Advertisers for years have relied on third-party cookies, Facebook’s tracking pixel and other trackers to make the necessary connections for measuring whether their ads led someone to take an action like making a purchase, downloading an app or subscribing to a newsletter. As those signals lose efficacy — as a result of browser tracker blocking, Apple tracking obstacles and the demise of support for third-party cookies in Google’s Chrome browser — advertisers still want to track conversions. So, the two largest digital ad sellers have encouraged them in the past year or so to implement what they expect will allow advertisers to continue tracking conversions (somewhat as they know it) in the future: server-side conversion tracking.
What does server-side conversion tracking do?
Server-side conversion tracking allows ad tech servers — like Facebook’s and Google’s — to connect directly to the servers that host an advertiser’s or retailer’s site. In the same way that a cookie or tracking pixel does today, the process sends information from an advertiser’s site to those tech platforms’ servers, telling them that someone clicked a product page or bought something, essentially showing that an exposure to an ad led someone to convert.
So, if it does what pixels and cookies already do, what’s new here?
Bypassing the browser. The fact that the advertisers’ server is communicating directly with Facebook or Google without the web browser getting in the way is what’s new. Those cookies and pixels — which must be set on the browser — are already not working as well as they used to and eventually will stop working all together. This is a way to circumvent browser blockades to ensure that ad platforms can access the data advertisers rely on to measure campaign performance now and in the future when a website’s server can no longer set trackers on the user’s browser. The browser is also referred to as the “client,” as in “client-side” tracking.
“Pixels and cookies still work for now, but you can also say, ‘I also want to send the same data to the server,'” said Farhad Divecha, founder of UK-based digital marketing agency Accuracast.
What ads can I use this for?
In the Google and Facebook contexts, the types of ads that are measured varies. Google’s server-side tracking allows advertisers to track how ad exposure ties to conversion for inventory bought in all types of channels including Google or Facebook ads, ads bought through demand-side platforms and via e-commerce platforms like Shopify. Facebook’s Conversions API, on the other hand, is used for tracking ads purchased through Facebook.
Ad tech companies including Mixpanel, TransUnion’s Signal and Twilio Segment also offer server-side conversion tracking for ads purchased from any ad seller.
OK, but how does the ad platform know an ad exposure led to a conversion?
Well, the process requires a piece of data to match an ad exposure to a conversion, so the advertiser has to pass along an identifying piece of data of some kind — an email address, phone number, IP address or some data point associated with the person who made a purchase or converted — to Google or Facebook. And, Google or Facebook can make the connection because it’s likely that person who saw the ad and converted is logged in to those platforms.
Isn’t this just an old-school audience matching?
Sure, firms like LiveRamp have enabled advertisers to upload offline or online data to gauge ad performance for years. But this server-side conversion method requires that data linkage to happen via a direct communication between the big tech platforms and the advertiser through their servers.
Wait a minute. So, now these platforms want more data?
Depending on how they implement the connection — either through a partner like Shopify or using their own in-house servers — advertisers have varying levels of control over what information is passed to Google and Facebook.
“The server-side option was built as part of our ongoing work to give advertisers more control over their users’ data,” said a Google spokesperson who noted that the company has rolled out regular updates enabling various aspects of server-side conversion tracking since launching server-side conversion tagging a year ago.
Facebook, which refers to its server-side conversion tracking as CAPI, or Conversions API, also says it is designed to give advertisers more control over what data they share and when the mechanism is implemented separately from the Facebook tracking pixel.
Ummm, what about privacy?
To look at it one way, server-side tracking keeps certain data out of the browser — especially sensitive stuff that advertisers want to make sure is secure. And ad agencies, which are encouraging their advertiser clients to implement server-side conversion tracking, said the information is encrypted.
In general, brands implementing server-side conversion tracking are relying on existing privacy policies and terms and conditions that note that people’s personal information might be shared with marketing partners, particularly if they already share data for purposes such as audience matching with Facebook and Google, said James Parker, chief solutions officer of data and planning at digital agency Jellyfish.
“Given the general practice of sharing data with the ad tech vendors right now, most marketers have covered this in one way or another already,” Parker said. However, in some cases, other requirements for people’s consent could prevent the necessary server-side data sharing. Europe’s General Data Protection Regulation “can be quite a roadblock,” said Parker.
So, is this a way for Facebook and Google to get around Apple’s privacy updates?
Not quite. While agency execs including Parker and Divecha say advertisers that have implemented server-side conversion tracking are detecting conversions they may have missed as a result of tracker deprecation, it won’t fill in the gaps Apple tracking changes have created.
“There are still details [advertisers] can’t get,” said Ty Martin, founder of Audience Kitchen, which helps advertisers uncover targetable audiences on Facebook and Instagram. “It’s still an incomplete solution.”