adlaneSign in Practical guide
Meta Ads campaign analysis: compare campaigns on one footing
Meta Ads campaign analysis starts with what each campaign was built to achieve and the reporting context it was read in. Delivery metrics explain exposure; attributed actions describe a later stage. Keep both visible, name the action you are counting, and do not assume every interaction carries the same business value.
By Adlane · Updated
Why Meta Ads campaign analysis starts with the intended outcome
A campaign optimised for video views and a campaign optimised for purchases will both report a cheapest action. Unless you name that action, the two look interchangeable, and the cheaper one looks better. The first question in any campaign analysis is therefore not “which campaign performed best?” but “what was each campaign asked to do, and did it do that?”
Write the intended outcome beside each campaign before opening a single chart. Reach, video plays, link clicks, leads and purchases are different stages of a journey. A campaign can be excellent at one stage and irrelevant to another. Judging every campaign by the same single number hides that.
Fix the reporting contract before comparing anything
Select one ad account, one currency and a complete period. Record the reporting level (campaign, ad set or ad) and the attribution setting used, because changing either changes the numbers without changing what happened. If two campaigns were read with different attribution settings, they are not yet comparable.
Use complete days in the account’s timezone. Partial days and recent days with immature conversions should be labelled as such rather than quietly included. When a comparison spans a creative, offer or budget change, note the date of the change so a shift in results can be read against it.
Read spend, delivery and the named action side by side
Use this worksheet to keep the three strands separate. Spend establishes cost; delivery explains exposure; the named action explains what the campaign was credited with. Rates such as cost per result only make sense when the underlying count is beside them.
| Campaign purpose | Action to name | Delivery context to keep | Avoid |
|---|---|---|---|
| Awareness or reach | Reach within the period and level | Impressions and frequency | Judging it by purchases it was never optimised for |
| Video engagement | The specific video play measure used | Impressions and placement mix | Reading plays as proof of interest in the offer |
| Website traffic | The specific click measure used | Impressions and click definition | Mixing click definitions between campaigns |
| Leads or purchases | One named action with its attribution setting | Spend, impressions and clicks | Adding overlapping action categories together |
Illustrative example: a video campaign and a sales campaign are not rivals
Illustrative example: in a fictional account, Campaign A promotes a product video and reports 4,000 video plays at an invented cost of 0.05 per play. Campaign B seeks purchases and reports 40 purchases at an invented cost of 25 per purchase. A table sorted by cheapest result puts A on top, yet nothing in the comparison says a video play is worth anything near a purchase.
The useful reading is separate: did A deliver video plays at a cost the business accepts for awareness, and did B deliver purchases at a cost the business accepts for sales? Those two questions have two answers. The numbers above are invented to show the structure of the comparison and are not benchmarks.
Find the largest contributors to change before inspecting creatives
When a period looks worse than the last, sort campaigns by the size of their contribution to the change in spend and in the named action. A few campaigns usually account for most of the movement. Read their ad sets and ads next, and check whether the offer, destination page or delivery mix changed during the period.
Platform-attributed outcomes can overlap with claims from other providers and from analytics. They are a view of what Meta credited under its rules, not a deduplicated record of business sales. Keep business order totals in a separate column and reconcile them separately rather than treating the Meta count as the sales figure.
Run the same Meta Ads campaign analysis in Adlane
In Adlane, connect Meta and select the ad account that belongs to the business you are reviewing. Ask a bounded question that carries the contract with it, such as “Compare campaigns for the last two complete weeks using the account’s current attribution setting, naming the result each campaign was optimised for.” Review the evidence it returns: the counts, the level, the dates and the attribution setting should all be visible.
Any change that follows, whether a budget adjustment or a new campaign, is prepared as a proposal that you review and approve; nothing is applied from the conversation itself. New campaigns start paused. Availability of fields depends on your Meta access and what the provider returns. Adlane is independent and not affiliated with Meta.