- الرئيسية
- المدونة
- عمليات الوكالة
- How to Agree on One Source of Truth for Ad Metrics: A Framework
How to Agree on One Source of Truth for Ad Metrics: A Framework
Alessandro Conti
خبير تسويق أداء أول
A source of truth for ad metrics is not the perfect number you have been hunting for — it is an agreement. Specifically, it is one agreed figure per decision, written down in advance and read by everyone from one shared surface. This framework gives you the four steps to establish that agreement so your meetings stop opening with reconciliation and start opening with the decision. The human keeps every judgment call; what disappears is the argument about which raw number to start from.
Quick answer: To agree on one source of truth for ad metrics, do four things: inventory the numbers and where they diverge, write a convention mapping each decision to the figure that governs it, move everyone onto one shared cross-channel view, and never relitigate an agreed metric mid-meeting. The agreement, not a perfect figure, is what makes it work.
Step 1 — Inventory the numbers and the divergence
You cannot agree on one number until you know how many you have. Most teams discover they are running on four to six: Ads Manager, the tracker, a manual spreadsheet, the client deck, and sometimes a separate analytics tool and the billing export. List every figure your team quotes in a typical week and, beside each, write where it comes from and what window it uses.
Then quantify the spread. Take one week, pull the same metric from every source, and put the numbers in a row. The gap is your reconciliation tax — and seeing it as a single number ("our revenue figure varies by €5,200 depending on who you ask") is what converts a vague frustration into a problem the team will actually fix.
The inventory is uncomfortable on purpose. When a team sees that its "revenue" exists in five versions spanning a 30% range, the instinct to find the one true figure dies — and the more useful instinct, to agree which figure governs which decision, takes its place. You cannot manage a divergence you have never measured.
Step 2 — Write the convention before the next meeting
This is the cheap step almost nobody does. Create a single page that maps each recurring decision to the one figure that governs it. It does not need to be sophisticated; it needs to be written and agreed.
A workable starting convention:
| Decision | Governing figure | Source | Window |
|---|---|---|---|
| Budget pacing | Spend | Platform billing | Day |
| Kill / scale a campaign | Cost per result | Shared cross-channel view | Trailing 7 days |
| Profit & true ROAS | Contribution margin | Tracker reconciled to orders | Order date |
| Client narrative | Platform-attributed result | Shared view, labelled | 7-day click |
Each row removes one future argument. The point is not that these are the only correct choices — your business may reconcile profit differently. The point is that the choice is made once, in advance, and written where everyone can see it. For the profit row specifically, lean on a deliberate method like the reported vs true ROAS framework so the figure you agree on is defensible rather than arbitrary.
Step 3 — Move everyone onto one shared surface
A convention on paper still leaves four people reading four exports. The reconciliation only ends when the exports do — when everyone reads the governing figure from one place, pulled the same way, on the same cadence.
This is where consolidation stops being a spreadsheet project and becomes infrastructure. Instead of each person exporting from each platform and reconciling by hand, you connect every channel once and read one cross-channel view. The mechanics of doing this for a single platform's many accounts are covered in the guide on consolidating Meta ad account reporting; the same logic extends across Google, TikTok and the rest.
The shared surface is what makes the convention real. A rule that says "kill decisions use cost per result from the shared view" only works if there is a shared view everyone can open. Without it, four people apply the convention to four different snapshots and arrive back at four different answers — the convention quietly fails while everyone believes they are following it.
Wevion is built for exactly this step. It connects Meta, Google, TikTok, Taboola and more through their official APIs and presents one cross-channel view, synced about every 15 minutes. Everyone — buyer, account manager, and through a shared report the client — reads the same spend and the same cost-per-result, pulled the same way. It is not live (the sync runs roughly every quarter-hour, not instantly) and it does not act on its own; it removes the manual reconciliation, not the judgment.
Step 4 — Hold the discipline
The framework's hardest step has no tool. It is behavioural: in the next ten meetings, when someone says "but Ads Manager shows more," the team has to answer "yes, and for this decision we agreed to use the shared-view figure" instead of reopening the reconciliation.
The convention is only worth anything if it is treated as settled. The moment one metric gets relitigated mid-meeting, all of them are back in play, and you are back to arguing arithmetic instead of deciding. Protect the agreement the way you would protect a budget cap — as a rule you do not casually override.
Discipline is the difference between a convention that works and a document that gathers dust. The teams that win this are not the ones with the best dashboard — they are the ones who decided, once, which number governs which decision, and then refused to renegotiate it under pressure. The tool gives them the shared figure; the discipline gives them the meeting back.
What you keep, and what you give up
What you give up is the fantasy of a single perfect number. It is not coming, and chasing it is how teams burn quarters building reconciliation models that shatter the next time a platform changes its attribution.
What you keep is judgment — all of it. Deciding which figure governs which decision is a human call. Deciding what to do with the figure once everyone agrees on it is a human call. The framework does not automate the thinking; it automates the data plumbing underneath the thinking, so the thinking is the only thing left to do in the meeting.
The behavioural shift this produces is the whole return on the effort. A 2024 analysis from data-integration vendor Funnel noted that marketers routinely manage data across dozens of sources without a reconciliation layer — which is precisely why fragmented numbers are the default. And Salesforce's 2024 State of Marketing report put the average number of data sources marketers use at roughly fifteen and rising. Against that backdrop, a team that has agreed one figure per decision and reads it from one surface is not just tidier — it is structurally faster, because it has removed the single biggest source of meeting friction there is.
If you want the shared-surface half of this framework without building it yourself, that is what Wevion's one cross-channel view is for. It will not pick your convention or make your kill calls — those stay with you — but it gives the whole team one figure to start from, across every channel, synced about every 15 minutes. The 14-day trial runs alongside the permanent free plan, so you can connect your channels and stand up the shared surface this week. For where this sits among full platforms, compare it against the best ads management platforms.
This guide is part of our agency tools hub — explore the full cluster for related playbooks.
الأسئلة الشائعة
The Ad Signal
رؤى أسبوعية لمشتري الوسائط الذين يرفضون التخمين. بريد إلكتروني واحد. فقط إشارات.
مقالات ذات صلة
Why Ad Teams Disagree on Metrics (Instead of Acting on Them)
When Ads Manager says one thing, the tracker says another, and the client deck says a third, the meeting stops being about the campaign and becomes a debate about whose number to believe. This is a narrative look at why ad teams argue about which figure is right instead of acting on it, where the conflicting numbers actually come from, and what every reconciliation argument quietly costs in trust and momentum.
Cross-Channel Ad Analytics: Fixing the Fragmented Reporting Problem
Every platform reports spend, CPA and ROAS its own way, in its own currency, on its own delay. Stitching them into one truthful view by hand eats hours and still drifts. This guide explains the fragmented reporting problem and how a cross-channel ad analytics layer fixes it without a spreadsheet ritual.
Reported ROAS vs True ROAS: A Framework You Can Trust
Platform-reported ROAS and true ROAS are two different numbers, and confusing them quietly steers your budget wrong. This is a practical, human-in-control framework for calculating a true ROAS you can trust, deciding which number to use for which decision, and building the habit into a weekly routine.