Cookie Free Marketing Measurement in 2025

Chrome quietly announced last month that they would drop their plans to require cookie opt-in, closing the door on a 5 year saga for marketers after launching their Privacy Sandbox initiative under regulatory pressure back in August 2019. But what is the landscape like in 2025 for cookies anyway?

Regulatory pressure remains – GDPR and PECR both require opt-in for use of cookies in almost every aspect of their use apart from their original intended use: persistence of essential website features, such as checkout carts.

The IAB’s TCF creates a legal obligation for privacy friendly Ad Measurement, and increasingly brands are requiring TCF’s privacy components, such as DNT (Do Not Track), and GPC (Global Privacy Control) to place media spend.

And of course, it isn’t just Europe which has stringent legal frameworks, increasingly more states in the US are following equivalents of California’s CCPA; and across South America, Africa, and Asia there are equivalent privacy-first legal frameworks to consider.

So, the legal obligation remains to remove automatic opt-in for cookies, eliminating the requirement for browsers to be the arbiter of choice. That said, more than 50% of the browser market in mature digital markets – like the UK & US – already have browser usage that blocks 3^rd party cookies by default.

In the UK, Chrome has just 49% marketshare, all other user-agents automatically block cookies currently.

While in the US, Chrome is only marginally stronger, with a 52% marketshare.

How is my marketing affected by cookies?

Today, legal cookie usage means that even 1^st party cookies – such as those used by Google Analytics 4 and Adobe Analytics – are routinely blocked by ‘Consent Mode’ functions, which require users to opt in to their usage. This means large chunks of measurable traffic are no longer available in UK & European markets. For most that figure will be between 25-35% of all event measurement, for some it will be up to 50% on average. In some regions like Germany, the rate can be greater still – though recent regulation will replace double opt in with a more standard single opt in requirement as of April this year.

This is clearly a critical problem for accurate understanding of marketing effectiveness.

I have long championed cookie-free measurement and ad activation since leading the development and launch of Corvidae AI to the market. There are sound data reasons to continue to follow that path.

First, measurement: losing up to a third of your measurement is for many businesses the same as giving up on measurement, and is in large part responsible for the return to digital marketing of onerous – and not actionable for performance marketers – MMM analysis to try to fill back in understanding on which channels are driving which sales.

This is a backwards step for performance marketers, who need to optimise on a campaign by campaign basis. Consider the value of having a fully complete and accurate breakdown of cross channel attributed insight for all your activity on any platform, as can be seen below, with an MMM which is directional to channel level at best, and even with modern AI advances still requires significant investment in infrastructure, data cleansing, and data caveats to run at all.

Unified reports which have a single attributed source of revenue that maps precisely back to your activity across any AdTech gives control back to the performance marketer to deliver efficient media allocation across individual campaigns and channels.

And moving away from cookie based measurement also allows deeper understanding of how your customers are interacting with your marketing activity in granular detail while remaining fully compliant with all of the legal aspects discussed above. For example, exploring how social campaigns mead to conversions in other channels with step by step breakdowns of aggregate paths gives complete insight into how Social plays in the mix regardless of which attribution model you have selected.

So how can I efficiently activate ads in a cookie-poor environment?

Measurement is only half the story of course – placing ads using the insight you now have is just as important. This is where following a targeting-light ad placement strategy offers the best value and comes with the added benefit of reduced costs in your MarTech stack.

Consider the costs associated with buying 3^rd party data for open web – which is principally harvested from cookies, but also increasingly from 2^nd Party ID vendors, such as the Trade Desk’s Unified ID 2.0; LiveRamp’s RampID, or ID5.

In principle the process of ID matching is the same across all ID vendors: matching your 1^st party data to their ID should allow you to serve an ad to a target audience you either select or allow a platform to build based on performance.

In practice, ad fraud is rampant in the open web, with early estimates from the ANA suggesting 25% of all ad impressions are on MFA (Made for Advertising) websites – a trend which has been slightly reversed in their latest study, though publisher verification remains mixed at best with The Trade Desk and Google not participating in this latter study. Anti-fraud solutions typically do not solve for the underlying issue that the majority of open web that is becoming available through these identifiers is junk traffic – either bots, or MFA sites which can be difficult to police when this is complemented by ad stuffing that is generally undetectable by these systems. Consider the stark failure to prevent even the most obvious fraud that is routinely discussed in the industry for its largest market analytics platform GA4.

And given the costs of attempting to use 2^nd or 3^rd party data to target your ad spending, the net results is that of every £1 spent on Display, only 36p actually reaches the publisher – meaning 64% of spend is lost to the cost of placement – too high a costs for effective uplift from the use of the 2^nd and 3^rd party data in the first place.

And of course, the lack of inter-operability with the major walled-garden players means your open web solution is likely to still incur the challenge of double serving your target audience via different mediums – Google, Meta, Amazon and Apple between them have artificially and actively created data siloing within advertising over the last 10 to 15 years that means cross channel measurement is almost impossible for brands or marketers.

However, consider the example I shared above about measurement accuracy with a tool like Corvidae. Here we can actually see inside the walled garden to see how much overcounting is going on in their own – again siloed – performance reporting:

In this report snapshot we can see overreporting of 3X for Meta campaigns, when unified to P&L via Corvidae AI Attribution.

If we have this level of accuracy in reporting we can manage our spend efficiently both within walled gardens, and we can then run media spend in the open web without 3^rd party targeting requirements using a whitelisting approach.

Whitelisting to serve open web efficiently and without cookies

Whitelisting is the practice of only opting in domains you would be happy to advertise on with no bid activation criteria – your ads would show to all visitors without prejudice to their cookie opt-in state or user-agent. No ads are shown on any other domains. Performance is measured using unified attribution, with high bounce rates indicative of poor ad serving practices and removal of the domain from the campaign. No other quality signals are required to manage high quality ad placement and brand safety as long as the selected domains are of a high quality for your market niche. Given that in the ANA report mentioned previously, only 70 domains were required to match the audience reach of over 44,000 used by non-whitelisting programmatic activity, that is a clear signal that we do not need to saturate domain lists for broad reach.

Indeed, with Log-level or Impression tracking in Corvidae, you can also track and view ad effectiveness regardless of domain – or even App Store – and stitch those touchpoints into your Journey Explorer reports to understand each and every influencing piece of the journey in an entirely cookie – and ID – free way.

In summary: 2025 still has little room for cookies or other ID solutions, instead embrace finding accuracy in measurement and unlock efficient optimisation to battle ricing CPAs and move away from legacy measurement and into a compliant, customer-first marketing world with modern AI solutions.