Conversion Tracking in a World Without Cookies

Balancing Measurement Needs While Respecting User Privacy

Measurement is a fundamental component of digital marketing used to determine ROI across media channels, gain insights on creative messaging and testing, understand which targeting is driving results, inform optimizations, and more.

Advertisers have become reliant on 3rd-party cookies, used to collect data and track behavior across the web, as a way to measure the whole marketing funnel from awareness through conversion. 

However, with the upcoming iOS updates and industry changes in how user data is collected and stored (to align with new user preferences and regulation) --  there's a shift away from tracking behavior of the individual user.

Although Google recently announced that it is delaying the removal of third party cookies on Chrome for about two years, advertisers should start planning now as the ability to leverage cookies and device identifiers will continue to decrease over time.

In addition to impacting the ability to directly measure conversions and ROI, the shift away from cookies will impact the ability to remarket, monitor and manage ad fraud, and set frequency caps. 

While cookieless measurement solutions will vary by platform, some of the more common themes across all tools are: 

  • Modeling conversion data to fill in gaps in the user journey
  • Aggregating user data to protect user privacy
  • Anonymizing user identities through IDs or hashed email addresses

We are also going to see a move towards server-side data collection and sharing to better protect individual user data and improve ability for users to manage consent.

Reporting is going to look different as platforms move towards new tracking methods; advertisers need to be prepared to adapt to the new environment by educating themselves and stakeholders, implementing solutions that align with business goals, and establishing new reporting benchmarks. 

In this post, we’ll be sharing tracking solutions from Facebook, Google Ads, Programmatic, and more designed to get advertisers the data they need to make informed decisions while remaining privacy-focused. We’ll also be sharing next steps advertisers can take to ensure they have a robust, future-proof, and privacy-centric measurement strategy in place. 

💡 Click here to skip to a checklist of all measurement considerations.


Changes to Platform-Specific Tracking


Server-Side Tracking Using Conversion API (CAPI)

Facebook has developed CAPI to allow advertisers to share data directly from their server, rather than through a browser, to bypass the loss of cookie tracking allowed in the browser. 

Conversion API data is fed directly into Facebook and handled similarly to pixel events, allowing advertisers to monitor conversions and make adjustments in Ads Manager. Advertisers can transfer a broader range of data than Facebook Pixel actually captures, like CRM data or lower funnel events.

Conversion API is great for industries with high-security requirements since there will be more control over what data is shared and where it is shared. 

Action for Advertisers:

  1. Become an early adopter of Conversions API to better understand and prepare for the evolving landscape.
  2. Review what can be sent with CAPI, and strategize which data points and sources (e.g. CRM) should be included and start preparing for integration.
  3. Run tests early to compare performance against historical data to understand measurement impacts and to inform recommendations on how to evolve strategy.
  4. Continue to utilize the Facebook Pixel as a compliment to Conversions API for full-funnel visibility and to understand the difference in data.


Conversion Modeling Using Machine Learning

Moving forward, Google Ads will be unable to directly measure conversions when website cookies with ad clicks are no longer available. Examples of when cookies might not be available are cross-device conversions, conversions that occur on browsers that are no longer storing cookies, or if a user clears their cookies. 

In these cases, Google Ads will use machine learning and historical data to “scale” the number of conversions and amount of conversion revenue that cannot be directly measured to close these measurement gaps and provide advertisers with a more comprehensive view of performance.

  • Without modeled conversions, advertisers may attribute conversions that should be credited to Google Ads to the wrong channels in instances where cookies are not available. There will also be holes in data about the customer journey path, limiting understanding of touchpoints that move consumers towards the point of conversion.
  • With modeling solutions, available tracking data informs algorithms so they can make use of historical trends to validate and educate measurement.

As of April 2021, advertisers using Consent Mode will now see modeled conversions in their Google Ads reports for all campaign types to account for gaps in tracking when the user has not permitted consent.

According to Google:

Modeling Helps Recover More Than 70% of Ad-Click-to-Conversion Journeys Lost Due to User Consent Choices.

Image Source: Google Ads

Action for Advertisers:

  1. Ensure you have a strong tagging infrastructure in place, to be able to model off as much high-quality data as possible.
  2. Advertisers should ensure they have implemented a first-party tag, like gTag.js or Google Tag Manager (GTM) and a Conversion linker tag on their site to continue to collect all observable data and build a foundation for modeling.
    • If advertisers are NOT currently using gtag for Google Ads conversion tracking or GTM (comprehensive tool to manage conversion tags across publishers), then the next step is to move towards these tagging solutions for their privacy-preserving yet sustainable features.



Server-to-Server Tracking Using Unique IDs & Integrations

For Programmatic providers, one of the biggest risks with moving away from cookies is the loss of visibility into view-through conversions across the different browsers to measure the impact of brand awareness efforts on the user’s end decision. 

Programmatic companies are developing alternative measurement solutions which vary by platform. As will be true for all channels, even with these solutions, we anticipate that the volume of available data will decrease over time for publishers and that new performance benchmarks will need to be established. 

One proposed solution is server-to-server conversion tracking which allows data to be passed to secure servers, without relying on the user’s browser. Unique IDs will allow advertisers to bypass the need for cookies, by storing a unique ID server-side when a user views or clicks on an ad unit. Once that same user converts at a later time, the same unique ID is mapped back to the original interaction. 

Companies like Taboola have already created server-to-server integrations within their platform to map unique IDs back to specific campaigns. Once advertisers set up the related tracking UTM parameters onto their URLs to capture these IDs, then the click  ID can be mapped to a campaign set up. 

Other publishers, such as Trade Desk, are pursuing their own tracking workarounds and solutions such as Unified ID 2.0 (UID2.). UID2.0 is an open-source framework that, in lieu of cookies, will leverage user’s anonymized email addresses. This is done via gaining consent by logging into a single specific website or app.

The Trade Desk also suggests that this approach is not only more privacy-first but that it’s also more operable between devices and platforms without increasing effort or frustration from the user. 

Action for Advertisers:

  1. Start cleaning, building and segmenting your first-party data: Invest in a CRM (Customer Relationship Management) platform that will be able to combine all your data into one place.
    • From here, you’ll be able to segment your CRM in a way that you can leverage your own first-party data to inform marketing strategies.
    • This also allows marketers to map the full customer lifecycle. You’ll also have the ability to leverage similar audience targeting off your first-party segments to continue to leverage prospecting tactics, without 3rd party audiences. 
  2. Improve Analytics Capabilities: Invest in your data cleanliness and tracking!
    • By having improved data tracking (re: server-to-server tracking), you’ll be able to understand the value your marketing efforts have on your bottom line. 
    • Reports in your Analytics platform e.g. Path to Conversion, Assisted Conversions, etc. will be invaluable to measuring full-funnel marketing.



Developing a Robust, Future-Proof Measurement Strategy

Now is the best time to analyze your current setup to ensure it is comprehensive, build a measurement infrastructure with data sources and tagging that work for your company, and implement these solutions with developer and/or partner resources. 

Below are the pillars of focus for a robust measurement strategy: 

1) Select the right data collection source(s)

Google Analytics 4

If you are currently on Google Analytics, consider moving towards GA4 (the sooner the better) which is built with machine learning and privacy at its core to prepare partners as there are more restrictions on cookies and identifiers.

Google is developing advanced modeling capabilities within GA4 to give a complete view of the user journey while building in additional data controls for both users and advertisers. The new analytics platform is designed to continue to adapt in this new environment with a more flexible approach to measurement and modeling capabilities. 

New call-to-action


Action Items for Advertisers:

  1. Leverage Google Signals: You can now use Google Signals as a reporting identity just as you would User or Device ID to identify users who have logged into Google and opted into Ad Personalization. This, combined with other GA4 features such as machine learning, enables you to get a more accurate picture of a subset of user’s behavior across all your platforms, even if they aren’t currently logged into Google. 
  2. Use Machine Learning to Fill the Gaps: GA4 comes with powerful machine learning that allows you to respect user privacy without losing data. Google’s Machine Learning can analyze the data of your subset of defined users we discussed above and fill the gaps in the larger dataset with a process called conversion modeling. So, if a user opts out of a particular tracking piece or suddenly no longer allows cookies, those gaps can be filled by machine learning.
  3. Data Deletion Made Easy: Deleting data in Universal Analytics is particularly painful because you can’t selectively delete portions of user’s data - you must delete all their data for all time. With GA4 this pain is no more. You can refine and select which pieces of user data need to be deleted, or were requested to be deleted, without losing all the data associated with that particular user.

In addition to analytics tools, advertisers should be investing in their own first-party data infrastructure, such as a CRM, and work with partners to ingest this data directly in platforms.

Remember, first-party data will not be impacted by this change so it is important to tie back results to marketing channels so there is a comprehensive view of how marketing is performing and data to leverage for reporting and optimizations. This will also help provide additional data to machine learning and fill in possible measurement gaps.

Google and other platforms are also working on audience expansion capabilities, to build off of available remarketing lists for further reach, so CRM data will still be an important component of audience strategies. 


2) Choose a durable tracking solution based on your needs and resources

As discussed above with conversion modeling, if advertisers are not currently leveraging Google’s durable tagging solutions of gtag or GTM and conversion linker, they should work on implementing a first-party tag, like gTag.js or Google Tag Manager and conversion linker to build a foundation of data collection for modeling to work off of. For advertisers using Google’s Marketing Platform, Floodlight tags are also considered a durable tagging solution. 

Depending on your digital maturity and your business challenges, consider server-side tagging via GTM’s server side container for higher quality data with the ability to control what is being sent to 3rd-party sites.

Server-side tagging involves running the GTM container in a server-side environment such as Google Cloud Platform compared to running the GTM container on the site visitors browser (aka client-side). 


Invest in Server-Side Solutions in GTM

Stay on Client-Side Solutions (gtag/GTM)




Control & Customization



Implementation Ease




3) Implement necessary and appropriate consent solutions to be compliant with legislation

Set up consent mode (beta) to enable tags to behave according to the current status of user opt-in consent.

With this feature, Google’s tags will dynamically adapt and only utilize measurement tools for the specific purposes that a user has given consent for. 


Checklist of Measurement Considerations

Check out Seer’s Pattern of Privacy hub to learn more about these industry changes overall.

Sign up for our newsletter for more posts like this delivered straight to your inbox! 


We love helping marketers like you.

Sign up for our newsletter to receive updates and more:

Sara Niemiec
Sara Niemiec
Manager, SEO