Not everything that creates value can be measured

One of the more interesting pieces of research I've read recently explored what’s referred to as the "missing percentage".

The premise is straightforward. Most purchasing decisions don't begin five minutes before somebody clicks an advert or responds to an email.

By the time that final interaction happens, people have often been forming an opinion for weeks, months or sometimes years. They may have heard your name from a colleague, attended an event, read something you've written or simply built confidence through repeated exposure over time.

That final interaction tells us where someone decided. It doesn't necessarily tell us what gave them the confidence to decide.

Recent research from BERA.ai put some numbers behind that idea. After analysing a wide range of marketing mix models, they concluded that, on average, around 15% of what is typically treated as baseline revenue could actually be explained by the cumulative effect of previous brand-building activity. In other words, value had been created, but the models were treating it as though it had always been there.

Organisations naturally place great importance on the things they can measure and they should. Good decisions depend on good evidence and understanding performance matters.

The challenge is that some of the things which influence long-term success don't lend themselves to the same level of precision. Trust, reputation, relationships and consistency all shape how people make decisions, but they tend to build gradually and reveal themselves over time rather than appearing against a single campaign or initiative.

At the same time, attribution models are becoming increasingly sophisticated at measuring interactions at the bottom of the funnel. The danger is that the quality of that data can sometimes lead us to overestimate the importance of the steps it measures.

If almost everyone can produce more content, more quickly, then the differentiator becomes something else. People still have to decide who they believe, whose advice they trust and which organisations they feel confident choosing.

The research is interesting because it highlighted something much broader than marketing attribution. Organisations can become very good at measuring outcomes without always understanding everything that contributed to them.

Evidence will always matter, but so will judgement.

Some of the most valuable assets an organisation develops reveal themselves slowly, accumulate over years and don’t appear in a dashboard in a way which truly demonstrates how vital they are.

Perhaps we've become better at measuring decisions than understanding what drives them.