What is the latest snake oil in AdTech? (with Memes)

Disclosure: These opinions are my own & do not call out specific companies

The industry of digital advertising is much like any other, one that is built on a set of foundations that become the norm over time. There is a lot of interesting innovation but of course many things that are effectively Ponzi schemes / completely the opposite what they are intended to be. Whilst I can write about this in pretty much any area of AdTech, let’s focus on some of the key ones currently.

Attribution

The state of measurement with regards to digital ads is at its most pivotal point you could argue, with the changes to technology / regulation directly implicating how to measure the success of ads. Attribution at its heart was leveraging user level identifiers to events to measure success.

This sector has had many attempts over time for the new “solution” that could do it all, but in reality could not. Whether this be looking at forms of third party redirects through impression tracking or fractionally attributing success in a blackbox, these generally speaking have fallen away once the majority have realised what their true capability is.

Whilst the evolution away from a cookie / mobile device id occurs, there is an interesting space developing with regards to some newer solutions emerging. Some looking at server side solutions. Some looking at conversion modeling / machine learning. Some looking at alternative identifiers. But the reality is that a lot of these solutions will not stand up when you dig deeper or aren’t actually futureproof at all. There are certain things that are a given, whether that is Apple’s approach (SKAdNetwork or Private Click Measurement) or Google’s Privacy Sandbox Attribution Reporting API. The smarter “solutions” will be looking at this as a means to power any alternative offering. And don’t be believing that click based attribution solutions are safe either. And that leads onto the next one.

COOKIELESS SOLUTIONs

This has got to be one of my favourite things to look at currently. No doubt the large majority of advertisers / vendors realise that the days of 3rd party cookies are numbered & are looking for alternatives. But at the same time, we are seeing the emergence of both existing vendors saying they are cookieless to begin with or new vendors showing up with some unheard capability to do something nobody thought was possible.

Ultimately for all the good privacy centric solutions that actually do not require a cookie to operate, there are loads of solutions that to this day, still depend on a cookie. Why? Two things really that seems to confuse the industry. One is the difference between 1st & 3rd party cookies. Whilst said cookieless solutions may pivot away from a 3rd party cookie, guess what, they are still leveraging a 1st party cookie in a core part of their methodology. And unless you are a walled garden, that is going to get you nowhere especially in the activation cross-site space. Clearly Apple are also not content with their privacy approach currently and are going out of their way to limit how 1st party cookies work for advertising, to which these solutions will have troubles with later down the line.

The second one is ultimately how the flow of data operates in AdTech. In particular in Programmatic, there is still heavy dependency on cookies, whether it is cross-site syncs with 3rd party cookies across SSPs / DSPs with client side pixels or how data is integrated into a buy side platform. Take DV360 for example, the large majority of data piping into the DSP is still going through their legacy cookie match endpoint / API. Theoretically, the “cookieless” approach is to leverage their Customer Match endpoint / API to match on a hashed PII basis, but the large majority of vendors have not integrated with this yet. This whole flow of data is having to evolve in order to truly become cookieless, to which some tech stacks are much more further down the line.

AI

The buzzword of 2023 not just in AdTech has to be AI, with the emergence of OpenAI’s ChatGPT and Google’s Bard bringing about a shift in efficiency across a large range of different use cases. Concentrating in advertising however, AI has been a core component for sometime in how the programmatic landscape operates, whether this be in algorithmic bidding or through the use of optimisation.

But as Generative AI in particular takes off, so has the amount of vendors & agencies suddenly showcasing some AI capability. Don’t get me wrong, a strong data science / marketing analytics approach combined with technology would have come to this conclusion by default, but what these newer AI products have introduced is a large amount of smoke & mirrors as to what truly is unique vs effectively theft.

Plenty of opportunities exists for AI to redefine how advertising is planned, activated & measured but its important to get the facts on what exactly is being delivered.

CONTEXTUAL

Another popular solution for ad targeting in a world of privacy centric marketing is contextual aka the art of leveraging the context of a page / app to serve a relevant ad to an user. Historically this was dominated by a few large vendors within the programmatic space.

But as user identifiers decline & context is seen as a core signal to bid effectively with, all sorts of solutions have come to market. The keyword nature is slowly evolving, which is not a bad thing but where the grey area really ramps up is in the actual methodology of some of these solutions.

Whether it is exactly how the context is derived, the emotive nature of it or the prediction of what behavioural traits to tag alongside it, technology has to be strong to do it accurately, which in certain cases is questionable.

At the end of the day, if a platform / vendor doesn’t have a contextual solution right now, they are at a disadvantage to marketers. Old school key value adserver techniques are still viable especially with curated publisher buys when tied with dynamic creative. Where the effective “war” is happening is between said publishers and vendors who may operate at scale to classify these pages. But once again, methodology and proof of what it is doing is key before you start paying for the service

CLOSING THOUGHTS

There are plenty more areas of AdTech that have a lot of noise blurring from the reality of what is actually real vs fiction. What is imperative really for anyone operating in the space, is to truly understand what is possible in the current / future state vs the methodology of some of these areas.

For sure this is easier said than done, where in some instances, you will really never be able to get into the weeds based on how these solutions have been designed. But AdTech is built on big data & big data can be used to prove instances where needing to go deeper. It is becoming a more complex ecosystem but leveraging critical thinking is an under-utilised process, to avoid funding things that are deceptive in nature.