A good agency measures success by the one number your business actually cares about: sales or revenue for an online store, qualified leads for a lead business, funded accounts for a fintech app. That number comes from your own source of truth, such as your store backend or your sales team, and never from the ad platforms' dashboards. Platform figures are useful for day-to-day decisions. They are a poor scorecard.
Start With the North Star Metric
Every client has one metric that decides whether the work is paying off. For e-commerce and DTC brands it is sales or revenue, and for more mature brands incremental revenue: the sales that would not have happened without the ads. For lead generation it is qualified leads, and ideally leads that became customers.
Clicks, impressions, click-through rate and cost per click all have a place, but none of them pay your bills. If your agency's report opens with them, ask where the sales are.
Why Ad Platform Numbers Look Better Than Reality
Ad platforms grade their own homework. Meta, Google and the rest have every reason to show strong results, and their reporting is built to do that.
The clearest example comes up all the time. Add the revenue Meta claims to the revenue Google claims, and the total is often higher than the revenue the business actually made. Both platforms take credit for the same sales.
Much of this comes from attribution windows. A platform can claim a sale because someone viewed an ad days earlier, even if they bought after a Google search. Google Analytics, for its part, tends to give Google Ads more credit. Each tool tells a version of the story that flatters its own channel.
Build Reporting on Your Source of Truth

The fix is to build reporting that pulls data from several sources and anchors it to the one that cannot be argued with:
Online stores: the store's own order data, such as Shopify
Lead businesses: the verdict of the sales agent who actually spoke to the lead
Fintech and apps: real deposits or funded accounts, tracked through a mobile measurement platform
Platform figures then sit underneath as guidance. They help decide which campaign to scale or which ad to pause. They do not decide whether the month was a success.
Measure What Each Channel Really Adds
The harder question is incrementality: what would happen to sales if a channel stopped?
When we start with one channel, we can see its real impact on the business. When we add a second, the change in total sales shows what the second one adds, which can be very different from what its dashboard claims.
For accounts already running several channels, there is a simple test. If the channels are fighting over who gets credit for the same revenue, keep one running and pause the others for a month or two. If total revenue does not change, those channels were adding little. If it drops, the drop gives you a rough measure of what they were really contributing. From there, geo tests, where ads run in some regions and not in others, give a sharper picture.
Lead Businesses Need Deeper Measurement
For lead generation, cost per lead is the most misleading number in the report. A flood of cheap leads should make an agency suspicious, especially for a premium product. An ultra-luxury property project should not suddenly generate huge demand at a low cost, and when it does, something in the campaign or its delivery has usually changed.
The right approach measures quality first. In real estate, that means following each lead until the sales agent has a clear answer, interested or not, and sending that answer back to the ad platform as an offline conversion so its algorithm learns what a real buyer looks like.
What a Good Report Shows
A report you can trust shows:
The north star metric against target, from your own data
What changed and why, in plain language
Platform metrics as supporting detail, clearly labelled as platform-reported
What happens next
The rule is simple: never take the numbers in the ad accounts as reality. Rely on your main data source, and expect your agency to do the same.


