DIGITAL ROUNDUP

April 2025

Welcome to this month’s edition of the roundup to some core updates in the digital advertising industry.

NB: the following represents my own opinion & not that of my employer

Adalytics drops another research on pre-bid & bots

Adalytics have made their latest research public, calling out the 3 largest verification vendors over their serving of ads to bots.

(https://adalytics.io/blog/prebid-bot-filtration)

DEEPDIVE

Ad Verification continues to be under the microscope for all parts of the industry and this time attention has turned to how pre-bid targeting is done, specifically around the identification and exclusion of bots from being served ads, in a theme to only target human users. The latest Adalytics research into this world suggests that the 3 leading ad verification tools in IAS, DoubleVerify & HUMAN have all been guilty of not actively preventing bot delivery, specifically with evidence shown that “known bots” as defined by the IAB / TAG were actively been served to.

Whilst this remains a convoluted topic, due to the dynamics of OpenRTB and what information may be exposed by the sell-side / buy-side, the idea of pre-bid ad verification targeting is something that the large majority of brands will tick a box and hope that it is actively doing something. It is hardly a cheap investment either, which is all the more reason why there will be many questions raised around the technology actually doing what it is selling it does.

In fact, you could argue this is about to get a whole lot worse once the rise of AI brings about the challenge of understanding whether an ad is being served to a human or some form of agentic solution trying to do something on a human’s control. However it is important to distinguish between the types of invalid traffic & whether it is defined as general (GIVT) or sophisticated (SIVT). You could argue that GIVT should be a standard exclusion at this point in time.

Both IAS & DoubleVerify have responded to this latest fallout, with DV in particular pointing out that they refund advertisers for this invalid traffic. But the bigger challenge here is the fact that it is happening on government media budgets, which has now seen Congress / FTC / DOJ start to get involved in the US.

If we take a step back, it is time to understand what exactly pre-bid targeting is doing before to invest in it. It is not a magical tickbox that seemingly solves everything. And on the flip side, investing too heavy in a pre-bid approach that is setup with the wrong methodology will inadvertently impact your programmatic delivery/performance. Optimisation to humans is more logical than technological, whether that be on the type of browser / device or controlling exactly where you are serving on with curated supply.

Scope3 seeks to redefine verification & curation with an agentic solution

Scope3 revealed their updated solution to tackle ad verification in the age of AI

(https://scope3.com/news/scope3-reimagines-media-buying-to-power-safe-and-sustainable-growth)

DEEPDIVE

Scope3 have been a very interesting company to follow, from its sustainability roots and more recently to their first acquisition of Adloox as an ad verification vendor. But their latest announcement to the wider ads ecosystem looks to shape the future of what ad verification means, bringing together a need to evolve legacy approaches in tandem with the rise of AI / agentic platforms.

The pitch is simple, make decisions at the supply level before it gets translated into the buy-side within DSPs / ad platforms. The means to curate not only where you are buying, but how you are buying through the use of data / algorithms is something that is slowly being adopted by leading SSPs in tandem with external parties. But factoring in an AI approach can significantly increase the scaling to such a model.

From a pure ad verification perspective of defining what is brand suitable / safe, this has been heavily implicated by a lack of outdated technology or technology that has been called out multiple times (as shown by plenty of issues over the years). Blanket keyword lists or using keyword blocking based on the structure of an URL has never worked. Notably even DoubleVerify have finally started to move away from this model as a means to reduce ad blocking on news publishers.

Scope3 has some major partnerships for their Agentic Media Platform at launch, notably both Meta & Amazon DSP. These 2 platforms being Walled Gardens are a great win and have both struggled from brand suitability challenges over the years. There is also a chance for the supply side in Index / Azerion and more agile DSPs in Hawk to lead the charge here. The intriguing question is this use of AI may somewhat contradict the sustainability lens, given that AI is not particularly sustainability friendly to maintain / use. Though the argument of saving irrelevant bids is a fair one.

Is this model a threat to not only DoubleVerify / IAS but the dynamics of a SSP to a DSP? Absolutely. The use of tickboxes and blunt setup feels a little archaic at this point in time, bringing with it a lack of true customisation to advertisers. But as outcomes become more the word of choice for advertisers, the majority of tech vendors have already made their intentions known to move towards this way. Though ultimately ad verification / brand suitability still matters and should not be based on a sample.

OTHER KEY CALLOUTS

SOCIAL / SEARCH HIGHLIGHTS 📱

  • Meta have announced a raft of platform updates at the annual Shoptalk event (Meta)

  • LinkedIn have announced some updates to Campaign Manager (LinkedIn)

  • Meta have rolled out Community Notes testing in the US to start with as of March 18th (Meta)

  • Microsoft have made some updates to their ad platforms, specifically around AI (Microsoft)

  • Google will be releasing the ability to pull search terms on PMAX campaigns (SEL)

  • Meta have revealed some additional insights into its AI models for Ads (Meta)

  • Google have announced new rules for PMAX brand guidelines (Google)

PROGRAMMATIC HIGHLIGHTS 🌐

  • Innovid have made the move to combine the legacy Flashtalking name with their existing solution, to provide an independent adserver (Innovid)

  • Google have made some adjustments to Google Ads display (Google)

  • Viant have made a large acquisition in the form of Lockr for privacy first party data toolkits (Viant)

  • Amazon have announced a new ad solution called Complete TV in the US (Amazon)

  • Google have announced the use of IP address for CTV targeting / measurement in the US for DV360 (Google)

  • Google have released the DV360 v4 API out of beta (Google)

IDENTITY HIGHLIGHTS 🆔

  • Infosum have announced an extended partnership with ITV in the UK (Infosum)

  • Utiq have announced a new partnership with Adobe, primarily with their real time CDP (Utiq)

  • Google have introduced support for consent mode on their own consent management platform (Google)

  • Amazon have announced they have integrated with the Google Privacy Sandbox for Attribution Reporting API (Amazon)

  • Google have issued a warning to all websites leveraging a Google Tag, with GTM set to auto-load for non-GA4. (Simo Ahava)

  • Microsoft have announced that their version of consent mode will be mandatory for all their ad platform by May 5th (Microsoft)

For any additional questions or comments, feel free to contact on the form below!