Why B2B Advertisers Are Abandoning Broad Reach for Vertical Precision

The B2B advertising market is going through a structural recalibration that most vendors have not fully internalised. For a decade the dominant playbook was straightforward: buy reach at scale, optimise for impressions, and hope some percentage of the audience is your ideal customer. The maths seemed to work. Budgets scaled, coverage expanded, everyone reported impressive numbers.

Then CFOs started asking harder questions about pipeline attribution, and the cracks appeared. Large portions of "reach" turned out to be irrelevant audiences, fraudulent traffic, or both. Click-through on broad B2B display sat below one percent. Marketing teams could not connect display spend to any downstream revenue metric. Stripped of vanity metrics, the performance data was damning.

What "Precision" Actually Has to Mean

Vertical precision — buying a smaller, better-defined audience instead of a larger, heterogeneous one — is the obvious correction. But "precision" has been claimed by so many vendors that the word has stopped carrying information. Almost every media seller and every intent platform now describes itself as precise.

So it is worth being concrete about what distinguishes one claim from another. There are three questions that separate them, and they are the ones a vendor should ask before signing anything.

Question One: Who Owns the Surface Where the Signal Was Recorded?

This is the question that determines everything downstream, and it is the one most often left vague.

The market has settled on a rough taxonomy. First-party signal is behaviour recorded by whoever owns the page. Second-party is signal a review site or publisher records on its own property and shares with you. Third-party is signal aggregated across a network of sites you have no relationship with, usually resolved to a company rather than a person.

The distinction matters because it decides what evidence can travel with the record. If a publisher owns the page, it can tell you which page, at what time, and what the reader did next. If the signal was aggregated from a co-op or a bidstream, the honest answer is a probability attached to a company name.

Neither is useless. Aggregated third-party data covers accounts that have never visited you, which owned-surface data structurally cannot. But they answer different questions, and a vendor paying precision prices for modelled coverage has bought the wrong product.

Question Two: Is the Deliverable a Person or a Company?

Most intent products resolve to an account. You learn that someone at a 4,000-person manufacturer looked at your category this week. You do not learn who, which means the sales motion that follows is still cold outreach — better targeted, but cold.

The gap between an account signal and a named person is where most of the disappointment in intent data lives. A rep handed "this account is showing intent" has to find the buying committee themselves, which is the expensive part of the job.

What a vendor should be asking for, and should expect to be told plainly, is what one record contains. In our case that is a named person with role, work email, LinkedIn and phone; their company, size band, industry and region; the vendors they put side by side, by name, with the timestamp of the comparison; the full engagement log of reads, downloads and pricing views; and the consent record with property, timestamp, IP and the exact opt-in wording. If a seller cannot enumerate their record that specifically, the record is probably thinner than the pitch.

Question Three: Can You Check the Audience Claim?

Vendors want audience verification, and they usually get audience assertion instead. A number appears in a deck — "2.4 million finance professionals" — with no method attached.

We deliberately publish audience sizes as bands rather than precise counts. A figure like "247,000 finance buyers" looks computed, which invites exactly the scrutiny it cannot survive; the width of an honest band is itself information. The deliverable number also moves with suppression, consent state and how tightly a program is specified, so a single precise figure would be wrong by the time anyone acted on it.

This is a commercially inconvenient position. Bands look less impressive than precise numbers in a competitive pitch. We think checkability is worth more than the impression, and a vendor evaluating any media partner should treat a suspiciously precise audience figure as a question rather than a credential.

The Economics, With Real Numbers

The case for vertical precision is usually made in the abstract. It is more useful with prices attached.

Published roundups of B2B content syndication vendors put the going rate at roughly $80 to $100 per lead at director level and above. That is the benchmark a vendor is working against when they budget a syndication program with the large established players.

Our published rate for content syndication is $45 per lead, qualified against a written spec, with out-of-spec leads replaced. A dedicated email send is $4,500. Data-as-a-Service and audience activation are quoted by category, because the price genuinely moves with spec tightness, category competitiveness, seniority, geography and committed volume.

The honest reading of that price gap is not that everyone else is overcharging. Scale costs money, and the large syndication vendors reach enterprise IT and security audiences at volumes we do not. What the gap reflects is a different cost structure: we are not buying access to an audience and reselling it, because the audience reads pages we already own.

The right comparison is therefore not cost per lead but cost per lead you can actually work. A $45 lead that arrives with the consideration set and a consent record needs less qualification than a $90 lead that arrives as a name and a job title. If that is not true in a given category, the cheaper lead is not cheaper.

Where Broad Reach Still Wins

It would be convenient to argue that broad reach is finished. It is not, and pretending otherwise makes the rest of the argument less credible.

Broad reach still does two jobs better than precision buying. It creates category awareness among people who are not yet evaluating anything, which matters enormously if you are defining a new category rather than competing in an established one. And it covers accounts that have never touched your category surface — the ones an owned-audience model cannot see by definition.

The failure mode was never breadth itself. It was using breadth as a substitute for evidence: buying impressions, reporting impressions, and calling the result demand. A vendor running both — breadth for awareness, precision for pipeline — with honest attribution on each is in better shape than one that has swung entirely to either pole.

What This Means Practically

If you are re-planning B2B media spend, the useful discipline is to make every seller answer the three questions above in writing. Who owned the page where the signal was recorded. Whether the deliverable is a person or a company. How the audience number was derived, and what it excludes.

The answers are usually available; they are just not volunteered. A seller who owns their surface will find those questions easy, and a seller who does not will tell you something interesting by how they avoid them.

That is the real shift behind the move to vertical precision. It is less about audience size than about whether the person selling you the audience can show you their working.