Dynamic creative is not a slot machine

Blog posted by Alex White
Comments (0)
VP Product Strategy at Sizmek, Alex White, talks about Dynamic creative optimisation (DCO) and the prerequisite of personalisation in digital advertising in 2016.

Personalisation is a prerequisite for digital advertising in 2016, making impactful creative the gateway to success. After all, serving the same creative to every audience in a targeted media buy is the equivalent of a politician campaigning to targeted demographics and special interest groups, but giving the same speech to every audience. At best it represents a missed opportunity, at worst, a fruitless endeavour.

This focus on targeted messaging has many brands searching for solutions to facilitate data-driven creative personalisation, and there are as many names for this type of technology as there are vendors. There are also significant differences between the various approaches to creative personalisation on the market, in addition to misinformation and false promises.

Dynamic creative optimisation (DCO) first emerged in the market primarily as a production tool with limited targeting capabilities. Many point solutions, especially those tied to media, focus on limited methods of creative personalisation, such as retargeting, and many have yet to build the infrastructure to leverage new innovations in data management. The leaders in dynamic creative, however, have grown with the rise of data and programmatic media to accommodate full-funnel, data-driven audience targeting while maintaining a commitment to scalable creative content management.

To capitalise on the emerging programmatic, data-driven market, many vendors in the space have begun offering programmatic creative solutions distinct from dynamic creative - asserting that DCO is too manual, while their technology automatically discovers the best creative for each audience with minimal input from the marketer. This is a flawed premise that should be challenged.

Dynamic creative is not a slot machine. Marketers cannot simply drop a bunch of values into each dynamic element ‘slot’, such as images, copy, calls to action, let the algorithm pull the lever and then expect it to pick the right creative for each user with a steadily increasing success rate. Casinos love the slot machine because they make money every time a random permutation fails to match the right pattern, which is statistically very likely. But as an advertiser, you are looking for a big return on your ad spend, which dwindles with every ineffective permutation served. Even with a smart algorithm optimising towards the right combinations, why leave it to chance when you can apply strategic thinking to narrow your options? Success comes down to how marketers inform their dynamic creative strategy with their knowledge of their target consumers.

Many RFPs indicate marketers want DCO, but expect that they won’t have to devote time or resources to make decisions up front. This isn’t possible – advertising is about the human connection, and how each campaign functions within the context of a holistic marketing plan.

The best DCO campaigns rely on context and human decision. Marketers know their products the best, and they have information on what kinds of customers are interested in or looking for those products and services. Dynamic creative is a tool through which they can bring their audience targeting strategy to light. It is not a magic wand, nor is it a substitute for strategic intelligence. Instead, DCO facilitates strategic intelligence – marketers can use DCO to implement complex targeted messaging strategies at scale, and quickly take action on insights gained from DCO, but it will not do all the work for them.

The issue here is that the seductive promise distinguishing programmatic creative from dynamic creative – that it completely automates creative personalisation -- goes against how brands have planned their advertising for decades. Brands have always maintained control over messages, especially the creative components and how those align with the wider campaign. Why would they suddenly turn that strategic component over to technology, simply because the technology exists? A machine will never be the best creative strategist. After all, computers are only as smart as the people using them.

So, what does human-guided DCO look like? The most successful marketers begin their campaigns by considering which audiences they’re looking to reach, what success looks like for their brand, and what creative elements will connect with each consumer segment across the purchase funnel. They view dynamic element values in the context of how they are combined, who they’re serving to and when, and they maintain fine control over this strategy. All of these elements tie into the goals of the larger campaign and brand identity, which only a human can define.

The campaign itself is not executed manually, of course, and DCO solutions eliminate hours of human effort, which can grow exponentially as the strategy grows in complexity. But when marketers go through this process, all of the decisions are human at the end of the day. In turn, those decisions fuel the auto-optimisation engine. This technology can deliver better when it considers dynamic creative elements in context of their combination and when those combinations are deliberately designed.

Let’s consider the importance of optimising creative combinations rather than individual elements with an example of a dynamic background colour and product colour. A slot machine approach may find that the blue background is performing very well on average, and will continue to serve that element more. But what if the blue background only performs this well when paired with a white product because it makes the product stand out, and terribly with a blue product because the product fades into the background? Slot machine optimisation serves a blue background regardless of how it looks or performs in context, while creative version optimisation only optimises the best-performing, deliberately designed combinations.

The consensus across the industry is that that there should be a union of creative and data. Even in this age of machine-automated marketing, the marketers themselves still make the best decisions about their campaigns, and they should keep making those decisions. This is especially true for creative agencies. No machine is more important than brainstorming and art direction – those processes aren’t going anywhere. This technology simply empowers creative decision making and brings a strategy to fruition.

DCO may help achieve powerful targeted messaging at a global scale, but it doesn’t know about a business or its customers. Marketers need to collaborate with solutions experts and apply strategic, creative thinking to get the best results. When marketers make decisions in the beginning, utilising the inferences and insights they’ve gathered, and use technology to power that strategy, they’ll get the results they hope for. When it comes down to it, personalisation will always require a person.

Add comment

Log in or register to post comments