TagMan is obsessed with tags and, as the first tag management system, we were among the earliest pioneers of the idea of server-to-server tags (or ‘servertags’ as we called the concept).
But, as things have changed and our thinking evolved, we are now of the view that server tags are an unwise solution to the challenges of marketing data collection, transfer and integration.Server tags enable data being collected on a website to be passed directly from the data collection system’s server to the servers of the 3rd parties who need to receive the data. This avoids the need for 3rd parties’ on-page tags using the website user’s web browser to execute the data transfer. The conceptual benefits of page speed improvements and data integrity were certainly exciting to us.
But, as our thinking around the implications of this technology, and the online data privacy landscape have both evolved over the years, we now have serious concerns as to the legality and wisdom of server-to-server tags in the United States and Europe. TagMan is committed to privacy and does not want to be perceived as enabling any clandestine user-data passing between hidden servers, undetectable to the user. Where the digital industry’s data collection practices are under such intense scrutiny, can we really afford one or more parallel systems of behind-the-scenes data passing?
But, it’s not just user privacy that has led us to move away from server tags as a data transfer mechanism, there are practical issues too. We have, in fact, three primary concerns with the server-to-server approach:
The lack of transparent opt-in, or in some cases even a basic opt-out, capability for end users to control their data being passed around from system to system, would seem to fly in the face of current trends and legislation. Furthermore, even if an opt-out system is in place, how does it fit into the larger, already overly complex opt-out ecosystem of browser-level do-not-track, vendor-level cookie opt-out, NAI, DAA, IAB, PrivacyChoice, etc.?
Online marketing vendors have enough to think about and do without being told they have to spend their valuable time and resources developing integrations with the next server-to-server technology provider that comes along. In other words, we are a very far cry from having an industry-wide protocol, with necessary functional and data-protection features, that is in a position to replace the current tag-and-cookie protocol in common use today – trying to force the issue simply leads to wasted effort and frustration.
3. Data loss
Anecdotal evidence suggests that, far from ensuring data integrity between systems, there are data loss factors between servers that actually lead to discrepancies as great or greater than the browser/cookie approach. And once that data is lost, it is next to impossible to recreate.
There is a long history of contentious developments in opaque online data methods. Consider the following:
- Flash cookies / Local Shared Objects (LSO’s)
- Browser history sniffing
- ISP data deep packet inspection (NebuAd / Phorm)
- Aggressive online/offline behavioral tracking (Doubleclick / Abacus)
In our view, the latest innovation in this trend is this direct server-to-server sharing of cookie data and, even if there were clear benefits in terms of the speedy and accurate sharing of data, the ever-chillier climate in which our industry’s data practices are being viewed suggests we should seek to evolve down a different path.