to produce ICPs, then combine them and look for where there are matches. An account that qualifies as an ICP for three product lines scores higher than one qualifying for one — because the total cross-sell and upsell potential is three times greater, the lifetime value is higher, and the acquisition cost is amortised across multiple revenue streams.
A single individual might be relevant to a consumer mobile contract, a business mobile deployment for fifty handsets, a broadband upgrade, and even a specialist IoT product. Siloed teams never see the full picture. The details about that customer are identical across every division, but nobody connects them.
That is the difference between a £30-per-month consumer customer and a £15,000-per-year enterprise relationship from the same individual. The ICP that captures that distinction produces fundamentally different pipeline economics.
The same logic works in B2C. A household with three cars on the drive and two teenagers approaching driving age has a fundamentally different lifetime value from the empty-nest household next door, even if both sit in the same postcode and income band. Single-product ICPs miss this. Cross-product ICPs catch it.
Most ICP exercises fail not because the methodology is wrong, but because organisational structure prevents the methodology from working. Product silos produce product-level ICPs. Company-level targeting requires company-level ICP exercises, and those require someone with the authority to bring all the product teams into the same room.
How Does an ICP Differ from a Buyer Persona?
This question generates significant search volume, and for good reason. The two are routinely conflated, and the confusion is not academic. It produces real operational damage.
The ICP defines which organisations to target. It is an account-level construct. The buyer persona defines which people within those organisations influence the decision and what they care about. It is a person-level construct.
The ICP comes first. Always. Personas are developed within the ICP boundary, not before it.
An organisation that builds personas before defining its ICP is answering “who do we talk to?” before answering “which companies should we be talking to at all?” That sequence produces beautifully detailed persona documents aimed at accounts that will never close.
| Dimension | Ideal Customer Profile | Buyer Persona |
|---|---|---|
| Unit of analysis | Organisation / account | Individual person |
| Scope | Which companies to pursue | Which people within those companies to engage |
| Data sources | Firmographic, technographic, intent, economic | Role, seniority, goals, pain points, content preferences |
| Timing | Defined first; gates all downstream activity | Defined second; within ICP-qualified accounts only |
| Role in demand generation | Determines where investment is directed | Determines how engagement is tailored |
| Failure mode | Too broad = wasted spend across wrong accounts | Too generic = messaging that resonates with no one |
The practical test: if your ICP and your buyer persona look interchangeable, you have not built an ICP. You have built a persona and called it something else. Go back and start with the account.
How Do You Map the Buying Group Within ICP-Qualified Accounts?
Once the ICP identifies which accounts to pursue, the buying group defines which people within those accounts collectively influence the purchase decision. This is not a contact list. It is a decision-making unit — larger, more experienced, and more autonomous than most demand generation programmes account for.
The 6sense 2025 Buyer Experience Report measures the average B2B buying group at 10.1 members. That is not ten people waiting to be educated. The median 40-year-old buyer has participated in approximately 8.6 purchase journeys within their category. They arrive with frameworks, preferences, and prior vendor experience already locked in.
These are not blank slates. They are experienced evaluators running parallel assessment tracks, and demand generation needs to reach the group as a unit rather than picking off individuals and hoping the account follows.
What a Buying Group Looks Like in Practice
The composition varies by deal complexity, but the roles repeat. A champion drives the evaluation internally. An economic buyer controls budget. Technical evaluators assess integration and architecture.
End users will live with the decision daily. And frequently there is at least one blocker — someone whose objections must be addressed even though they may never appear on a form fill or attend a webinar.
Forrester’s 2026 State of Business Buying analysis extends the picture further, measuring buying groups at 13 internal and 9 external stakeholders — a wider lens that captures advisors, analysts, and peer influencers outside the organisation. The methodology differs from 6sense, but the direction is consistent: buying groups are growing, not shrinking.
Consider a pattern Leadscale sees repeatedly in defence and aerospace procurement: a Tier 1 account evaluating a critical component supplier on a multi-year contract worth £5–10 million. The buying group spans seven stakeholders across four departments.
Engineers need performance specifications and testing data. Finance cares about efficiency and total cost of ownership. Sustainability teams need environmental credentials and carbon reporting. Procurement wants compliance documentation, export licensing, and supply chain resilience evidence.
Each stakeholder needs different material, calibrated to their specific evaluation criteria. A demand generation programme that sends the same white paper to all of them is not engaging a buying group. It is broadcasting.
How Should Demand Generation Engage the Buying Group?
The implication is structural. Lead-based programmes measure individual contacts. Buying groups operate as collective decision-making units. Measuring one while trying to influence the other produces a fundamental misalignment between what the programme measures and what actually drives the deal.
This is not a CRM configuration problem. It is a programme design problem. The targeting layer must define not just which accounts to pursue, but which roles within those accounts constitute the buying group and what each role evaluates.
→ What Changed in B2B Buying examines how this structural shift in buyer behaviour broke the lead-based model. And → AI Search and AI Agents in B2B Buying documents how buying groups are increasingly conducting their evaluation inside AI environments where no vendor has direct visibility at all.
Why Is the Anti-Target the Most Economically Significant Targeting Decision?
The anti-target is the least discussed and most consequential component of the targeting layer. It is a deliberately defined exclusion — an account type, segment, or use case that the organisation chooses not to pursue. Not because those accounts would never buy, but because the economics of pursuing them do not justify the investment.
Most organisations have informal exclusions. Sales teams know which industries are a poor fit. Marketing avoids certain segments by instinct or through acquired scar tissue. But informal exclusions leak.
They are inconsistent across teams, invisible to automated systems, and impossible to measure. A formal anti-target makes the exclusion explicit, measurable, and enforceable across every channel.
Why Anti-Targeting Has Become Urgent
Two forces have pushed anti-targeting from a theoretical refinement to an operational requirement.
The first is economic. For every pound spent acquiring a lead, the downstream conversion cost runs five to ten times higher — nurturing, sales engagement, servicing, follow-up, all compounding against the original acquisition (Leadscale client analysis, 48 engagements). When a substantial proportion of acquired data is structurally unsuitable — wrong industry, wrong size, incompatible procurement process — the waste is not in the acquisition cost. It is in the multiplier applied against data that was never going to convert.
A UK residential windows and home improvement company discovered this during a 12-month lead audit. Roughly 40% of their acquired leads were structurally unsuitable, concentrated in segments where installation complexity and customer service costs outweighed the revenue per job. They could extract some sales from that segment, but servicing costs ran at 1.8 times the margin. Discarding the segment entirely and redirecting investment toward accounts with genuine structural fit produced a measurable reduction in wasted servicing spend and a corresponding improvement in cost-per-acquisition for the remaining portfolio (Leadscale engagement, 2023).
The second force is reputational, and it is accelerating. The Forrester 2026 State of Business Buying analysis confirms that buyers are increasingly sensitive to irrelevant outreach — and that poorly targeted contact erodes vendor credibility before a conversation even begins.
Gartner predicts more than 2,000 “death by AI” legal claims by end of 2026, driven by automated, high-volume prospecting that irritates potential future customers and degrades brand equity. When AI-driven outreach systems can send thousands of personalised messages per day, the absence of formal anti-targeting is not just wasteful. It is a brand safety risk.
Every avoidable contact with the wrong audience is a preventable harm. This is the logic behind what we call Vision Zero anti-targeting — the total elimination of preventable waste by removing entire categories of unsuitable accounts from the targeting pool before any campaign activity begins. Not filtering out bad leads after the fact. Removing the conditions that produce them.
How Do Anti-Targets Differ from Negative Personas?
These are not the same thing, and conflating them creates different types of errors.
An anti-target operates at the account level. It removes entire segments from investment based on structural fit: wrong industry, wrong size, incompatible procurement process, insufficient deal size to justify servicing costs. An anti-targeted account never enters the pipeline. It never consumes SDR time, never receives a nurturing sequence, never occupies a CRM record.
A negative persona operates at the person level within accounts that otherwise qualify. It filters out specific individual profiles: a serial form-filler who downloads everything but never engages sales, or a role that has no purchasing influence whatsoever. Negative personas refine engagement within valid accounts.
Both matter. But the anti-target carries ten-times economic leverage, because it prevents the servicing multiplier from ever being applied.
What Criteria Should Guide Anti-Targeting Decisions?
Functional anti-target criteria typically fall into four categories: structural misfit (the account’s industry, size, or operating model makes the product irrelevant), economic misalignment (the deal size cannot justify the cost of acquisition and servicing), procurement incompatibility (the account’s buying process is fundamentally misaligned with the sales cycle — government procurement timelines for a product requiring quarterly iteration, for example), and historical evidence (the segment has been pursued before and conversion data confirms the economics do not work).
The discipline is making these criteria explicit and enforcing them in the campaign targeting logic, not just the strategy document. An anti-target that exists on a slide but not in the exclusion lists is decoration.
And the criteria must be reviewed. Markets shift. A segment that was structurally unfit two years ago may have evolved through new procurement processes or technology adoption.
Equally, a segment that converted reliably may have deteriorated as competition intensified. Anti-targeting is a continuous discipline, reviewed against conversion data at least quarterly, not a one-time exercise filed away after the strategy offsite.
How the Three Layers Work as a System
The three constructs are a single targeting system, not independent exercises run on separate timelines by separate teams.
The ICP sets the boundary. It defines the universe of accounts where the product can deliver value, where deal economics justify investment, and where the conditions for long-term commercial success exist.
The buying group maps the people within those accounts. It identifies who influences the decision, what each role evaluates, and how the demand generation system must engage multiple stakeholders simultaneously — rather than treating leads as isolated individuals who happen to share an email domain.
The anti-target sharpens the boundary by exclusion. It removes segments where the economics collapse, where structural misfit would trigger the five-to-ten-times servicing multiplier against data that will never convert, and where automated outreach would create more brand damage than commercial opportunity.
The sequence matters: ICP first, buying group second, anti-target third. But all three defined before execution begins. This is what separates account-based marketing programmes that produce pipeline from those that produce activity reports.
ABM without a complete targeting layer is just expensive broadcasting aimed at a named account list. An organisation that launches campaigns before completing this sequence is not generating demand. It is generating activity, and activity without targeting precision is the single most expensive mistake in B2B demand generation.
With only 10% of the total addressable market in-market at any given time and buying groups arriving with preferences already formed before first contact (6sense 2025), the targeting layer is not optional infrastructure. It is the foundation on which everything else — content, campaigns, scoring, nurturing, sales engagement — either works or does not.
→ The Demand Generation Operating Model translates this targeting layer into the three-phase system (create demand, capture demand, accelerate pipeline) that makes the precision operational. → What Is Demand Generation establishes the system model this targeting layer sits beneath.
What Comes Next
The targeting layer is a prerequisite, not a destination. Defining the ICP, mapping the buying group, and building the anti-target creates the foundation. It does not create demand on its own.
What it does is ensure that every subsequent investment — content, campaigns, nurturing, sales engagement — is directed at accounts that can convert, aimed at the people who collectively make the decision, and protected from the waste that compounds when targeting is imprecise. The difference between organisations that build this foundation and those that skip it shows up in pipeline conversion rates, cost per opportunity, and the relationship between marketing spend and revenue.
The demand generation operating model translates the targeting layer into execution: how the three-phase system (create demand, capture demand, accelerate pipeline) operates against ICP-qualified accounts, engages buying groups at the role level, and uses anti-target exclusions to prevent waste before it starts.
The targeting layer comes first. Everything else follows from the quality of that decision.
