When you can’t trace revenue back to its source

The attribution gap doesn’t form because your team is unsophisticated or because you chose the wrong tools. It forms because the modern marketing stack was never designed as a connected system. It evolved as a collection of point solutions: one tool for ads, one for email, one for web analytics, one for CRM. The average B2B company uses 12 to 16 marketing tools, each built by a different vendor, each tracking users with its own cookies, its own identifiers, its own definition of a “conversion.” Nobody sat down and designed an architecture where a single customer could be recognized across all of these systems. The stack grew tool by tool, each purchase solving an immediate need, and the gaps between them became invisible until someone asked a question that required data to cross the boundary.

Here’s how this plays out concretely. A buyer clicks a LinkedIn ad on Tuesday morning from their work laptop. They browse your site for four minutes and leave. On Thursday, they return via organic search on their phone, read a case study, and subscribe to your newsletter. Two weeks later, they open three emails, click through to your pricing page from the last one, and submit a demo request. That buyer’s journey touched 8 distinct interactions across 14 days and 2 devices. In LinkedIn’s reporting, this is an ad click with no conversion. In Google Analytics, it’s two separate anonymous sessions followed by a known conversion attributed to email. In your CRM, it’s a new lead with source “website.” Three systems, three stories, zero overlap. The buyer’s actual journey, the one where LinkedIn created the initial awareness, organic search built consideration, and email closed the loop, doesn’t exist in any single report. You’d have to reconstruct it manually, and nobody has time to do that for every deal. One company we worked with had 1,200 closed deals per year. Reconstructing each journey manually would require an estimated 15 minutes per deal: 300 hours of analyst time annually, roughly $45,000 in loaded labor cost, just to understand what already happened.

The organizational response to this confusion is predictable and damaging. Teams default to last-click attribution because it’s the simplest model and the only one that gives clean, unambiguous numbers. CFOs accept it because it produces a tidy spreadsheet. Channel owners accept it because at least someone gets credit. But last-click systematically lies about how B2B buying actually works. Research consistently shows that the average B2B purchase involves 6 to 20 touchpoints over 27 to 90 days. Last-click ignores all but the final one. It favors the channel that catches the customer at the moment of decision and erases the channels that created the conditions for that decision to happen. One of our clients discovered that when they paused their $8,500/month content program because last-click showed it “generated zero conversions,” their paid search cost per acquisition rose 38% within two months, from $185 to $255, because fewer buyers were arriving pre-educated and brand-aware. They were treating the symptom while deepening the disease.

The technical debt here is real but often misunderstood. The problem isn’t that attribution is philosophically hard. It is, but that’s not what’s killing you. The problem is that most organizations lack even the basic plumbing required for attribution to be attempted. In an audit of 23 B2B marketing operations, we found that 78% had inconsistent or missing UTM parameters on more than half their campaign links. Sixty-one percent had no shared identity layer connecting anonymous sessions to known contacts. Eighty-three percent tracked only pageviews, not meaningful engagement events like content depth, video completion, or feature-page return visits. Only 3 of the 23 had a data warehouse that unified ad platform, analytics, and CRM data. Without these foundations, attribution isn’t inaccurate. It’s fiction.