We decided in the room, not in the data
In most conventional B2B companies the marketing budget is set by whoever has the most experience and the most conviction. Experience is not worthless — but it is a hypothesis, and hypotheses are supposed to be tested.
- Marketing Analytics
- Attribution
- Data-Driven Decisions
The annual marketing budget meeting in a traditional B2B business follows a reliable script.
Someone says the industry exhibition is non-negotiable, because that is where the customers are. Someone says the trade publication works, because a customer mentioned the advertisement once. Someone says our buyers are engineers, they do not use LinkedIn. The most senior person in the room agrees with two of the three, and the budget is allocated.
Every one of those statements might be true. Nobody in the room knows whether they are, and that is the actual problem.
Experience is a hypothesis
It is easy to frame this as data versus intuition, with intuition cast as the villain. That framing is wrong and it loses the argument, because the people with twenty years in the industry usually do know things you do not.
The accurate framing is narrower:
Experience produces hypotheses. Good ones, often. But a hypothesis that has never been tested and cannot be falsified is not knowledge — it is a habit with seniority attached.
"Our buyers are engineers, they do not use LinkedIn" is a testable claim. It costs very little to test. The reason it never gets tested is not stubbornness; it is that no measurement layer exists, so there is no mechanism by which the room could ever discover it was wrong.
Why the habit persists
The original evidence was real. The exhibition genuinely did produce orders — in 2011, when it was the only place buyers could see the equipment. The observation was correct; it has simply expired, and nothing has been measured since to notice.
Failure is invisible. If a campaign produces nothing, you find out. If a ₹30 lakh exhibition produces nothing, you attribute the three orders that arrived that quarter to it anyway, because there is no way to tell.
Attribution defaults to whoever tells the story. In the absence of tracking, the enquiry gets credited to the last person who touched it — usually a salesperson, occasionally the exhibition, almost never the six months of search visibility that actually made the buyer aware you existed.
What to instrument, in order
The instinct is to build a comprehensive dashboard. Do not. Instrument five things, in this sequence, and stop.
1. Enquiry source, captured automatically. Not "how did you hear about us?" on a form — buyers answer that badly. UTM governance on every outbound link, and source captured at the point the enquiry is created.
2. Time to first response. The single most controllable variable affecting conversion, and almost nobody in conventional B2B measures it.
3. Enquiry-to-order conversion, by source. This is the number that ends arguments. Not enquiry volume by source — conversion. A channel producing 30 enquiries at 25% conversion beats one producing 200 at 2%, and volume reporting hides that completely.
4. Cost per qualified enquiry, by channel. Include the fully loaded cost of conventional channels: stand, build, travel, staff time, collateral, logistics. The comparison is usually uncomfortable, which is the point.
5. Pipeline value by segment and stage. Once leadership can see pipeline rather than enquiry counts, marketing stops being a cost line.
Five metrics. Everything else is decoration until these are trustworthy.
The honest caveat
Attribution in long-cycle B2B is genuinely hard, and anyone who tells you their model is precise is overselling it. A buyer who reads three application notes over eight months, sees you at an exhibition, asks a colleague, then searches your brand name directly and fills in a form is recorded as direct traffic. That is not a solved problem.
So do not claim precision you do not have. Claim direction. You do not need to know exactly what produced an order to know that one channel produces qualified enquiries at a fifth of the cost of another. Directional truth, consistently measured, is enough to reallocate a budget — and over-claiming precision is the fastest way to lose credibility with a finance team that understands measurement error better than most marketers do.
What changes when the number exists
The change is not that people become data-driven. It is that the conversation changes shape.
Before, the discussion is about opinion, and opinion is settled by hierarchy. After, the discussion is about whether the measurement is right — and that is an argument anyone in the room can participate in, including the person with the least seniority and the best evidence.
That shift is worth more than any individual insight the dashboard produces.
What I would carry forward
- Do not attack the experience. Test it. Framing it as a hypothesis lets the senior person stay right about the past while you find out about the present.
- Measure the before state. It is free exactly once, and "it improved" is a much weaker claim than a number.
- Five metrics, trusted, beat forty that nobody believes.
I write about digital transformation, B2B marketing and applied AI. Get in touch if you are working through something similar.