Skip to content
targetLAB

Web analytics and conversion tracking

The reports disagree because they are counting different things

targetLAB defines business events, implements straightforward GA4 and tag setups, tests consent behaviour, and specifies and verifies complex integrations.

Real enquiries can exist while Google Ads, GA4 and the sales record show different totals. Some gaps come from broken tags or duplicate events. Others come from consent, browsers and attribution rules built into the platforms.

The work starts by naming the event that matters: a qualified form submission, a measurable call, a booking, an order or a meaningful step towards one. The site, consent platform, tags and advertising accounts are then checked against that definition.

targetLAB implements straightforward analytics setups and small site changes directly. Complex CMS work, application changes and non-standard integrations are specified for the client’s developer and verified after implementation. This is technical implementation and testing, not legal advice or preparation of privacy and cookie documents.

GA4 · Google Tag Manager · Search Console · Looker Studio

Measurement changes

Consent and attribution now shape the dataset before the report opens

A clean tag is only one part of measurement. Eligibility thresholds, Czech consent rules and different platform clocks all change what appears in the interface.

01

Google added consent signals for EEA advertising

Since March 2024, Google has required two additional signals for EEA traffic: ad_user_data and ad_personalization. Without them, linked advertising products restrict audience use, remarketing and parts of measurement.

Google Ads Help and Analytics Help · requirement in force since March 2024; checked 9 August 2026

02

Conversion modelling has an entry threshold

Google states that consent-mode modelling requires at least 700 advertising clicks over 7 days for each country and domain grouping. Reaching the threshold makes the account eligible; it does not guarantee that modelling will activate or reconstruct every missing conversion.

Google Ads Help, About consent mode modelling · checked 9 August 2026

03

Czech consent law and regulator guidance are separate layers

Czech law has required prior opt-in for non-essential cookies since 1 January 2022. Separately, ÚOOÚ recommends a reject option in the first banner layer, considers 12 months a reasonable consent period and advises against asking again within 6 months after refusal. The dates come from law; the banner periods are regulator guidance.

ÚOOÚ Cookies Q&A · checked 9 August 2026; ÚOOÚ press release · 2 August 2023

04

Google Ads and GA4 use different clocks

Google Ads uses a 30-day click window by default, while GA4 key events look back 90 days by default. Ads reports the conversion against the ad-click date; GA4 reports it on the conversion date. Depending on the configured Ads window, a conversion can be added to an earlier click retrospectively for up to 90 days.

Google Ads Help and Analytics Help · checked 9 August 2026

These rules explain why reports diverge. They do not create a single universally correct total, restore all unobserved activity or provide a legal conclusion on a consent setup.

What the work covers

Define the event, test the path and document the limits

The useful outcome is a measurement system whose events, consent behaviour and known gaps can be inspected rather than guessed.

  1. 01

    Write the measurement plan

    The plan names the business events to observe: qualified forms, measurable calls, bookings, purchases and useful intermediate steps. It records where each event occurs, which system receives it and how duplicates are prevented.

  2. 02

    Audit the current setup

    GA4, Google Tag Manager, advertising tags, the consent platform, Search Console, forms and existing reports are checked for missing events, duplicates, wrong triggers, conflicting definitions and unexplained gaps.

  3. 03

    Implement straightforward tracking

    targetLAB configures ordinary GA4 and tag-manager events, advertising conversions and measurable form or call actions. Small site changes needed for reliable tracking are included in the work.

  4. 04

    Connect consent technically

    The consent interface and tags are tested so that the chosen state reaches the relevant platforms and triggers the agreed technical behaviour. The work verifies implementation; legal basis, wording and document drafting stay with the client’s legal adviser.

  5. 05

    Run quality assurance

    Events are tested across relevant devices and routes, duplicates are removed, data retention and attribution settings are reviewed, and known browser or platform limits are written down rather than hidden in a footnote.

  6. 06

    Reconcile reports

    A dashboard and metric dictionary show which system answers which question. Differences between advertising platforms, analytics and business records are traced to definitions, dates, attribution and implementation wherever the data allows.

  7. 07

    Diagnose the conversion path

    Forms, order steps, devices, speed and exit points are examined for practical friction. Models may flag anomalies in anonymised aggregate data; targetLAB decides which hypotheses merit a change. Larger development work is specified and verified after implementation.

What remains in the client’s hands

  • Measurement plan
  • Event and metric dictionary
  • Tracking issue register
  • Configured events and advertising conversions
  • Description of the technical consent behaviour
  • Quality-assurance protocol and test record
  • Known data and attribution limitations
  • Dashboard with an explanation of report differences
  • Conversion-path backlog and development specifications

The documentation states what can be counted, how it is counted and where the limits sit. Email addresses, phone numbers and form contents are not sent to AI tools. Legal review and privacy or cookie documents are outside this technical service.

Analytics questions

Why the totals differ and what can be fixed

Why do Google Ads and GA4 show different conversion totals?

They can use different attribution rules, reporting dates, identity signals, consent states and definitions. Ads is built to explain advertising interactions; GA4 describes behaviour across the site. The review first removes implementation errors, then documents the differences that are normal platform behaviour.

Will Consent Mode restore every missing conversion?

No. It communicates consent state and can support modelling where Google’s eligibility and data requirements are met. It does not observe a user who has not consented in the same way as a consenting user, and it does not turn an estimate into a complete sales record.

Does targetLAB check the cookie banner?

targetLAB checks the technical behaviour: what loads before a choice, which consent state reaches the tags and what changes after acceptance or refusal. The legal basis, banner wording, interface assessment and privacy documents require legal advice.

Can calls be measured in Czechia?

Depending on the platform and setup, measurement can record a click on a phone number or call metadata such as the call time, duration and attributed advertising interaction. The service does not include call recording or legal assessment of recording and consent requirements. Platform and number availability are checked for the specific setup.

Who implements the changes on the site?

targetLAB implements straightforward tag, event and small site changes with the required access. Complex CMS work, application changes and non-standard integrations go to the client’s developer with a written specification, followed by verification against the measurement plan.

When do server-side tagging, Google tag gateway or Meta CAPI make sense?

They are considered where the platform mix, data flow, consent design, infrastructure and value of additional control justify the extra implementation and maintenance. They are not automatic upgrades, and they do not bypass consent or browser rules.

Discuss the number that nobody can reconcile

Share the current reports, tag setup, consent platform and the business event that matters. The conversation can start with missing conversions, duplicate events, unexplained differences or a form path that loses people before contact.

These settings are available only while analytics is configured. You can allow or decline analytics, then change that choice later through the footer.

Necessary

Always on in these settings and cannot be switched off. This cookie stores only the policy version, analytics choice and decision time, with no random visitor ID.

On