Numbers with narrow meanings
Advertising produces a great deal of data. Most platforms can report dozens of metrics for every campaign, and dashboards make them look equally important. They are not. Each metric measures something specific and narrow, and most of the confusion in campaign reporting comes from asking a metric to answer a question it was never designed for.
Understanding what each common metric actually records — and what it does not — makes reports easier to read and decisions easier to make.
Impressions and reach: was it delivered?
An impression is recorded when an ad is served. Depending on the platform and format, that may or may not mean anyone actually saw it. An ad served at the bottom of a page that nobody scrolled to still counts.
Viewability metrics try to narrow this by recording whether a portion of the ad was on screen for a minimum time. They are a useful quality signal, though still not proof of attention.
Reach estimates how many different people saw the ad at least once, and frequency how many times on average. Reach and frequency together tell you whether the campaign is touching enough of the intended audience often enough — or too often. Note that reach figures are usually estimates, particularly across devices.
What these metrics tell you: whether the campaign was delivered at the scale planned. What they do not tell you: whether anyone cared.
Click-through rate: did it prompt an immediate action?
Click-through rate (CTR) is clicks divided by impressions. It indicates how often people who were served the ad chose to click on it there and then.
CTR is genuinely useful for comparing ad variations within the same channel and placement. If two search ads run against the same keywords, the one with the higher CTR is probably more relevant to searchers.
It becomes misleading when used to compare different channels or to judge campaigns whose job is not immediate response. Display and video ads often influence people who never click. Search ads naturally attract higher CTRs because people are actively looking. Comparing them directly says more about the channels than about the campaigns.
Accidental clicks inflate CTR in some mobile placements. A high CTR paired with very short visits to the landing page is often a sign that clicks were not intentional.
Conversions: did something meaningful happen?
A conversion is whatever action you have told the platform to count: a form submission, a purchase, a sign-up, a call. This is where metrics start to connect with business value — but only if the right actions are being counted correctly.
Three questions are worth asking about any conversion figure. First, what exactly is being counted? A page view of a thank-you page is not the same as a qualified lead. Second, is each action counted once? Duplicated tracking is common and inflates results. Third, does the platform's figure match what the business actually received? Comparing platform conversions with your own records is one of the most revealing checks available.
Cost per acquisition: was it efficient?
Cost per acquisition (CPA) divides spend by conversions. It is an intuitive efficiency measure and helpful for comparing campaigns with the same objective.
CPA hides quality, however. A campaign producing inexpensive leads that never become customers can look better than one producing fewer, more valuable leads. Wherever possible, CPA should be read alongside some measure of outcome quality — sales, qualified opportunities or revenue — even if that information arrives later and has to be matched manually.
The same caution applies to return on ad spend (ROAS). It is only as reliable as the revenue and attribution data behind it.
Attribution: who gets the credit?
Most people encounter several ads and channels before acting. Attribution is the method used to assign credit for a conversion across those touchpoints. Every method — last click, first click, position-based, data-driven — is a model built on assumptions, not a record of fact.
Each platform also tends to measure its own contribution generously, often using different time windows and counting rules. Adding up conversions claimed by every platform will frequently produce a total larger than the conversions that actually occurred.
A sensible approach is to compare more than one view: platform reporting, independent web analytics and the business's own records. Where they broadly agree, confidence is higher. Where they diverge, the difference is worth investigating rather than ignoring.
What metrics cannot easily show
Some of advertising's most important effects are the hardest to measure directly. Building familiarity, shaping how a brand is perceived and influencing decisions made weeks later in conversation all leave faint traces in platform data.
Privacy protections, consent choices and changes to browsers and devices also mean that a growing share of activity cannot be tracked at the individual level. This is appropriate, and plans should accept it rather than try to work around it.
Methods such as controlled tests, geographic comparisons and survey-based studies can help estimate effects that tracking cannot see. They require planning in advance and usually a reasonable budget, but they answer questions that dashboards alone cannot.
Reading a report sensibly
A few habits make campaign reports far more useful:
- Start with the objective, then look only at the metrics that relate to it.
- Look at trends over time rather than single-day snapshots.
- Be cautious with small numbers; a handful of conversions can swing wildly.
- Compare platform-reported results with independent data where possible.
- Write down what you plan to do differently as a result. If the answer is "nothing", ask whether the report is measuring the right things.
Metrics are tools for making decisions, not scores to be admired. Read in context, with their limits understood, they can tell you a great deal. Read in isolation, they can tell you almost anything you want to hear.