Can you recover conversions lost to cookie consent?

All posts

Not as observed conversions. A visitor who rejects cookies is never tracked back to the ad they clicked. You can give Google what it needs to estimate some of them. That means Advanced Consent Mode, all four signals set before your tags load, pings that actually leave the browser, and an account above Google's modelling thresholds. Here is the order, and how to check each step.

What do you lose when consent is denied?

Rejecting cookies does not delete the visitor. They still land, browse and, often, buy. What disappears is the thread that ties that purchase back to the ad they clicked. Without cookies, the identifiers that stitch click to conversion never get written, so Google Ads sees the spend but not the sale, and your reports quietly stop crediting the campaigns that are working.

That is why teams watch conversions fall overnight without touching a single campaign. The size of the drop depends on how many of your visitors decline, which varies by region, audience and banner design, so read your consent tool's reject rate before you judge a campaign. It can look like a performance collapse when it is a measurement gap. Modelling can narrow that gap, but only if the setup underneath it is right.

How do you set it up, in order?

This is not one setting. It is a short chain, and each step depends on the one before it: skip step two and the later steps have nothing to work with. Do them in this order and check each one before moving on.

1. Turn on advanced consent mode

   Switch Consent Mode from basic to advanced. In basic mode, Google tags stay blocked until the visitor chooses, so a refusal sends nothing and Google can only apply a general model. In advanced mode, the tags load with every signal denied and send cookieless pings, which let Google model from your own site's data. Google's consent mode overview sets out both. Check with whoever advises you on privacy that pings before consent are acceptable in your markets.

2. Send all four signals before tags load

   Set ad_storage, analytics_storage, ad_user_data and ad_personalization to their default denied state in the page head, before any Google tag initialises, then update them the instant the visitor chooses. All four have to be present. A common failure is shipping only the two analytics signals and wondering why Ads never models anything.

3. Check the denied pings leave the browser

   This is the step most setups skip. Open the network panel, decline consent, and confirm the cookieless collect request actually fires with gcs=G100 and the four signals attached. If your banner initialises after the Google tag, or an extra consent check blocks the tag, the ping is never sent and the model has nothing to learn from. Nothing warns you when that happens.

4. Add enhanced conversions for the yes group

   Advanced mode deals with the visitors who said no. Enhanced conversions help with the ones who said yes. At the moment of conversion, the tag hashes a piece of first party data, such as an email address. The conversion can then still be matched when cookies are missing or the journey crosses devices. The tag only sends that data when ad_user_data is granted. Google's guide to enhanced conversions covers the setup.

5. Move reporting to the server

   A server container receives events from the browser on your own domain and forwards them to Google and Meta from there. Delivery can be more reliable, because a first party request is less exposed to ad blockers and to browser limits on third party cookies. It does not change what the visitor agreed to. The consent state has to travel with every event, as we explain in how to pass the consent signal to server side tagging.

6. Confirm modelling actually switched on

   Modelling only starts once you clear Google's thresholds, and below them nothing tells you so. In Google Ads, open the conversion action and read its consent mode status on the Diagnostics tab. Once modelling is active, it shows the uplift per domain and country for four weeks. In GA4, check the property against the behavioural modelling prerequisites. A setup that looks right but sits under the threshold models nothing.

What are the honest limits?

Three limits are worth stating plainly. First, modelling estimates the conversions Google cannot see from the ones it can. It is a statistical fill, not a replay of lost data, and nobody can tell you in advance how much of the gap it will cover. Second, it depends on volume: below 700 ad clicks over 7 days per country and domain grouping, Google Ads does not model at all. Third, none of this reaches backwards. Modelling only uses data collected while consent was set up correctly, so conversions from a broken or missing consent period stay unmeasured.

The model is also only as good as your banner. A banner that pushes people into accepting produces consent that is legally fragile and a signal that is noisy. Clear choices give you cleaner data.

Velo handles the consent side of this chain. It sets the Consent Mode v2 defaults as soon as its snippet runs, sends the update when the visitor chooses, and records each choice with a timestamp. Put the snippet above your tags so they read the defaults first. Enhanced conversions, server side tagging and Google's thresholds stay in your Google accounts.