8 minutes

Bedrock – Guest Post | How to maintain accurate conversion data in a privacy-first world: What marketers need to know

Bedrock’s Analytics & Digital Strategy Director, Scott McMullan runs us through the items affecting your data accuracy, the impact they are having and what you can do to retain the most accurate marketing data possible…

For those that prefer to watch, you can recap on the recent webinar Scott delivered on this topic using the video below, or for those that fancy indulging in a longer read, check out the full article in depth below!



Getting accurate behavioural & conversion data is no easy task… 

It used to be easy. Companies could collect data willy nilly with very few restrictions for users that landed on their websites or used their apps. With the rise of privacy laws, browser updates and Ad Blockers, it is becoming much harder to get an accurate view of performance. 

In this article, we’ll run you through the items affecting your data accuracy, the impact they are having and what you can do to retain the most accurate marketing data you possibly can.


GDPR, CCPA & Country Specific Privacy Laws

The GDPR needs no introduction. Though for those not familiar, it prohibits the collection of data for users that have not provided consent. This consent is typically collected through a cookie banner asking to opt into tracking. For a complete overview, you can visit the GDPR website

What is interesting is that other countries/states are following suit and implementing similar legislation. There are at least 13 different countries with GDPR-Esque tracking laws now, and it looks like this trend will only continue. 

We’ve implemented GDPR tracking for many clients and typically see opt-in rates of between 50% – 75% depending on banner design. Below is an example of a client with a 29% decline in sessions reported after moving to GDPR compliant tracking.



Improving Cookie Banner Opt-In Rates 

There are ways to improve the number of users opting into tracking through testing. Many vendors that provide cookie tracking banners, such as One Trust will provide facilities to A/B test different types of banners, positioning, and copy. 

You should definitely take advantage of these to ensure you get the most accurate data sample you possibly can. We’ve seen uplifts of up to 50% in the amount of data being captured by a/b testing forms.


Ad Blockers 

Ad Blocking is no longer a niche sport for technical users. According to Statistas study users that have an ad blocker installed now sits at 27%. That is a considerable number.


This adoption rate is a worry for marketers as some of the most prominent players, such as Ad Block, give users the option to block trackers as well. 

We need to make some assumptions here about the percentage of users using the advanced blocking features. If we assume this to be about 50% of users, your total exposure from a data accuracy standpoint would be a loss of 12.5%, which is significant.


How to collect data from users leveraging ad-blocking technology

How to maintain accurate conversion data in a privacy-first world: What Marketers Need To Know 2 

Disclaimer: Remember to follow the laws in your country, such as GDPR. If a user has opted out, make sure not to continue tracking them. 

If your user has opted into tracking but uses an adblocker, we can still collect this data legally. Ad Blockers will block traffic to domains including facebook.comgoogle.com, etc. What we can do instead is moving to a more sophisticated tracking setup that uses Server Side tagging. 

Server-side tagging allows us to collect data in the first-party context as the tagging server uses your website URL; this will not block tracking. Google has recently come to market with their own server-side offering, which you can find out about here

Setting up a server-side tagging platform is a complex process and comes with a cost. There are many other benefits to running a server-side implementation besides data accuracy. If you’d like to find out more about this, don’t hesitate to get in touch. I’d be happy to talk you through the process and what it entails scott@bedrock42.com.


iOS 14.5 Update 

The iOS 14.5 update was another big privacy update from Apple on April 26th 2021. With the release, all iOS users will automatically be opted out of ad tracking on all apps. Users will now be prompted by apps when they are installed to opt-in or out of tracking. This has a sizeable impact on paid ads, remarketing and audiences building in particular. 

Analysts from Flurry Media have estimated tracking has an opt-in rate of 5% in the US, rising to 13% for users outside of the US. Before the update, tracking opt in was estimated to be accepted for around 70% of users. This is a huge impact, but what does it mean for ad platforms?


iOS accounts for 14.5% of the smartphone market, which is sizeable, particularly if your target customer tends to use one of these devices.

Performance Changes 

Now that tracking is blocked on most iOS apps, any of your campaigns that target iOS apps specifically will see a drop in performance. This has also hindered the algorithms temporarily. Typically marketers would set up automated bidding strategies to make the most of their marketing dollars. This change in data availability and performance data will have thrown the models, but chances are they are retrained now to be more effective. Bear in mind this update will not have had any effect on search ads. 

Some companies have begun redirecting their ad spend from iOS into Android if they are capped by budget. This way, they have access to more robust data sets to understand performance. Facebook and Google recommend splitting out your iOS campaigns from your others so that this lack of data does not pollute other campaign reporting accuracy.


Statistical Modelling For Lost Conversions 

The performance of ads is a very important topic for marketers. If they’re not returning an ROI we like, they’ll quickly get the chop. On the back of this, vendors are beginning to employ statistical modelling to fit the gap. 

Google call it Enhanced conversion modelling. Facebook has decided it’s statistical conversion modelling, but whatever the name, it’s roughly the same idea. Vendors will use machine learning to estimate lost conversions due to lack of data, as well as cross device conversions and view-through conversions.


Audience Sizes 

Now that there’s a small number of users accepting tracking, it’s going to really shrink the size of audiences available from these different vendors. Particular when it comes to demographic or interest targeting. This will impact your remarketing audiences and available target audiences.


What you can do 

Maybe reconsider on which devices you spend your marketing monies on. If you are capped by spending, potentially shifting it across to Android makes sense if that’s the device your customers use. Be cautious of statistically modelled conversions. The larger your data set, the more accurate these will be. Stay vigilant if you’re dealing in small numbers, as the modelling will not be as accurate.


3rd Party Tracking Cookie 

The final big change that will be impacting the accuracy of your tracking data is the updates around 3rd party tracking cookies. Many of the large platforms, including Google, Apple, Mozilla and others, are beginning to make the use of these cookies very ineffective. 

3rd party cookies are cookies that are stored under a different domain than the one you are visiting, for example, a cookie set by facebook.com on your example.com domain. These are used to track the movement of users around different websites and remarket to these same users. 

Ever feel like there are some ads following you around the internet? This is how that technology typically works. The biggest change will come in 2023 when Google blocks these cookies altogether. You can read more about that here.

Many companies like Apple have been squeezing the effectiveness of these cookies to 1 day and were the first company to begin moving in this direction. The importance of Google and 2023 is that Chrome has around 65% of the browser market.


How to protect against 3rd party changes 

The easiest method is to begin moving all of your tracking and pixels to the 1st party context. This is where a 1st party cookie is set instead of a 3rd party one. Luckily lots of the bigger players like Google Ads and Facebook have already made that move. Smaller publishers are a little further behind, and you should check with all of your partners what the process to move across looks like. 

There may be some technology on your site that may cease to function if there is no 1st party alternative. Our recommendation is to create a list of all of your technologies and partners and contact them directly to see what their provisions are. 

There will be some technical work to move across to the first party, and you should contact your analytics partner or resource to support you in this.


Final Thoughts 

You need to work hard to get the most accurate data possible. It’s an ongoing piece of work that changes all the time, but accurate data is worth it 10x over when it comes to optimising campaigns and budgets. 

Here are a few final takeaways: 

Move to server-side tracking

Server-side tracking has the ability to quickly boost the accuracy of your data by an estimated 10%. There are numerous other benefits to server-side, including boosting site speed and enhanced data security. 

First-Party data is now more valuable than ever 

You should begin collecting as much first-party data as possible – this could be through signup forms on-site, feedback, sales data, login data etc. This means you will not need to rely as heavily on third parties in the future Ensure you move all technology and partners to the first-party context.

Create and use Customer Lists and Similar/Lookalike Audiences 

As audiences are shrinking, creating lists and similar audiences through Facebook and Google is more important than ever. Upload your first-party data to these to begin advertising to new audiences similar to your buyers. 


It’s hard to keep your finger on the pulse, but hopefully, this article has cleared some things up. We’re data nerds and love keeping up to date with new trends. If you have any questions about this topic or anything else, be sure to get in touch through our website or scott@bedrock42.com.