How Chilean B2B startups measure blind

We scanned 8 Chilean B2B startups. None of them were measuring their marketing well.

That's not what I expected to find. And it's probably not what you'd expect either, if you work at one: half of those 8 had plenty of tools — Google, Meta, LinkedIn, a CRM, sometimes more. The other half had almost nothing. They landed in the same place anyway: deciding based on a number that wasn't the real one.

There's no benchmark for this in Spanish, and not much of one in English either. There's plenty of content about B2B attribution, but it's written for a RevOps team with Salesforce, HubSpot and a data warehouse behind them. Here, the person deciding is one person, running 6 or 8 tools held together with duct tape, trying to figure out why the number on the screen doesn't match what's happening in the bank account — a version of what happens with attribution dashboards at large agencies, except there the problem is incentives, and here it's technical first.

More tools isn't the answer

The first assumption that falls apart is the obvious one: that the problem is a lack of instrumentation.

Two cases, opposite extremes. One didn't even have Google Analytics installed — nobody at the company could say how many people had visited the site that week. The other had six tools connected at once: Google, Meta, LinkedIn, a CRM, a heatmap tool, a chat tool. The most complete stack of the 8 we reviewed.

Both ended up with the same problem: a number that isn't useful for deciding anything. The first one is missing the data. The second one has too many versions of the same data, all disagreeing with each other.

Instrumentation isn't the axis. The shape the gap takes is. And across all 8 cases, that shape landed on exactly four patterns — two startups in each.

The four patterns that repeat

1. Not instrumented

Zero or nearly zero measurement tools installed on the site.

One case had absolutely nothing: no Analytics, no Tag Manager, not a single tracking snippet from any ad platform. The other had exactly one thing installed, Google Analytics, and nothing else — no Meta, no Google Ads, no LinkedIn.

In both, you can answer "how many people came in?" today. You can't answer where they came from — and that's the only question that decides where to put the next dollar.

2. Single point of failure

All measurement flows through a single Tag Manager container, with nothing set directly in the site's code.

One case ran every connected platform's tracking through that one container: if it gets published wrong on a given day, every channel goes dark at once. The site keeps working exactly the same. Nobody notices until someone asks why the numbers dropped.

The other operated across several countries from the same site, and all three shared the same container. Ad spend is split by country. Results aren't. There's no way to tell which of the three markets is actually paying for itself.

It's convenient to manage. And it's exactly the kind of failure that stays quiet — it doesn't announce itself, it doesn't show up anywhere, and by the time it's noticed, weeks have already passed.

3. Broken funnel

There's a form on the site. There's nothing catching what it collects.

One case sends every form submission straight to an inbox. No CRM, no record. The trail between "came from an ad" and "became a deal" gets cut right there, and nobody notices because the form itself works fine — the email arrives, someone replies, life goes on.

The other had an email marketing tool connected, which looks like it solves the problem. But that tool tracks who signed up, not who ended up paying. The number that gets checked most often — signups this week — is, without anyone deciding it that way, the one that least resembles revenue.

4. Double counting

Two or more platforms claim credit for the same conversion.

The case with the most built-out stack of the 8 is also the one suffering most from this: with six tools watching the same button, each one reports a different number for the same event. Five booked meetings can show up as eight, split across Meta, Google and LinkedIn, each platform claiming full credit.

The other lives it across channels: it measures conversions on the website with one tool and on the app with another, and each system uses its own rules to decide who gets credit for the sale. The same person can get counted twice without anything in the report giving it away.

Which one is yours

Four yes-or-no questions. Whichever one you say yes to first probably places you.

Do you have fewer than two measurement tools installed? → Not instrumented.

Does all your measurement flow through a single Tag Manager, with nothing set directly in your code? → Single point of failure.

Does your form land in an inbox, or does your email tool only see signups, not payments? → Broken funnel.

Does more than one platform report the same meeting or the same sale? → Double counting.

If you said yes to more than one, you're not alone: half of the 8 cases we scanned also qualified for more than one pattern. We logged whichever one weighed most on the decision it was distorting.

What to do with this

There's no single checklist fix, because each pattern calls for something different. If you're not instrumented, install. If you have a single point of failure, pull your main tracking out of the container and set it directly in the code. If your funnel is broken, get a CRM — even the simplest one that exists. If you have double counting, pick one source of truth and let the rest inform, not decide.

The first step, before any of those four, is knowing which one you're in. That's exactly what the scan calculates: not another tool to add to the pile, but a diagnosis of which pattern your measurement has today and what that means for the decisions you're making with it — the same principle behind our guide to running your own ad campaigns without an agency, applied here to the technical side of the problem.

These are 8 scans from an outreach campaign, not a national study. This doesn't extrapolate to "all Chilean startups," let alone SaaS in general. But the pattern repeated often enough to earn a name, and if you work at a B2B startup, there's a reasonable chance one of the four is yours.

Frequently asked questions

How do I know if my startup is measuring its marketing well?
Check four things: how many measurement tools you have installed, whether everything flows through a single Tag Manager with nothing set directly in your code, whether your form reaches a CRM or just an inbox, and whether more than one platform reports the same conversion. If any of those hit home, that's where the problem starts.

What is a "single point of failure" in marketing measurement?
It's when all your measurement depends on one piece — usually a Tag Manager container — with no tracking set directly on the site itself. If that piece fails or gets misconfigured, you lose every channel at once, and the site keeps working exactly the same: nothing alerts you.

Why do Meta and Google Ads show different results for the same sale?
Because each platform only sees what happens inside its own ecosystem and uses its own rules to decide which ad gets credit for a conversion. If both were active before a sale, both can claim it in full. Adding up what each one reports gives you more sales than actually happened.

Do I need a CRM if I'm just starting out?
You don't need the most expensive or complete one. You need something — even a well-kept spreadsheet — that connects each form submission to what happened next: did it become a meeting, a deal, or nothing. Without that, the number you check daily (signups, leads) is the one that least resembles revenue.

Is more measurement tooling always better?
No. The case with the most complete stack of the 8 we scanned was also the one suffering most from double counting: six tools watching the same button, each claiming credit for the same conversion. More tools without a clear source of truth just add versions of the same number, not clarity.

If you want to know which of the four patterns your own measurement falls into, scan your site free: I'll tell you in plain terms what your marketing is measuring today, what's leaking along the way, and how far you are from a startup at your same stage. No agency and no airplane-cockpit dashboards.