In part one we explored five different factors that would result in a perfect ad. Perfect defined as every impression leading to a business outcome. This required suspending belief at a few points, but overall each of the elements ARE technically possible. But, the actual creative asset needs to do a lot and be specific to each person viewing it. It is unrealistic to make a different ad for each impression so, logically, we need to make the ads dynamic.
Before going deep on the issues, let’s highlight what I am talking about. Imagine one ad unit, in this example a simple 300×250 display ad, (but dynamic can and should be applied to all channels and formats). This particular ad has a background image, a heading, the company logo and a call to action. In a traditional work flow, a designer uses software such as photoshop or illustrator to design the initial version of the ad, get necessary approvals and then make different variations. Dynamic units, have a similar starting point, but all of the variations are done dynamically from a feed. In the graphic below you will see how a simple ad unit can be broken down into its core components, all of which can then be variables loading in different assets depending on what we know about the target user.
This is a very simple example of what can be done today. So, again why aren’t all ads dynamic? The reasons are various and complicated. Way too in-depth for me to explain, but there are many smart people discussing some of them here:
- Media planning and creative separation. The designers should be involved in the targeting conversations.
“But media and creative agencies don’t always agree who owns dynamic creative as a service for clients. Dynamic creative emerged at media agencies because they had the skills to do audience-based targeting. But as the technology has evolved to use machine learning to deliver personalized messages to individuals, it will benefit from more creative thinking, said Diaz Nesamoney, CEO at Jivox.” - Overwhelming amount of data. Creatives need to able to understand and take actionable insights from all available sources.
“There is more data than ever before, and it is only as good as what you do with it. And creativity is still the best way to solve really chunky problems. There has to be more of this marriage.” - Historical processes limiting the advancement of new technologies
“…many traditional marketers and their creative agencies are still largely inclined to make ads that tell the same story to the whole country or planet. And the disjointed nature of ad agency holding companies are a big obstacle to advancing programmatic creative”
Progress is being made and the IAB has outlined a Dynamic Content Ad Standard that will help standardize production. But we still have a long way to go. I would like to keep the dream going and believe that all ads will be dynamic. But not necessarily in how we see them today. Right now, dynamic creative assembles ads from multiple pieces. Pulling from a pre-defined list of messages, images, call to actions and products to assemble what it believes is the ad you are most likely to engage with.
But dynamic creative today essential assembles an ad from a bunch of premade assets of other information. This could be a product image, or a different background, or even something like sports scores or if it is sunny outside. The ads are templates that fill in the placeholders from a set of options. This can be very effective and gets us close to truly personalized ads. A great example of this comes from SpongeCell who produced a campaign for Tennessee Tourism that pulled together video clips dynamically and made a custom spot for you. It is a great example of what is possible today. But even that Cannes Lions-winning awesomeness was compiling an ad from a bucket of pre-made assets.
Perhaps, that is good enough? — You could argue that if there are enough combinations available to make a personal connection to every single person, then we’re pretty close to reaching that perfect ad we defined earlier. But I believe to really get there, we need to go further. TRULY dynamic ads, would generate the imagery and messaging in real-time, based on the thousands of inputs available to them. The closest example I can use to illustrate this is how far video games have come in the last decade. Rendering a scene as you are playing and changing the visuals depending on what is happening. Graphics and processing power are at a point now where it’s photo-realistic and fast. So, if you are a sports brand selling sneakers, rather than making hundreds of image files of different combinations of how the runner looks (male/female/young/old) or what type of shoes they’re wearing (running/gym/casual/fashion) in a different environment (indoors/outdoors/city/countryside) instead of making all of those possible variations, the ad can ‘generate’ the right image based on all of the information it has about the user and the brands business objectives. You program the ad to deliver the right combination. I’m excited about dynamic creative, not because I can show you a different image based on if it is sunny outside, but by the thought that the role of a “creative” (a designer and the ad itself) in the future will not be limited by what combination of pixels you can put together in software. But by what your knowledge and imagination can communicate to a machine, that will then create something, checking all five of the boxes we outlined earlier. Producing the perfect ad… well, set of ads, there really is no such thing as a perfect ad as everyones idea of perfect is different.
Programmatic advertising isn’t killing creativity — it is opening it up to a new world of possibilities.
Phew! — ok back to earth… next week will be more real. I’m gonna make a simple ad in HTML5 using three different tools and show you how I did it. For me, today’s reality is exciting, but tomorrow’s potential is freakin’ incredible!
Next week: Let’s make an ad!
