Why does every tool claim the same sale?
Each tool sees its own part of the path and credits itself under its own rules. Meta counts a sale if it followed a click on a Meta ad within its window. Google Ads counts it if a Google ad was clicked. Google Analytics 4 looks across channels and shares the credit out. Each is consistent with itself. The trouble starts when their totals are added together, which can easily come to more sales than you made.
That is why attribution figures are for comparing campaigns inside one tool, and why your real sales total is the referee between tools. Each model and window gets its own section in the full guide.
Who gets the credit for one Bunbury bridal appointment? (example figures)
A bridal boutique in Bunbury takes appointments online. An engaged couple click an Instagram ad on a Saturday. On Thursday they search the boutique's name on Google, click the boutique's own brand ad and book a fitting. Three weeks later they buy a $2,400 gown in store.
| Who is counting | The rule it uses | Credit for the booking |
|---|---|---|
| Meta Ads Manager | A sale within 7 days of a click on a Meta ad | 1 booking to the Instagram ad |
| Google Ads | Data-driven, across the Google ads they clicked (only one) | 1 booking to the brand ad |
| Google Analytics 4 | Data-driven, across every channel it saw | Shared between paid social and paid search |
| The boutique's booking form | "How did you hear about us?" | "Instagram" |
One booking, two full claims and a split. If the owner adds Meta's bookings to Google's for the month, this couple is counted twice. Judged alone, the brand ad looks like the hero, yet the couple already knew the name and would probably have found the boutique without it. The Instagram ad did the introducing, and the gown sale in store never reached any of the tools.
Which models are left in Google's tools?
Fewer than there were. In 2023 Google retired the first click, linear, time decay and position-based models from Google Ads and Google Analytics 4, leaving data-driven and last click (Google Ads Help, First click, linear, time decay, and position-based attribution models are going away). Data-driven is the default for most conversion actions. Meta sets its credit with attribution windows instead of models. The entries on last-click attribution and data-driven attribution explain the two that remain.
How does CoraVision handle numbers that disagree?
CoraVision reads conversions from Google Ads, results from Meta and key events from Google Analytics 4 every night, each counted under that tool's own rules. The CoraVision dashboard's Trust Score says how far those numbers can be trusted, and when the tools disagree by more than the sales you made, that gap is a finding in itself. Your own sales total, from the till or your store, stays the referee.
Questions
Which attribution model should a small business use?
Mostly the defaults, used consistently. Switching models changes the story without changing a single sale. The bigger gains come from checking the tools against your real sales total each month and asking new customers how they found you.
What is an attribution window?
The number of days after someone clicks or sees an ad during which a sale can still be credited to it. Meta's default counts a sale within 7 days of a click or 1 day of a view (Meta Business Help Centre, About attribution settings). A longer window credits the ad with more sales.
Is asking "how did you hear about us?" good enough?
It catches what tracking cannot see, such as a friend's recommendation, a radio spot or a sign on a ute. People misremember, and they tend to name the last thing they noticed. Use the answers beside the ad accounts' figures, as a second opinion rather than a replacement.