Analytics & measurement

Measurement is the instrument, not the report

Most businesses treat analytics as an end-of-month report nobody reads. It should be the opposite: the instrument you steer with day to day. Good measurement answers the questions that decide where money goes — which channel brings customers and not just clicks, where in the funnel people drop, which change actually moved the needle. Without it, marketing is a series of bets; with it, a series of informed decisions you can defend with figures.

That's why this is foundational to everything else we do. You can't optimize what you don't measure, you can't prove what works without trustworthy data, and you can't feed AI bidding or experimentation good signals if the underlying numbers are wrong. Measurement isn't a deliverable that sits beside the marketing; it's the nervous system that makes the rest of it intelligent.

The trust gap most businesses don't see

Here's the uncomfortable truth: most analytics setups can't be trusted, and the people relying on them often know it. Only about 37% of businesses trust their analytics data enough to use it for major strategic decisions, and poor data quality drains an estimated 20–30% of marketing budgets through misallocated spend. Decisions get made on numbers that are incomplete or simply wrong, and the cost is invisible until you look for it.

Most "GA4 problems" are setup problems

When reports look wrong, the cause is usually the foundation, not the tool. In practice, most GA4 complaints trace back to tagging issues: duplicated page_view or purchase events, conversions never properly defined, broken UTM discipline, internal traffic polluting the data, or consent settings misconfigured. More than 60% of implementations carry problems like these, and a shaky setup makes every downstream report unreliable — including the ones AI bidding and experimentation depend on.

So we start with an audit and a clean rebuild: a single properly configured GA4 property, events that fire once and consistently across templates, conversions defined around your real business outcomes, disciplined UTMs, and internal-traffic filters that work. It's unglamorous engineering, but it's what turns analytics from a source of arguments into a source of answers. Firms that get this right and use GA4 for genuine insight report around a 34% improvement in marketing ROI — the upside of measurement you can actually believe.

Attribution: the last click isn't the whole story

One of the most expensive measurement mistakes is judging channels by last-click. A search ad, a social post or a piece of content often creates the demand that converts later through another channel — and last-click hands all the credit to whatever happened to be last, so teams cut the very channels building their pipeline. This matters even more now that AI search answers questions without a click, pushing value into the pre-click part of the journey where trust is formed.

The fix is attribution that reflects the full path. GA4's data-driven attribution is the default and uses machine learning to spread credit across touchpoints, though it needs enough conversion volume to work well and still has blind spots across devices and platforms. We configure it correctly, read the assisted-conversion and journey reports rather than last-click alone, and complement GA4 where it falls short — so budget decisions credit what truly contributes, not just what closed the deal.

Value metrics, not vanity

A dashboard full of pageviews, sessions and likes feels reassuring and tells you almost nothing about the business. In 2026 the line between vanity metrics and value metrics is a line between looking busy and growing. We build reporting around value: cost per lead, cost per acquisition, conversion rate, customer lifetime value and return — the numbers that map to revenue and tell you where to invest more and what to cut.

// Trustworthy data, honest attribution, value metrics
{
  "foundation": "one clean GA4: events, conversions, UTMs, filters",
  "consent": "consent mode v3, privacy-first by design",
  "attribution": "data-driven + full-journey, not last-click",
  "closed_loop": "GA4 <-> CRM/ads: first click to closed deal",
  "report": "value metrics: CPL, CAC, LTV, ROAS",
  "rule": "if it isn't measured, we don't bill it"
}

And we close the loop. By connecting GA4 to your CRM and ad platforms, we follow a customer from first click to closed deal — closed-loop reporting that shows which marketing produced sales, not just clicks, and feeds clean conversion data back to the platforms that need it.

Built for a privacy-first web

Measurement now has to work under privacy rules, not pretend they don't exist. Low consent opt-in can push up to 70% of visitors into modeled rather than observed data, browsers block much of what used to be tracked, and consent mode is non-negotiable in markets with consent requirements. Ignoring this doesn't keep your data clean — it just makes it wrong in ways you can't see.

So we design for it: correctly implemented consent mode, first-party data, and, where it's warranted, server-side tracking that can recover 20–40% of the signal lost to browser restrictions while centralizing privacy control. Done carelessly, server-side setups actually lose 20–30% of events, so this is engineering that has to be done right, not just turned on. For businesses serving multiple markets, that also means respecting each region's rules while keeping the measurement consistent enough to compare. The aim throughout is the same: data you can trust, on a web that's harder than ever to measure.

The reports that quietly mislead

Some of the most dangerous numbers are the ones that look fine. A traffic chart inflated by your own team visiting the site. A "conversions" figure counting the same purchase twice because an event fires on every page load. A channel that looks unprofitable on last-click but is actually creating half your demand. A spike that's really a tracking change, not a real shift in behavior. These aren't exotic edge cases — they're the everyday ways a dashboard tells a confident story that isn't true, and decisions made on them quietly waste budget.

Guarding against this is ongoing work, not a one-time setup. Modern analytics stacks drift: a new template breaks tagging, a consent change alters what's collected, a campaign launches with sloppy UTMs. So we treat measurement as something to monitor continuously rather than configure and forget — watching for duplicates, discrepancies and attribution blind spots before they corrupt a quarter of decisions. The goal isn't a prettier report; it's a number you can stake a budget on without a quiet doubt that it might be fiction.

How we work

We build measurement you can actually decide on, foundation first:

• Audit & rebuild: a clean GA4 — events, conversions, UTMs and filters that work.
• Consent & privacy: consent mode and first-party data, privacy-first by design.
• Attribution: data-driven, full-journey reading instead of last-click.
• Closed loop: GA4 connected to your CRM and ads, first click to closed deal.
• Dashboards: value metrics — CPL, CAC, LTV and ROAS — not vanity.
• Recovery: server-side and modeled data, done right, for a privacy-first web.

The result is a measurement system you trust enough to act on — the foundation that makes CRO honest, AI bidding effective and every channel accountable. It's the same principle that runs through everything here: if it isn't measured, we don't bill it.

Frequently asked questions