Goal

If you switched a channel off tomorrow: which revenue would stay away, and how do you know?

Attribution says who gets credit for a conversion. It does not say whether the conversion would have happened without the channel. That second question decides budgets, and it is rarely asked.

What we do: We define which number steers, wire consent cleanly to the tags and build tests with control groups; you know which revenue would stop if a channel were switched off tomorrow.

Comparison groups instead of attribution: only what exceeds the group without the channel is impact. The rest would have happened anyway.

Why this is hard right now

Google Ads today knows two attribution models: last click and data-driven, with data-driven attribution used by default for most conversion actions. The four older models first click, linear, time decay and position-based have been retired. Whichever model applies: a channel can look good in every one of them and still be doing nothing but collecting conversions that would have arrived anyway.

On top of that, systems count differently, and they are supposed to. GA4 measures event-based and assigns to a session and its source path; the ad account assigns to the click. As long as both are set up cleanly they will differ, and that is not an error. It becomes critical when budgets are cut on the basis of whichever number is smaller, with the convenient source picked in every meeting.

The third construction site is consent. Google steers measurement via ad_storage and analytics_storage plus, added in November 2023, ad_user_data and ad_personalization. Missing conversions almost never come from the parameter itself but from consent dialog and tags not speaking the same language. The loss does not just hit the report; it hits the automated bidding, which learns on less data.

The questions we work through with you

Which number decides in your company, and will it still hold tomorrow?

How you can tellAd account, analytics and business system report three different numbers, and depending on the meeting the convenient one is used. A second sign is a break in the time series on exactly the day an attribution model was switched: before and after can no longer be compared.

What we doWe agree with you which model carries the steering, and document every switch with its date so time series stay readable. Every source gets its purpose: one for steering the campaigns, one for business reporting. We write that down so the discussion is not reopened monthly.

How do you measure what a channel adds?

How you can tellThere is no test in which a channel was ever held out. So there is no statement about what it adds. The typical follow-on pattern: budget is shifted by last click, channels early in the journey lose, and then total revenue drops without the last-clicked channel having gotten any worse.

What we doWe separate attribution from impact. Attribution steers the day-to-day; for the question of incremental contribution we build tests with control groups, say regional tests or deliberately holding out audiences. Such tests need preparation and a runtime that matches the buying cycle, so we plan them into the year instead of improvising them.

Why do GA4 and your ad account disagree?

How you can tellThe report shows the same metric from two sources with different values, and nobody can say which one holds. A common side cause: enhanced measurement events run unchecked, so scrolling or outbound clicks count as success. The event count rises, revenue does not.

What we doWe quantify where the difference comes from, say attribution model or consent, instead of bending one of the two numbers into line. Before that stands a measurement plan: which events exist, which parameters they carry and which of them are key events. Where Google offers a recommended event with a defined name, we use it; custom names are the exception.

What happens to your measurement when consent rates drop?

How you can tellThe drop coincides with the day the dialog was changed, not with any shift in the market. Or the dialog reports consent and the tags still run in restricted mode. This can be checked with the Tag Manager preview on a test page.

What we doWe first establish which parameters are sent today and how high consent actually is. Then ad_storage, analytics_storage, ad_user_data and ad_personalization are wired consistently to the dialog. Whether basic or advanced consent mode is permissible is a call for you and your legal team; we build the result and do not judge it legally.

Do your conversions even arrive in the platforms?

How you can tellA noticeable share of deals is missing in the platforms although they exist in the shop or the sales system. In businesses where a human sits between click and deal, a shared identifier is often missing, so offline deals cannot be matched to a lead.

What we doWe compare deals in the business system with what the platforms report and quantify the gap before building anything. Enhanced conversions pass first-party customer data such as email address, name, address and phone number, hashed with SHA256; the leads variant joins offline deals with website leads. After the rebuild we verify against the business system that the number is right.

The expertise we combine for this

Measurement & AttributionSets the model that carries the steering, and builds the tests that answer what a channel truly adds.
Tracking & AnalyticsPuts measurement plan, events and parameters into an order in which two systems may count differently without causing arguments.
Data Protection & ComplianceWires the four consent parameters consistently to the dialog, so what may be measured is what gets measured.
SEA / Paid SearchUses the verified numbers for steering instead of optimizing on a metric that does not reflect the business.

And the technology to match

Your own data and signals prepared so they are comparable across channels. Only then can anyone argue about contributions without arguing about definitions. Outages and deviations are flagged automatically, so they surface the same day and not in the quarterly report.

How expertise and technology work together at hurra.com

Frequently asked questions

Which attribution model should I use?

Google Ads today still offers last click and data-driven attribution, with data-driven used by default for most conversion actions. The four older models have been retired. More important than the choice is staying with one model and documenting every switch with its date.

Why does Google Ads show more conversions than GA4?

Because the two assign differently. The ad account assigns to the click and uses data-driven attribution by default, GA4 assigns to a session. With a clean setup, deviations are normal and should be explained, not aligned away.

Which consent parameters does Google require?

ad_storage and analytics_storage plus, added in November 2023, ad_user_data and ad_personalization. Anyone using consent mode must send the two newer ones as well.

Do I lose conversions without consent mode?

Above all you lose signal for the automated bidding. Fewer measured conversions also means less basis for bids, and that weighs heavier than the missing line in the report.

Do I need server-side tagging?

Not necessarily. Google names better site performance, more detailed privacy controls and better data quality as reasons for the server container. The effort is not in setting it up but in running it: the container has costs, needs its own subdomain setup and must be monitored. For a small site with two tools it rarely pays off.

Does server-side tagging solve my consent problems?

No. Consent applies regardless of where a tag is executed. The consent parameters have to travel with the setup, otherwise you measure technically cleanly and still without a legal basis.