Most facebook ads reporting is a list of everything the platform will export, arranged in a grid, delivered on a schedule. It gets opened, scrolled and closed, and no decision follows. The problem is not the data. The problem is that nobody decided what question the report exists to answer, so it answers all of them badly and the reader has to do the analysis themselves, every week, in their head.
Useful facebook ads reporting starts from the decision it supports. A media buyer deciding where tomorrow’s budget goes needs a different artefact from a founder deciding whether the channel deserves more money next quarter. Those two needs conflict: the buyer wants granularity and recency, the founder wants stability and context. Trying to satisfy both in one dashboard produces something that serves neither, which is why we build them separately and accept the small duplication.
Three reports, three audiences
The daily operational view exists for the person with their hands in the account. Campaign and ad set level, spend pacing against budget, cost per result against target, learning phase status, and a flag for anything that changed in the last twenty-four hours. It is allowed to be dense and it does not need to be beautiful, because one person reads it and that person knows exactly what each column means.
The weekly commercial view exists for the marketing lead. It rolls up to campaign objective rather than ad set, compares against the previous four weeks instead of the previous day, and puts platform-reported results alongside the numbers from the back office. Its job is trend and divergence, not detail. The monthly view exists for whoever signs off the budget, and it answers one question: what did this channel contribute, and is the contribution stable enough to plan against.
Comparison windows decide what the report says
Choice of comparison period changes the conclusion more than any metric selection, which is why we fix it in advance rather than letting whoever builds the view choose. Day over day is noise for almost every account. Week over week is defensible when volume is high and the business has no strong weekly cycle. Rolling seven day against the prior rolling seven day removes weekday effects and is our default for accounts that spend steadily.
Attribution settings deserve the same treatment. Pick a window, write it on the report itself, and do not change it because a different setting makes a month look better. If you need to compare under two windows, show both columns permanently rather than switching between them, because a silent change in attribution setting is the single most common reason two people looking at the same account reach opposite conclusions.
What facebook ads reporting should leave out
Cutting is most of the work in facebook ads reporting. A metric earns its place only if someone can name the action they would take if it moved. The items we remove from almost every dashboard we inherit:
- Reach and impressions at the top of a performance report, where they read as achievement rather than as cost inputs
- Relevance and quality rankings shown without the cost figures that give them meaning
- Video view counts at low thresholds, which correlate with autoplay behaviour more than with interest
- Ad level detail in any view intended for someone who does not edit the account
- Percentage changes on small denominators, where a shift from two conversions to three shows as a fifty percent improvement
What replaces them is usually less exciting and more useful: spend, results, cost per result, the ratio the business actually cares about, and a short written note explaining what changed. The note is the part clients tell us they value most, and it is the part no dashboard tool generates.
Reconciling platform numbers with the business
Platform-reported conversions will not match your order data, and pretending otherwise wastes a meeting every month. Modelled conversions, view-through attribution and cross-device matching all mean the platform counts things your back office does not. The fix is not to pick a winner but to show both, track the ratio between them over time, and treat a sudden change in that ratio as the alert it is.
A stable gap is manageable information. If platform-reported revenue consistently runs a third above what you bank, you can plan around it and set targets that account for it. It is the week the ratio moves without explanation that matters, because that usually means a tracking change, a consent change or a platform release nobody told you about.
Build it to be read
Put the conclusion at the top. The first thing on the page should be a plain sentence saying whether the account is on track and what is being done about it, followed by the three numbers that support the sentence. Everything else goes below, for the reader who wants to check the reasoning rather than accept it.
Then make somebody responsible for reading it on a fixed day. An unread report is worse than no report, because it creates the impression that the account is being watched while nobody is actually watching. Reporting that nobody opens should be deleted, not redesigned, and the time it took to produce should go into the account instead.
Keep reading: Facebook Ads · Social Media Marketing