In this interview, Ben Jeger, VP EMEA at Moloco, discusses the company’s journey, mobile app advertising, and the future of the adtech industry.
What sets adtech leaders like Meta, Google and Amazon apart in the industry?
What sets the likes of Meta, Google and Amazon apart from the others is they excel at achieving results. Return-on-ad-spend (ROAS) and driving real performance is what helped them grow to the behemoths that they are today. These adtech leaders make it extremely easy to get started in performance marketing, which, in part, is what draws advertisers to them. But what really makes them stand out is their use of machine learning (ML). The level of sophistication around ML in performance marketing is clear. They have realised its potential and, as a result, capitalise on it. So, if others want to drive true ROAS outside of these walled gardens, they ought to turn to ML. This technology has the potential to level the playing field and allow the best companies to rise to the top.
What role does ML play in mobile app advertising?
It’s important to firstly understand how ML works. ML uses algorithms and data to teach a system to self-learn and produce outcomes that otherwise couldn’t have been designed or achieved by humans alone. If you think about this in terms of playing chess, for example, you can teach the system to teach itself and become better than you. That’s how ML works. In mobile app advertising, there is a huge amount of user data and signals stemming from installs, in-app purchases, and other engagements. ML can be used to augment the data, analyse bid requests in real time, and then automatically determine the ad format and bid price for a specific impression. To demonstrate the scale of this, at Moloco we receive about 5.3 million bid requests per second, and our ML solution responds to these in about 50 to 80 milliseconds. The scale and speed at which ML systems operate is almost unimaginable, and certainly give mobile apps the edge when it comes to advertising.
What impact has the introduction of Apple’s ATT framework had on the mobile ad industry?
When Apple introduced its ATT framework, it was supposed to offer an alternate measurement solution to the classic mobile measurement partners (MMPs). The issue is this alternate measurement solution was a very flawed attempt on many fronts, it was unclear and this led to the market shifting to probabilistic modelling. The consequence is that in this ecosystem where you have ad networks and demand side platforms (DSPs), there are various players in the market that are incentivised to capture the attribution of user interactions, such as a download, because they get paid on a cost per install, for example (CPI). And so now, it all comes down to the last click. Every advertiser wants to ensure they fired the last click, so they get credited for user attribution. And as a result, the industry is seeing Ad companies render ads in a way that makes it extremely difficult for users not to click.
How is this affecting the overall user experience?
We’re all too familiar with a shrinking ‘x’ button on an ad, or worse still, no ‘x’ button at all. This is incredibly harmful to the user experience. For a lot of companies, the ability to show relevant ads to users has diminished and the actual content is often not relevant and impersonal, because targeting capabilities have been impacted. Ad quality has deteriorated, and a click is not necessarily a click anymore.
App publishers are seeing an increase in users that are unhappy with the ads they are being shown. There is pressure on the publisher side to filter out and remove bad ad experiences, however the issue is that there is a trade-off between user experience and the revenue you can generate from ads.
How can those in the industry (e.g. mobile advertisers, adtech platforms and publishers) navigate these challenges?
It’s important for those in the industry, including measurement partners, publishers and advertisers, to drive transparency into the market. Often, advertisers are unaware of how their ads show up, and that a click is not actually a click. This education is a good first step, but I think we also need to move more in the direction of ads verification, so advertisers can evaluate the quality of traffic and to make it easier for them to understand the incrementality of a channel. And of course, marrying this with attribution is necessary. For instance, advertisers might be happy with receiving lots of clicks and installs as a result of a campaign. But what they should be asking is, did it actually drive incrementality? This is where looking at it from a different angle can prove beneficial — what if we hadn’t run this campaign? How many good quality users would we have acquired without it? Where should we be investing ad spend?
The reason I say this is because it is advertisers who have real value in the system. Regulators do to some extent, but those who are buying ad space have the potential to make decisions that will ultimately lead to a better ecosystem.
What upcoming industry trends do you think will shape 2024?
Featured image: RDNE Stock Project / Pexels