Attribution in the AI era: measuring what escapes measurement
More and more research ends in an answer instead of a click. What that means for reporting, budget allocation and the question of channel contribution.
When a customer prepares a purchase decision inside a chat window and only then goes to the brand directly, the entire research path shows up in no report. Direct traffic grows, paid channels look worse, and nobody can say whether that is good or bad.
What stays visible
- Branded search volume as an early indicator of attention created elsewhere.
- Direct traffic and its trajectory relative to media spend.
- Self-reported attribution at checkout: “How did you hear about us?” — imprecise, but directionally strong.
- Geo tests: switch off budget in comparable regions and measure the difference.
No item on this list replaces a clean click path. Together they produce a more robust picture than an attribution model that only sees the final metres.
Consequences for planning
Steering purely by attributed conversions systematically shifts budget to where measurement still works — not to where the effect happens. Incrementality tests take more effort than a dashboard glance, but they answer the question that was actually asked.
Author
Keep reading
Creative production: how much does AI really take over?
Generated creatives have become cheap. That did not reduce the work — it moved it. A look at where the bottleneck sits now.
Building a small RAG system: what you actually need
A RAG system sounds like heavy infrastructure. For a small project, four building blocks are enough: documents, an embedding model, a vector database, and a language model. Here's how they fit together conceptually, using Google Gemini and ChromaDB.
Agents in the mid-market: what really happens after the pilot
Almost every company is running an AI pilot by now. Few make it into day-to-day operations. Why that is — and what the exceptions do differently.