MarketingThe marketing lead at a 20–200 person company

When six activities run at once, I want to know which actually brings customers, so budget goes where it works instead of being spread evenly for comfort.

Khi tôi chạy sáu hoạt động cùng lúc, tôi muốn biết cái nào thật sự mang lại khách, để dồn ngân sách vào chỗ có tác dụng thay vì chia đều cho yên tâm.

Job context

Who: A marketing lead running a small budget
When: At each quarterly budget review
Trigger: Being asked what to cut and having no answer
Situation: Buyers pass many touchpoints before deciding
Constraints: Cannot switch a channel off to test because revenue drops immediately

One customer passes many touchpoints
High
Crediting the last touch is the easiest and the most wrong.
Root cause: Only the last touch leaves a technical trace
Measurable channels crowd out unmeasurable ones
High
Budget flows to where the reports look good, not to where the effect is.
Root cause: Whoever allocates must defend the decision with numbers
Brand effects arrive months later
High
The reporting cycle is a quarter; the effect cycle is a year.
Root cause: The review rhythm is shorter than the effect rhythm
Sales and marketing claim the same credit
Medium
The same customer appears in two reports.
Root cause: There is no shared definition of who created an opportunity
Cannot switch anything off to test
Medium
The only conclusive test costs revenue.
Root cause: Nobody accepts the risk of a real experiment

A solution is not the same thing as a product — a customer can hire a behaviour or a workaround too.

Last-touch attribution analytics
Product
Technically accurate and causally wrong, and it is the default everywhere.
Ask new customers how they heard
Behaviour
Imprecise data that is roughly causal — the exact inverse of analytics.
Run a geographic holdout test
Behaviour
The only conclusive test, and few small companies dare run it.
Split budget evenly
Workaround
The defensive move when nothing is known, guaranteeing no channel reaches a working threshold.
Tag every link
Workaround
Marketers build the attribution the tools lack, and it breaks at every offline touchpoint.

The trade measures the last touch because it leaves a trace, while allocation needs to know which touch created the demand.
Why existing solutions fail: Technical traces exist only where there is a click, and everything that creates demand — a conversation, an article, a recommendation — has no click.
Potential opportunity: Pairing click data with the direct question and trusting the answer over the trace.
Whoever allocates must defend with numbers, so the measurable channel always beats the effective one.
Why existing solutions fail: The personal risk of spending on an unprovable channel far exceeds the reward if it works, so the safe choice is always the measurable one.
Potential opportunity: A share of budget declared unmeasurable from the outset, so it never has to compete on numbers.