Stop. Counting. Leads

Walk into almost any European marketing meeting and you will hear a version of the same sentence. We hit the lead-MQL target, Pipeline did not move.

Then the hunt for a culprit starts. The content was too generic. We selected the wrong AI, Sales did not follow up fast enough. The nurture flow needs rebuilding. We should buy intent data. And then there is always the one that suggests a new tool. Etc..

I came back from the US eighteen years ago. Ever since, I have listened to marketing and sales leaders complain about the same thing, in Brussels, Amsterdam, Paris and London. I was not sitting in those meetings any more. I had sat in enough of them, and that is exactly why LeadFabric was founded. The culprit is almost never anyone in the room. It is the number at the top of the report.

A lead is one person. A decision is not.

A lead is one person, one form (assuming it gets filled in), one score, one handover to sales. That is the unit almost everything in B2B marketing is built on. Scoring assumes it. Routing assumes it. Attribution assumes it. Your agency invoices against it. AI won't fix it. And also... sales leaders still seem to want it.

Gartner for Marketing & Communications puts a typical complex B2B purchase at six to ten decision makers. 6sense counts 11. Others say closer to 20. The exact number does not matter. What we buy and sell keeps getting more complex, and it has to satisfy more requirements, more departments and more rules than it did five years ago. So the number keeps climbing. The only thing worth stating with confidence is that it is not one individual.

On larger deals my experience says the group runs bigger still, and that its composition shifts halfway through. The person who reads your paper is rarely the person who decides, and often is not in the room when it is decided. Meanwhile a security architect, a procurement lead or a sceptical peer of your sponsor can stop the whole thing without ever visiting your website.

Side note: ask yourself a question. How much of your marketing goes into content built to disarm those people? Not to convince your champion, who is already convinced, but to answer the objection before anyone says it out loud. Because once a detractor voices it in the room, the deal can be already gone, and nothing you publish afterwards will reach them.

So you optimise, report and get paid on a unit that does not buy. Everything downstream inherits that error. No amount of better AI copy repairs it. Not new tooling will.

Nobody in the chain wants to change the unit

This is the part I find hard to say politely at conferences, so I will write it down instead.

Agencies are measured on and even sometimes paid per lead, per MQL, per cost per lead. Change the unit and you change the invoice, the staffing model and the case studies. Most would rather add a new channel.

Most marketing technology holds a person and an account, and nothing in between. There is no place to put a buying group, so it never gets stored, and what you cannot store you cannot orchestrate, engage, score, route or report. We have been dragging this problem along since 20+ years.

Across the board...

Technology like marketing automation and CRM: Salesforce, supposedly the CRM leader, has no buying group object. The usual workaround is Opportunity Contact Role, and it has one fatal catch. An Opportunity Contact Role cannot exist until sales has created an opportunity. The group formed months earlier. The object turns up exactly when marketing no longer needs it. Instead marketing was chasing just one person and was content they were able to hand one over when they 'converted' one. Only to discover the contact was not part of ... an active buying group.

Vendors solve this in vocabulary long before they solve it in the data model. Putting the word NEXT in front of your "new" marketing cloud platform does not do the trick. The data model has to carry an object called Buying Group with the same standing as Contact and Account, populated from the first signal, not from the first opportunity.

Reputation, Branding, Influencer marketing and Search: LLM chatbots are replacing search, and a tooling category has arrived to measure what that means for brands. SEMrush is moving in this direction. AEO specialists like Profound are built for it. They tell you how visible your brand is when someone asks a chatbot a question.

Useful. But two mistakes are being made here, and they compound.

1. The first is treating search as the buying cycle.

In transactional B2C, someone types "best running shoes under 100" and buys. B2B doesn't work that way. Nobody opens a chatbot and asks "give me the best 5 supply chain management tools" as their opening move — and if they do, they're at the very end of a process that started months earlier, by which point the thinking is done and the shortlist is largely set.

The questions that actually shape the outcome get asked far upstream, long before anyone is looking for a vendor. Is this even our problem. What does good look like. How have others solved it. What would this cost us to ignore. Search is one step among many, and not the one where the deal is decided.

2. The second is treating it as one person asking.

There is no single buyer. There is a buying group, and each member arrives at a different moment with a different question in different language. A CMO asking about CRM functionality is not asking what a revenue ops leader is asking, and neither is asking what the IT architect is asking. Different vocabulary, different concerns, different definition of a good answer.

Yet almost every technology vendor or thought leader on LLM visibility and AEO skips this entirely. The challenge gets flattened into "we need to be discoverable." No mention of personas. No mention of whether the brand is discoverable in the language each member of the group actually uses.

Which is the part that matters. A single brand visibility score tells you almost nothing. The unit of measurement isn't the brand — it's persona × journey stage. Until the tooling and the thinking catch up to that, we're measuring the easy thing and calling it the important one.

The elephant in the room

Those are just two examples of the same thing: the elephant in the room gets ignored. Even now, in the middle of an AI transformation everyone can feel, the buying group barely comes up. We see lots of new tools and solutions being developed to help B2B marketers but few actually take into account the buying group.

Orchestrating against it is harder than most people assume. Your tactics, your programs, your campaigns all have to run as a function of how well you're engaging each member of that group — and how badly you're engaging the rest. Get it right and it opens an entire new set of possibilities. It also means considerably more work.

This is where AI actually earns its place. Rules-based automation cannot carry it. The permutations are the problem: every member sits at a different level of engagement, and the combinations multiply far faster than any rule set you're willing to build and maintain.

Because what you say to a given member, and when, depends on how far you've progressed with the rest of the group. A CFO who surfaces while the IT architect is still unconvinced needs something different from the same CFO arriving after IT has signed off. Same person, same role, same asset sitting in the library — different moment in the group's collective progress, different thing to say.

That conditionality is the part rules can't reach. It's also the part AI can.

Accountability

Lead volume with a cost per lead beside it makes a clean board slide. Group coverage and group progression do not, at least not until someone builds the reporting. So the board keeps asking for the number that is easy to produce, and the number that is easy to produce keeps everyone pointing at the wrong problem.

Three self-interests, all pulling the same way. That is why this survives every new tactic.

Check the data model, not the pitch deck

Some of the martech vendors in this market now say buying group somewhere on the homepage. Some of them mean it. The claim is testable, and it takes an afternoon. Sit with your own admin and ask three things.

• Where does a buying group exist as an object, separate from a person and an account?
• Can we score that object, and does the score change when a second or third role joins?
• Can we report progression at group level to someone outside marketing?

If the answer to the first question is no, the other two are already decided. I have watched teams spend a year and a serious budget on a platform that could never hold the thing they had just paid a consultancy to research.

Side note, and probably the real easter egg in this post: we're developing the European GTM for Phave, a new MAP platform built by one of Marketo's co-founders. AI-native, not rules-based, buying group at the core of the data model. Long overdue, and IMO the first real answer to "the playbook is broken" rather than another post about it. General Availability (GA) is later this year, but I can show you a first glimpse now. Ping me or the team.

Four questions, before any tooling

These are the four we ask on day one with every client. The answers decide everything after them, including whether we are the right partner at all.

• How do your customers actually buy? Which processes they run, which roles take part, who holds a veto, and above all which triggers pull someone into a buying group in the first place. A reorganisation. A renewal date. A new VP. A regulation. A project that failed last year and left someone with a reputation to repair. Find out more here: https://leadfabric.com/buyer-insights-and-personae
• Where is the data on those triggers, and on intent across the whole group rather than one visitor? Some of it sits in your own systems and nobody has joined it up. Some of it you buy or build. Clay and its equivalents have made this far cheaper than it was three years ago, which is the main reason this conversation is possible now and was theoretical in 2021. Find out more here: https://leadfabric.com/clay-offering
• Does your technology carry the group in its data model? See above.
• And who designs and runs the work on top of that? Because insight that arrives in a slide deck and dies there is expensive entertainment.

Campaign and program are not the same word

We use these two words precisely, and I would encourage you to as well, because the sloppiness costs money.

A campaign is a theme based go to market approach for one product/service/solution to market combination, across every marketing function. Theme? Think about NIS2 if you're in financial services, or Back to School. A campaign makes sure all the supporting function are articulating your solution in context of the theme.

Function? Think branding, demand generation, sales engagement and sales enablement, plus the intelligence layer underneath that feeds all four.

Before you build theme based campaigns you of course have to have your standard GTM defined, your ICP and you overall messaging to each of the members of the buying group. We call it "generic" campaign framework, many lately started calling it GTM nowadays. We comply.

But a program is one building block of that campaign, or of the wider theme agnostic GTM strategy. An account scoring programme. An ABM demand generation program aimed at existing customers. One component, built properly.

Most agencies sell a program and call it a campaign. That is why so much B2B marketing looks busy and moves nothing. The parts run. The whole was never designed. If your sales enablement knows nothing about the roles your demand generation is reaching, you do not have a campaign, you have a collection of programs with a shared logo.

We have said this many times. Strip out the complexity of the roles, settle for a name to hand over, and you are not adding value, you are adding admin. Sales notices. Not in one quarter, but over a year or two, sales leadership quietly stops believing marketing can help them, and that trust is far harder to win back than a pipeline number.

Building programs is what we do, weaving into your GTM Fabric and building campaigns as well. When we build programs we do it in context of the GTM (campaign). Sometimes the client already have defined this sometimes we work with the institutional memory of the team. But anyone who tells you they are good at advertisement 'campaigns' to help you create revenue and that does not understand (or asks anything about) how your buyer's buy, what the buying groups look like and how you treat the individual response, how you enrich data and where you source signals from, is most likely not helping you.

See it rather than read it

Three recordings, if you want the longer argument with the data behind it.

Stop chasing leads, on the MQL question and five more: https://www.youtube.com/watch?v=ow2kcpVBFE4

Where B2B marketing stands: https://www.youtube.com/watch?v=flwW-0FeMRY

A client, in their own words. Watch from six minutes in, that is the part that matters: https://www.youtube.com/watch?v=EIwZiKL4zDg

We built LeadFabric as a consultancy, a data partner, a technology implementer and an agency in one team. Not because that sounds complete on a website, but because a buying group falls apart at every handover between those four.

Want to hear more? The team is here: https://leadfabric.com/request-for-proposal