Ad fraud: why there are mismatches in ad spend and ad performance

Programmatic advertisers are buying ad space for targeted campaigns at astounding volume in 2023. But one problem looms large for advertisers as they deploy campaigns – the mismatch that continues to grow between what advertisers spend on campaigns and the performance that they deliver.

The digital advertising performance gap is directly attributed to ad fraud. This is driven by the rise of bots and other fake data that forces advertisers to use their budgets to target fake accounts and other irrelevant audiences. 

Any combination of tactics used to stop digital ads from being delivered to the audiences, or spaces for which they were intended, is considered ad fraud. These tactics can include the use of bots, software development kit (SDK) spoofing and click spamming (a.k.a. click injection), which allow fraudsters to periodically siphon off companies’ ad spend dollars. Meanwhile, the ads themselves fail to generate brand exposure, leads and sales because they were never seen by an actual person. Besides hurting the advertisers, ad fraud also hurts the publishers that serve the ads because it decreases the overall value of their own platform and its advertising business. 

This dynamic comes as no surprise when we look at the state of the internet as a whole, with an estimated 38% of web traffic made up of either automated accounts or bots – 24% of which are bots used for fraud and theft. Additionally last year, advertisers lost an estimated $68 billion in ad spend due to fraud, with $23 billion of those losses occurring in the United States. According to a recent digital advertising survey, 73% of U.S. businesses say that ad fraud is a problem for them and costs them an average of 4% in lost revenue. 

In short, the business of ad fraud is booming. So what do digital advertisers do when it is continuing to cost more and more money to get the same performance on an ad? Here are three tangible ways programmatic advertisers can improve their return on ad spend amid the rising tide of ad fraud.

Tailor CTV campaigns for “fully on-screen” ads

Connected TV (CTV) is any device that connects to a television to support video content streaming – for example Roku, Amazon Fire TV, Apple TV, and others. As traditional cable declines in favor of the streaming revolution, ad spend is increasingly moving towards CTV. By 2026, CTV will account for more than 5% of all ad spending in the U.S. according to Statista. Moreover, nearly six out of 10 connected TV users in the U.S. found CTV ads to be more relevant than ads on linear television.

While CTV enables viewers to watch streaming content without a cable subscription, the relatively new technology also presents issues for advertisers. Namely, one issue that has recently come to light is that when viewers power down their TV screen without quitting a CTV app, programming and ads will continue to run while the viewer isn’t actually watching the ad. When this happens, from the advertiser’s point of view, ads have been served to the audience they paid for. However, since the screens are off, the ads are never actually seen.

By ensuring advertisers are targeting ads that are fully on-screen, they can ensure their ads will be viewed while a device is on instead of off. According to a recent report, ad fraud is 83% less likely to occur when advertisers tailor their campaigns for “fully on-screen” CTV environments. 

Lean on Deterministic Data

When deploying campaigns, advertisers have the choice of using either probabilistic or deterministic data to target their campaigns. Probabilistic data is widely available, and can easily be purchased online from a number of audience vendors. But the issue with probabilistic data is that it is an approximation of a digital identity based on things like page views, time spent on page, or click-throughs. This data is grouped by the likelihood that a user belongs to a certain demographic, socio-economic status or class, but it is not a certainty.

Alternatively, deterministic data is an identifier like an email address or a cookie ID, that is directly linked to an individual and has a likelihood of being 100% accurate. Deterministic data provides a solid foundation for marketing operations because the data is based on fact rather than probability. For example, if a user signs up in one year and gives their current age, it is a fact that the following year they will be one year older. Utilizing and prioritizing this type of deterministic data in ad targeting is another way advertisers can ensure they are targeting the audiences they want.

Expand audiences with Identity Resolution

One of the best ways for programmatic advertisers to ensure that they’re targeting the audiences they’re aiming to target is to solicit first-party data on their existing customers. By identifying similarities within an existing customer base, advertisers can target other people with these characteristics and find other relevant audiences for their products and services.

Many advertisers also assume that ad targeting with first-party data means they are only targeting existing customers. However this rich, first-party data can be used to build custom audiences with machine learning. By analyzing and comparing existing customer data to other deterministic, first-party datasets, advertisers can solicit new deterministic lists of audiences that are similar to their customers. 

To expand your deterministic audiences and learn more about BDEX’s Identity Resolution ID Graph, visit our website here.