Every business that advertises on more than one platform eventually builds a spreadsheet to combine the numbers. It never quite agrees with the platforms, and the discrepancy is not an error — it is the platforms measuring different things.
Understanding why is more useful than trying to make the totals match.
Why the numbers disagree
Attribution windows differ
Each platform credits itself for conversions within its own window after an interaction. Those windows differ, and they are configurable. A conversion two weeks after a click may be counted by one platform and not another.
Every platform claims the same conversion
A customer sees an ad on one platform, searches, clicks an ad on another, and buys. Both count it. Add the platform totals and you have more conversions than sales — which is why the spreadsheet never ties out.
"Conversion" means different things
One platform counts a view-through. Another counts only clicks. Your own analytics counts sessions with a purchase. These are three different measurements of one event.
Currency and timezone
Platforms reporting in different timezones split days differently. Over a month it mostly washes out; for a specific day it does not, which matters when evaluating a campaign that ran for three days.
What to do instead of reconciling
Pick one source of truth for revenue
Your own system — the store, the CRM, the books. That is what actually happened. Platform numbers become directional indicators of which channel is working, not the record.
Compare against a consistent baseline
Rather than asking which platform produced a sale, track total business against total spend over time. Blended cost per acquisition — all marketing spend divided by all new customers — is less sophisticated and considerably harder to fool.
Use tracking parameters consistently
Tag every link with the same conventions everywhere. Inconsistent tagging is the most common self-inflicted reporting problem, and it cannot be fixed retrospectively — data collected untagged stays untagged.
Agree the convention, write it down, and apply it to every campaign including the ones someone sets up quickly.
Ask on the enquiry form
"How did you hear about us", recorded somewhere durable. Imperfect, biased towards what people remember, and still valuable — particularly for channels that are hard to track, like word of mouth, which is frequently the largest.
When a spreadsheet is genuinely enough
One or two channels, modest spend, and someone reviewing monthly: a spreadsheet is fine. Reporting tooling costs money and time to maintain, and for a small programme the manual version is cheaper.
The case for tooling arrives with several channels, multiple people needing the same numbers, or decisions being made weekly rather than monthly — because at that point the consolidation effort recurs often enough to automate.
The number that matters most
Customer acquisition cost against customer lifetime value. Almost every other marketing metric is a proxy for these two, and businesses that track them make better decisions than businesses tracking twenty platform metrics.
Neither is visible in a platform dashboard, because both require data the platform does not have.
About the author. Kislay Chandan has worked with schools since 2014. Rather than gathering requirements from the outside, he worked inside a school to see what actually goes wrong day to day, and built what he found into the software. He still works with Shemrock Glitters in Patna — most changes to the product start with something that came up there.
Gallery
