You will write a short guide to how your business reads credit across channels and which numbers it trusts.
Every few months, someone in a small business reopens the same argument. The freelancer says the ads are working because Meta shows forty leads. The owner says they are not, because the booking book shows fifteen. Someone quotes GA4, someone else quotes a different GA4 report, and the meeting ends with nobody changing their mind. The argument is not really about the numbers. It is about which numbers count, and nobody ever wrote that down.
This exercise writes it down. You produce a one-page attribution note: a short guide to how your business reads credit across channels, which numbers it trusts for what, and how it fills the gaps. The worked example is Priya's, for her maths centre in Tampines.
Start with one line for each result that matters: where the real count lives. This is the system from lesson 5.2 that records each actual sale or lead once, with a name or number behind it.
Priya's note says: new students are counted from the enrolment spreadsheet, and trial classes from the booking tool. Enquiries are counted from the website form submissions in her inbox plus the admin's weekly tally of new WhatsApp chats. If any other number disagrees with these, these win.
Add who keeps each record and how often it is updated. A source of truth that is three weeks behind is not much use for a monthly review.
Next, say which attribution model your GA4 property uses and how you read it. For most small businesses the answer is the default, data-driven attribution, from lesson 5.1, Attribution models and what each one rewards.
Priya's note says: GA4 uses data-driven attribution, the default. Once a month we look at the model comparison report against last click, to see which channels start journeys and which finish them. We do not cut a channel because it looks weak under last click alone.
If you have changed the model, say why. If you are not sure what your property uses, check the attribution settings in the admin area before you write this line.
Now list each ad platform you use, with what its numbers are for and what they are not for. This is the heart of the note, because it is what stops the arguments.
Priya's version:
Google Ads conversions: used to compare campaigns and keywords within Google Ads, and to feed its bidding. Not used as a count of enquiries or students. Meta Ads Manager results: used to compare campaigns, ad sets and ads within Meta. Not used as a count of enquiries or students, and never added to Google Ads numbers. GA4 key events by channel: used to compare channels with each other on one basis, and to see what visitors did on the site. Treated as an estimate, because of cookie choices and device switching.
Add the monthly side-by-side check from lesson 5.2: each platform's claim, GA4's count and the real count, with a note on how the gap has moved.
Then describe the question that catches what tracking misses, from lesson 5.3, What attribution cannot see. Write where it is asked, the exact answer options, and how you will combine the answers with tracking.
Priya added a required "How did you first hear about us?" question to her trial booking form. The options are: a friend or family member, a WhatsApp or Telegram group, Google search, Google Maps, Instagram or Facebook, a flyer or banner near the centre, and other, with a short text box.
Her note explains how the answers are used. Each month, the admin copies the answers for students who enrolled into the enrolment spreadsheet. Priya then puts two columns side by side: students by self-reported source, and GA4 key events by channel. Where they agree, she trusts the channel's numbers more. Where they disagree, she writes down what she thinks explains it. In her first month, 11 of 19 new students (example figures) said a friend or a WhatsApp group, while GA4 credited most enquiries to organic search and direct. Her note records the likely reason: parents hear about the centre from a friend, then search for its name.
The finished note has four short sections matching the steps above, plus a date and the name of whoever owns it. Priya's fits on one page with space left over. She sent it to the freelancer who runs her ads and to her admin, and it now sits at the top of the folder where the monthly reports go.
Review it when something changes: a new channel, a new booking system, a change to the GA4 model. Otherwise, a yearly read-through is enough.
The note will not end every debate about which channel deserves credit. What it does is make sure the debate starts from the same numbers. With Priya's note as a model, your own will look different, because your source of truth, your platforms and your customers' habits are different. What should carry over is the shape.
Produce a one-page attribution note with your source of truth, chosen GA4 model, how platform numbers are used, and your self-reported source question.
Junxiong-WFG Organisation is an authorised representative of AIA Financial Advisers Private Limited (Reg. No. 201715016G).