Marketing Is Not Numerology
A practical distinction for organizing dashboards, reports and meetings: which metrics should open the conversation, and which ones are useful for diagnosis afterward.
A distinction I took longer than I should have to learn.
There was a moment, around mid-2022, when I was presenting results to the sales team at Efecto Educativo. I opened the report, pointed to the CTR of the latest email campaign, said something like “we are up almost four points from last month,” and everyone nodded happily. It was a good meeting. The problem is that the campaign, with its spotless CTR, had generated exactly 14 leads. Fourteen. And four of them were already customers.
That is when I understood something I should have understood much earlier: some metrics look good in a report, and some metrics force you to make decisions. They are not the same.
Over time, I built a way of organizing data from the general to the specific, platform by platform. This is not a list of “important metrics,” because there are millions of those. It is more a criterion: which number would you start the meeting with?
Google Ads: I would start with cost per conversion
When you start with Google Ads, the temptation is to look at CTR first. It is the most visible metric, the one that rises or falls most dramatically, and the platform itself puts it in front of you.
But there is a well-documented problem with that: you can have a campaign with decent CTR and still bring expensive or low-quality leads. There are situations where ads with lower CTR have better conversion rates. In other words, CTR tells you whether the ad gets attention. It does not tell you whether the campaign works.
At UPAGO, I had a product awareness campaign that reached a 6% search CTR, which looked excellent on paper. But when we crossed that number with real conversions, submitted forms and demo requests, the cost per lead was twice what made sense. The ad attracted clicks from curious people, not buying people. That told us something about the ad, but above all it told us something about the keyword and the search intent behind it.
The first thing I would look at in Google Ads is cost per conversion (CPA). That tells me whether the result is sustainable. The second metric would be conversion rate, because it separates two different problems: if I have many clicks and few conversions, the problem is probably after the click, in the landing page, form or promise. If I have few clicks but good conversion, the problem is volume or keywords.
CTR is still useful, but as diagnosis, not as the protagonist. First I look at conversions. Only then do I look at CTR, CPC, Quality Score and search terms to understand why.
Meta Ads: cost per result and result quality
Meta has a classic trap: it shows many metrics that look good. Reach, impressions, views, interactions, clicks, CTR. The problem is that none of those tells you whether something commercially relevant happened.
I remember a lead campaign we ran at Efecto Educativo for a training program. The first week looked “good”: low cost per click, many impressions, CTR above average. But when Sales started calling those leads, the story changed. Many were students with no budget, or people who had filled out the form without reading it carefully. Cost per lead was low, but cost per qualified lead was a different movie.
The industry has documented this quite a bit: Meta promotes Lead Ads as a lower-friction option, but that often comes with a significant drop in lead quality. A cheap CPL can be a trap if those leads do not move through the sales process.
The first thing I would look at in Meta is cost per result, understanding that “result” depends on the campaign objective: a submitted form, an initiated conversation, a purchase, whatever matters. The second thing is the quality of that result: for ecommerce, ROAS; for services, how many of those leads are actually reachable, qualified or moving through the pipeline.
CTR, CPM, frequency and reach are useful for diagnosis. Is the creative tired? Is the audience too narrow? But I would not let them run the meeting.
GA4: key events, not traffic
Analytics without well-configured key events is basically an elegant visit counter. I say this with some guilt, because I spent a while looking at users and sessions as if that told me anything useful.
GA4 moved toward events because GA4 understands that what matters is user behavior, not just arrival. That makes sense when you think about it: a website can grow in sessions month after month without generating any commercial progress. Traffic without intent is noise.
What I would configure as key events depends on the business, but generally they are actions that show real intent: form submission, WhatsApp click, email click, download, booking, purchase, demo request. Those are the metrics the business cares about.
The second metric I would take is key event rate segmented by channel, because that is where the truth appears. You may discover that Instagram brings a lot of traffic but converts little, or that organic Google brings fewer visits but much stronger intent. That helps answer a concrete question: which channel brings traffic that actually does something?
Sessions, users, bounce rate and similar metrics move to a second layer. They do not disappear, but they should not be first on the dashboard.
Search Console: clicks by query and unused opportunities
Search Console has four simple metrics: clicks, impressions, CTR and average position. The mistake is looking at them as totals. Global average position can be very misleading: you might rank 4 on average, but be first for your most valuable keywords and 40th for the rest, and the number flattens everything until it says nothing actionable.
The first thing I would look at is organic clicks by page and query. Not “how many total clicks,” but which specific pages are generating real traffic, which searches are bringing it and whether that traffic comes from branded searches or new-intent searches.
The second thing, which to me is one of the most actionable metrics available, is identifying pages with high impressions and low CTR. That is a concrete opportunity: Google is already showing you, but people are not entering. There may be several reasons: the title is not attractive, the meta description does not resolve intent well, competition is strong in that SERP, or the content promise is unclear. But the point is that there is concrete editorial work to do, not just “keep publishing and wait.”
The underlying criterion
If I had to summarize it in one idea, it would be this: some metrics inform and some metrics decide.
The deciding metrics are few. In Google Ads, how much it costs me to convert. In Meta Ads, how much the right result costs and what quality that result has. In GA4, which channels generate actions that matter to the business. In Search Console, which searches bring real traffic and where visibility has not yet become clicks.
The other metrics are not useless. CTR, CPC, CPM, impressions, average position, engagement all have a place. But that place is diagnosis, not headline. They help explain why something works or does not work, once you already know which result you are evaluating.
Changing that order, putting decision metrics first and diagnostic metrics second, is a small-looking difference that completely changes the conversation with Sales, the client or the team.
Humberto Newman is a digital marketing specialist focused on growth and conversion. He has managed funnels for databases of more than 30,000 contacts and built performance dashboards in education and fintech companies.