Measure Camp Brno 2025 – Valuable Takeaways from the Conference

Topics: ,

An (un)conference where the programme is born on the spot. Paper cards, a time schedule, 25-minute slots, a mix of Czech and English and hundreds of people who voluntarily spend a Saturday on analytics. This year the eleventh edition took place at the premises of Gen Digital (formerly Avast) in the wider centre of Brno. Networking from breakfast to the afterparty, five rooms and lots of practical sessions about GA4, GTM, server-side tracking, AI and measurement without cookies. Of course there were also “therapy sessions” – meetings not directly about techniques and implementations but about interesting topics on the line between psychology and analytics.

Why do we like it? Because practice meets practice here. No marketing varnish, but specific procedures, demonstrations and discussion of real problems.

The conference was attended by Martin Štacko and Marek Janošec, each of us choosing different topics so that we would cover at least part of the talks from the board. We prepared a short summary of some of the talks:

Google Tag Gateway (GTG): A Smarter “Middleman” Between the Website and Google Servers

Speaker: Karolina Wrzask

GTG is a new “intermediate step” between your website and Google services (GA4, Google Ads, etc.). Instead of the classic “web → GTM → Google” it adds Google Tag Gateway, ideally on your own domain (CDN). The speaker Karolina went on to show the benefits of GTG and its implementation.

The talk mentioned that thanks to GTG signal delivery improves by up to +14%. Because of the implementation on a first-party domain there are fewer blocks than with 3rd-party, and in addition modelling works better with more valid signals, which means more accurate data in Google tools. Another plus is security, thanks to greater control over the data.

Types of implementation:

  • Via a CMS – Google is working on a solution for popular CMSs.
  • Fastest via a CDN (Cloudflare) – “one-click” in GTM (note: it works for 1 GTM container).
  • Manually via another CDN – according to the developer guide.
  • Via sGTM – that works too, but watch out for costs (more requests = more money in the cloud).

Bugs in Tracking: Errors Will Come – Be the First to See Them

Speaker: Pavel Šabatka

Why deal with it: Automated campaigns (PMax and the like) stand on quality signals. When measurement breaks, the algorithms lose their compass and you burn budget. Errors can occur on the side of the website itself, in GTM, or they can be external influences such as spam, policy violations on the given platforms, or legislation (GDPR). On the website side they can be code changes, JS in the wrong place or conflicts of it, changes in the data layer or the Content Security Policy.

How to catch errors in time:

  • Error listeners in GTM (JS Error Listener trigger) + subsequently sending the error information to GA4
  • Try/catch in custom scripts + logging errors as events in the data layer.
  • Tag Monitor / addEventCallback and Additional Tag Metadata for a better audit.
  • Custom insights in GA4 and alerts in BigQuery.
  • Test templates and use checksum validators.

Practical impression: A client will appreciate it when you know about an error before they do, and you have a log of it and a clear remedy procedure.

Server-Side “Tricks” That Save Money and Nerves

Speaker: Zuzana Bečková Mikyšková

During her talk Zuzana described several tips and tricks on what to think about when implementing server-side measurement, what to avoid, or how to improve the efficiency and functionality of the measurement itself. We took away several interesting points from the talk:

  • Do not give yourself away with the name: avoid words like sgtm, analytics, tracking in the URL/host – ad blockers already know them and block the requests. It is better to name it with random words unrelated to measurement.
  • Check what you are really sending: not everything has to go through GA4; part of the events you can send straight to the target platform (less noise in GA4, fewer requests).
  • The difference between Cloud Run and Networking in the basic server setup because of its effect on price; budget alerting was also mentioned.
  • Watching event_id: key for deduplication and correct matching of conversions.
  • Stape Data Tag: handy for inspection, for precisely determining the data for Facebook CAPI
  • How to do Anonymous measurement for GA4

GTM Variable Audit (Mája Remešová): Get Things in Order and Speed Up the Website

Speaker: Mája Remešová

Large containers often suffer from duplicates and variable dependencies. Mája showed an approach to mapping and cleaning variables and speeding up GTM loading. A common result? Fewer variables, a shorter load time, easier debugging. During the talk the listeners learned useful information on these points:

  • Why mapping and organising variables is important for managing GTM
  • How duplication can affect website speed or effective management of GTM
  • Real data from “cleaning” containers and the resulting benefits

A question for everyone: When did you last do an audit of tags, triggers and variables in your GTM?

Legal Measurement Without Consent (for Analytics and Media)

Speaker: Vašek Jelen

In his talk Vašek provided an introduction to how Consent Mode works, and there was also a short discussion on data modelling. In practice, companies today often combine consented measurement for full data with anonymous/cookieless measurement for a basic overview (where there is no consent), which is important for reporting. It was also explained how to work with consent in Facebook Ads and the risks of Advanced Consent Mode, for example in the German market.

What we liked:

  • GTM Data Tag (Stape.io) for precise control over what and when we send for anonymous measurement
  • BQ streaming via sGTM (custom template) – real-time data into BigQuery.
  • Identifying the “session” from the referrer + sources via gclid/fbclid/utm – even without cookies you get a decent basis for reporting traffic by source.

A note on law: for IP anonymisation, consider at least truncation (ideally complete removal); it depends on the context of the implementation and the goal of the measurement. Some do not use the IP at all, some remove (anonymise) its last part.

What We Are Taking Away from Brno

Start with signal stability: without reliable measurement PMax and the like make no sense.

Consider Google Tag Gateway: ideally via Cloudflare CDN – quick implementation, fewer blocks, better models.

Log errors into GA4/BQ: so you know before the campaigns do that something has gone wrong.

Server-side with sense: do not hide behind flashy names and do not send everything everywhere.

Audit GTM variables: order = a faster website and easier management.

Think about law and security: anonymisation, hashing, governance – choose the solution according to the goal, not according to “everything that is possible”.

MeasureCamp in Brno once again confirmed that analytics is a team sport. Whoever can connect business goals, technical execution and the legal framework wins. We are already testing some of the novelties with clients – from GTG pilots through better error logging to cleaning up variables in GTM. If anything caught your interest and you want to deploy it yourself, let us know – we will gladly prepare a quick audit and a proposed procedure for you.