How self-service advertisers can optimize performance

Missed a session at the Data Summit? View on demand here.

This article is contributed by Jeremy Fain, CEO and co-founder of Cognitiv.

It’s easy to see the appeal of self-service programmatic ad buying. Instead of limited transparency on pricing and placement, advertisers can pinpoint exactly where their campaign spend is going and how much they’re paying for each impression. But as many traders have sadly discovered, this freedom comes at a serious cost. Not only does it take a lot of time and energy to effectively optimize performance, it’s incredibly difficult to produce consistent results on the scale needed. With deep learning at their disposal, advertisers can avoid this endless grind of tedious, unsuccessful work and instead rely on an AI algorithm delivered by integrations such as dynamic Private Marketplaces (also known as DealIDs or PMPs), which automatically, continuously optimize media purchases to maximize execution.

At the moment there seems to be a fairly even split between the number of brands and agencies that rely on managed services and the number of brands and agencies that prefer self-service. A recent survey by Advertiser Perceptions found that 56% use some form of managed service, while 46% say they use self-service. At the same time, 52% of buyers said they plan to increase their self-service spending this year, while only 17% plan to increase their managed services spending. The same study found that The Trade Desk, Amazon Advertising and Yahoo! have each grown into a majority self-service platform in the past year, addressing the widespread desire for greater transparency in programmatic, especially regarding fees.

However, for all the optimism surrounding self-service, it doesn’t solve many performance advertising challenges. Self-service, for example, requires a large contingent of merchants to effectively manage day-to-day operations. As new information comes in about market conditions, consumer preferences, trends, and so on, traders need to be able to quickly synthesize that information to run campaigns efficiently and accurately. However, humans are not robots – we need time to sift through information and parse the relevant patterns before we can design an effective strategy. Given the tight deadlines many marketing teams work with, there is no real way for marketers to consistently produce perfectly optimized campaigns, which leads to wasted spend in the long run. It is also rare for a trading team to have enough traders to effectively cover all of their clients’ campaigns. They usually have to spend most of their time with two or three of their most important customers, while the rest are given less time and effort.

This system of constant, non-scalable trial-and-error also makes it incredibly difficult to work at scale. Many tactics start out strong but quickly fade, leaving marketers scrambling to gather new ideas as they attempt to manually optimize their campaigns. This just makes self-service programmatic more tedious and inefficient than it needs to be — and makes it much harder for marketers to achieve long-term success.

According to Advertiser Perceptions, the main reason given by most advertisers for moving to self-service is “the desire to understand programmatic fees.” More than half (56%) of advertisers cite compensation optimization as a primary reason for making the transition – which is understandable given the strain the pandemic has put on marketing budgets. If advertisers can’t effectively optimize their spend, moving to manual self-service may not deliver the cost savings they hoped for.

Self-service has become indispensable, but the time pressure and the difficulty of finding scalable tactics is a huge limiting factor for success. Self-service advertisers need to find solutions to address these issues. Solutions that will do the tedious work of optimization for them, instead giving them more time to cover all their campaigns equally, and focus on strategy and longer-term concerns. Notably, various forms of machine learning, such as deep learning, have been used by brands such as DoorDash to ensure they can optimize their ad spend at scale.

Deep learning is a valuable tool because of its self-learning, ever-evolving predictive skills. For example, if you trained a deep learning algorithm on customer data, it could identify key characteristics of your target consumer and use that information to make predictions about how new prospects will respond to your ad. This allows the algorithm to avoid advertising to people who are unlikely to convert, while targeting those who are. Best of all, it does this automatically and in real time, and it will adjust its predictions as it learns more about your audience and how they react.

These algorithms are sophisticated and powerful enough to evaluate each media buying opportunity individually. As a result, this enables discreet and dedicated campaign analysis even while the campaign is live, ensuring that real-time market conditions and consumer behavior are continuously taken into account.

Manual self-service ads are difficult, if not impossible, to master over the long term and on a large scale. It requires taking into account many different elements that can change in the blink of an eye, while ignoring the complexities of human behavior. With deep learning algorithms, self-service advertisers can be relieved of the endless pressure to improve performance, instead relying on technologies like AI-driven dynamic PMPs that continuously optimize campaigns and ultimately give them time to complete all their campaigns.

Jeremy Fain is the CEO and co-founder of Cognitiv.

DataDecision makers

Welcome to the VentureBeat Community!

DataDecisionMakers is where experts, including the technical people who do data work, can share data-related insights and innovation.

If you want to read about the latest ideas and up-to-date information, best practices and the future of data and data technology, join us at DataDecisionMakers.

You might even consider contributing an article yourself!