Marketing attribution demo · sample

Which marketing view should guide the next dollar?

A focused comparison of first-touch, last-touch and position-based attribution—showing what each view can support and where only an experiment can answer the question.

Sample figures were created solely to demonstrate the measurement workflow; they are not client results.

01 · The business question

Three tools. Three answers.

The ad platform reports strong returns. The website shows fewer conversions. The CRM tells a different customer story again.

Leadership needs to know which numbers reconcile, what each attribution view can support, and where an experiment is still required.

02 · The investment baseline

Start with measures everyone can reproduce.

Net revenue$128,400+12.4% vs prior week
Marketing spend$31,200+18.1% vs prior week
Blended ROAS4.12×Down from 4.36×
Blended CAC$38.70Up 7.8%

Sample figures for demonstration only.

03 · Before interpretation

Check the information first.

Reporting period aligned

Each source covers the same seven-day window.

Spend reconciled

Platform totals match the approved finance export.

Orders de-duplicated

Repeat transaction IDs and refunds are handled once.

04 · Attribution comparison

Show the difference instead of hiding it.

ChannelFirst touchLast touchPosition based
Paid social$42,600$31,900$38,200
Paid search$28,400$39,700$33,800
Email$16,900$25,100$22,400
Affiliate$12,300$8,500$10,700

The views answer different questions. None proves that a channel caused the sale.

05 · The leadership brief

Turn the numbers into decisions.

CONFIRMED

Efficiency softened

Spend rose faster than revenue, increasing blended acquisition cost.

INVESTIGATE

Search is closing demand

Its last-touch share is high. Check assisted paths before increasing budget.

NEXT DECISION

Test before reallocating

Keep budgets stable and design a controlled lift test for paid social.

What stays human

Reliable numbers support judgment.

The system reconciles defined measures and makes uncertainty visible. AI may organize the explanation, but it does not calculate the measures, invent a cause, or change a budget.
See Marketing Intelligence on the website →