Consent Mode : Cookies, Personalisation, GDPR

Jump to section:

UPDATED: Consent Mode and the Cookie banner

On January 18, 2024, Google declared that after March 2024 there will be increased enforcement of the GDPR policy for audience and measurement solutions. Google Ads and Google Analytics in particular, causing loss of data, disconnection of GA4 and Google Ads, as well as blocking all user level data collection. 

Indeed this has started to happen for website’s with non-compliant solutions.

What is surprising to us is how surprised and off guard this has caught many ecommerce retailers. In particular those who are heavily reliant on advertising. We work closely with our ecommerce partners, as part of our digital marketing maturity roadmap, to ensure the setup and data collection is optimised as much as possible for robust optimisation and reporting. This includes using advanced conversion tracking like enhanced conversions and measurement protocol API.

Google defines conversion modelling as;

“Conversion modelling uses Google’s AI to analyse observable data and historical trends, quantifying the relationship between consented and unconsented users. Then, using observable user journeys where users have consented to cookie usage, our models will assess attribution paths for the unconsented journeys. This creates a more complete and accurate view of advertising spend and outcomes – all while respecting user consent choices.“

Limitations to advanced Consent Mode user modelling

As of 22 Feb 2024 phone call conversion events are not supported for data modelling. It is also worth understanding that to provide anonymity minimum thresholds need reaching, so ecommerce retailers with low sales volumes and Ad budget may not be able to benefit from user data modelling. A minimum of 700 Ad clicks is needed per week per country. Once the threshold has been met it takes 4 week for the machine learning to start showing modeled conversions in Google Ads.

Check Google Analytics (GA4)

If you are being affected by consent setting and management issues you will see an “Action Required” notice within the data streams section of the system as you can see below.

This shows your data is currently being affected by data loss as a result of non-compliance. If you see this or a flag within Google Ads Conversions, checkout your setup.

A brief history of user data consent

GDPR initially came into force 2018, with the ePrivacy Directive here since 2002, so hardly a new concept. Indeed, everyone is now very familiar with clicking their own usage options on cookie banners as we traverse the Internet. These laws clarify you, as a website owner, must collect opt-in consent e.g. click to agree, rather than click to turn off and deny tracking. 

Of course this only applies to ecommerce stores targeting target people in the European Union or the European Economic Area, including the UK (where this is actually being rolled out slower/ behind the EEA, but will certainly follow suit). Despite being around for so long many websites, including large ecommerce stores still fail to comply with these now old privacy laws with basic cookie consent banners, ultimately flouting the law for better targeting capabilities.

If it has been around so long then why has it become such an issue now?

In short, because of a lack of consistency and compliance across the Internet in collecting user consent via cookie banners. This initiated privacy regulators to enforce compliance at a technical level – through Internet browsers like Firefox, Edge and Google Chrome.

It took this long because Google managed to delay regulation and forcing compliance through automatically blocking third party cookies in the world’s most popular browser Google Chrome, until now. Google has long argued the move would ‘severely affect the economy and advertising ecosystem’ causing ‘billions in lost revenue’, until an alternative technology is created.

They have been working on it for some time and now a range of new Ad technologies powered by APIs are being released. Topic based interest targeting, Google’s Sandbox API and privacy compliant browser-based ad serving technology, are here, but the new technology and eco-system is open to scrutiny by the regulators to ensure it is fit for purpose. The he go ahead will be given in Q4 2024 when Google Chrome will then automatically block third party cookies. 

To use additional AI conversion modelling (a fancy way of saying that any lost data will have the gaps filled in by generalising other results and user behaviour), conversion tracking and personalised advertising (remarketing) Consent Mode V2 needs implementing. Compliance from Google’s point of view is sending them user choices on the storage of personal data for analytics and advertising purposes, as well as processing this information for advertising and remarketing purposes.

Despite Chrome blocking third party cookies later in the year, people are seeing issues with data loss and consent notifications to drive companies to be compliant with existing legislation. The roll out will also be completed gradually and already affects 1% of the global users of Chrome. Fancy seeing what it would be like without cookies? In Chrome you can disable cookies already and try the new privacy settings as a user, including setting topic interests.

What does this change mean for PPC advertising?

Just get Consent Mode V2 implemented already!

Resistance implementing a valid cookie banner was because marketers know very few people opt-in to marketing, which causes a loss of data (we see between 70% and 90% loss across our clients). If your website didn’t have a cookie banner, or the default option granted the use of data for advertising, implementing one will cause a loss of data.

Google claims that 65% of conversion data loss can be recovered using the data modelling available with Consent Mode V2. This remains to be seen, but ultimately the alternative is to lose all advertising and marketing capabilities on Google. Working to new ways of tracking and behavioural modelling should begin now to ensure the latest tools are being used. Alongside this new dataset, a value-based model of marketing and assessment of how and where budgets should be spent is having beyond an overly simplistic “£1 in to get £5 back” mentality.

Performance Max has less data to optimise and perform like it did

AI driven “smart bidding” tools like Performance Max will become less effective at reaching the right customer at the right time, as a result of their being less data available to work with. Given this data also removes a lot of demographic data it has less signals to work with and AI needs accurate, high quality data with enough volume to function. Less data and quality means less optimisation capability and therefore results.

To combat this we work closely with our clients to ensure all the data collection methods are used with as many other sources and tools to improve the data available for campaign optimisation. Our approach focused on segmenting campaigns around the stages of the marketing funnel and specific marketing objectives uses a blended approach, rather than solely relying on Performance Max campaigns. We use a blend of Google Shopping, PPC Search, Display and Performance Max to deliver an optimal mix of control, automation and data learning. 

Move to topic and interest based targeting

Accuracy and segmentation of audiences is a lot more difficult and less accurate. The reality is that marketers have become so (perhaps too) used to collecting a vast amount of demographic and audience data to really narrow targeting and improving the efficiency of advertising spend.

Image credit: Google, 2024 

As a Bournemouth PPC agency, Onefeed is adept at collating and using a wide variety of data to enrich campaigns and ensure they are highly profitable and effective at driving sales at a low cost per acquisition and cost of sale. However, audience targeting is particularly effective due to its highly granular nature. However, there are other approaches like topic and interest based targeting, which also allow for increasing the relevance of those who see our client’s Ads. As an integrated marketing agency we also use marketing mix modelling and marketing funnel segmented reporting to break out the impact Ads are having on attracting and converting customers. 

Get closer to the commercial reality of your marketing

This means a new appreciation of data is needed, which is based on the understanding that analytics and reporting tools, have been and always will be based on sampled data. There has always been an <15% discrepancy with revenue data shown in Analytics/ Google Ads vs what is seen in the back end of an online store. Analytics is a tool for measuring relative performance of cross-channel marketing and not intended as a “Source of truth” with 100% accuracy.

It’s crucial to ensure revenue and sales are measured using actual internal sales figures. Onefeed has access to many of our client’s sales data (without customer data), to measure bottom line impact using blended ROAS or cost of sale KPIs. We still use Google Ads and GA4 data to understand and make campaign/ Ad level optimisations, but look at trajectories and how metrics are performing relatively.

Events are still trackable 

It is important to remember that although personally identifiable information is being reduced this doesn’t affect the standard Campaign / Source / Medium metrics captured by UTM tags and auto-tagging in Google Ads. So it is still possible to get insight into the ROI on Ad spend overall. By ensuring these parameters are setup and applied to all marketing channels a cross-channel assessment is still possible.

Recommendations on what to do next

  1. Implement alternative tracking methods. Run GA4 and GTag tracking side by side as they use different methodologies and having access to both can help gain new insights 
  2. Benchmark sales volume and revenue against GA4 and Google Ads data as soon as possible to understand how much of a discrepancy there may be.  
  3. Review and ensure UTM parameters are setup and universal rules are setup. Check these within GA4, ensuring you do not see (unassigned) or (not set) values. 
  4. Setup enhanced conversions in Google Ads using a privacy compliant method. This involves sending encrypted user data to Google. We can support you on implementing this as part of your PPC advertising campaigns
  5. Use first party data e.g. email lists, offline customer sales lists in Google Ads to help Google Ads identify people with similar interests in topics and interests to improve targeting and optimisation.

We are building the technical and data level stages into our digital marketing maturity model. This helps our clients understand and plot where they are currently, areas for improvement, ways of scaling and taking their business to the next level. Should this be of interest get in contact and we’ll be happy to discuss options.

Technical Requirements for Implementing Consent Mode V2

This section provides a top line summary, but additional information and links to external resources can be found on our Consent Mode V2 blog

Requirements for implementing a Consent Mode V2 Banner:

  • Datalayer, javascript, cookies etc
  • Needs to communicate each choice
  • Banner must let you opt-in/out
  • Must use either gTag or Tag Manager sitewide tagging
  • Consent option setting must be added to every page 
  • Needs to set default consent upon first page load as ‘deny’, until a user accepts consent options
  • 4 defined ‘flags’ must allow consent choices for the options below, which can be updated with any changes made to consent options: 

Standard (Basic) consent mode options

ad_storage : Google Ads and user data storage

analytics_storage : GA4 and other analytics user data storage

V2 (Advanced) additional consent mode options

ad_user_data : Set online advertising consent to allow processing of data for advertising purposes

ad_personalization : Set permission Google Ads to read the cookie for personalised advertising like remarketing

Practically the above can be achieved either through a certified Consent Management Platform, who provides compliance as a service as is often charged for monthly, but requires no coding or development time, other than adding the necessary script. Otherwise plugins and extensions specific to your CMS or ecommerce platform can be found to handle this, but ensuring a solution is fully compliant with testing will be needed. Finally it is possible to manually add the code necessary to provide compliant user consent options. There are a number of guides and help articles available to help guide your developer, like this excellent one for setting up consent mode for Tag Manager.

Key issue on scripts loading on the web page:

When using Google Tag Manager the Datalayer must fire first to push the related consent and tracking information to Google. Consent settings must then be set before tracking scripts on the page are loaded, for valid consent to be sent to Analytics and Google Ads.

Tag assistant will show a warning should this be the case:

Scripts on a website can be checked through viewing the source code in any web browser. Below a search for “consent” shows the code to collect consent is on line 666, but searching for the Analytics ID shows the tracking is loading higher on the page. 

This is important as website optimisation tools (particularly SEO optimisation and site speed tools like W3C Total Cache etc) can combine and move the position of scripts on the page and can change their respective loading position, leading to issues.

How to test compliance with consent mode

Consent settings can be sent to a website through cookies, Local Storage or the datalayer. All of these can be read by using Google’s Tag assistant which is available when you have the necessary accesses to the website’s tags either using Tag Manager or Google Ad’s gtag.

By clicking on the tag at the top of the screen, the “Consent” tab on the left and then again in the output window you should see the event information and consent state for each variable as default (deny).

Next, interact with the cookie banner and allow some options to check whether these values are updated with granted to ensure it updates correctly.

Already accepted the websites cookies and cannot see it to test? On the web page right click and select “inspect”. A pane of developer tools will open. This allows you to then hold down the left mouse button on the page reload button.

This will remove all the cookies and data stored on your computer for that website and display the cookie banner if there is one. It is also possible to do this within the website window loaded by Tag Assistant.

bonus tip: If there is a Universal Analytics tag displayed this now defunct code is still on your website and slowing it down. It can be removed!

We hope this guide has provided a complete base of information from the background of how we’ve arrived here in the first place, to what impact it is having and what you should be doing next. Onefeed are helping dozens of businesses migrate and gain serious competitive advantage in an increasingly changing and competitive landscape. If you want to find out how we can help you then get in touch below.