The Evolution of Dynamic Creative Optimization

By Sachin Jogia, Senior Product Manager at AOL UK
02/27/12

Here’s a look at how dynamically personalized ads emerged onto the display advertising scene and what the future possibilities can be. Join us at eTail West, 2/28-3/1, to learn more about our newest generation of Dynamic Ads.

Online Display Landscape

Online display ads have come a long way in the last 14 years since the birth of Advertising.com. Dynamic Creative Optimization is one of the newest and (arguably) most interesting areas of the market today. But, how did we get here?

Here’s a quick look back at the evolution of the online display landscape. In the early 1990’s, banner ads were embedded into publishers’ sites and a direct transaction happened between an advertiser and a publisher... then ad networks came along.

These ad networks, such as Advertising.com, provided the ability to serve the ads on behalf of the publisher and monetize their site for them, usually via a revenue-share agreement, ultimately delivering higher quality and broader reaching content which was optimized for each advertiser. This optimization was a revolution for online display advertising, as it gave advertisers insight into the channels that were performing well for them via a key metric known as the click-through rate.

 

Leveraging Display Data

The next natural step was to start optimizing towards users and their behaviors. The impression and click-tags (which monitored whether a user carried out an action after viewing an ad) could be used to track the behavior of the user and hence start to bucket them into audience segments that could be targeted too.

Over time, tracking became more precise and allowed advertisers to understand both how users were interacting with their brands as well as whether their ads were delivering a decent return-on-ad-spend. This enabled conversion (or action) driven targeting and performance was regarded as being more important than straight branding campaigns for advertisers.

Logically, various companies then decided to capitalize on various gaps in the market, including yield optimizers (who became supply/sell-side-platforms or SSPs) giving power to the publisher, Demand Side Platforms (DSPs) giving power to the advertiser and dynamic retargeters giving power to the user and ultimately everyone in the chain. The latter is an area that is progressing in leaps and bounds.

Now, we’re able to track performance right down to a product level and understand what exactly delivers results for advertisers against which users.

 

Creative Optimization

So what is dynamic retargeting? This technology can be used to produce creatives that are optimized based on the users that are interacting with them, as well as including everything we know about these users, anonymously.  The technology allows advertisers to pass product-level information back to them in order to then serve product-targeted ads across an external network, engaging the users much more effectively and ultimately driving conversions and performance for the advertiser, as well as revenues for the publishers. In its simplest form it can deliver a 100% increase in performance to the advertiser while delivering a considerably higher return to the publisher. (AdOpsInsider.com, Jan. 2011)      

Taking this one step further, the technology is able to optimize any element within an advertisement including the products, titles, fonts, backgrounds and logos, to name a few. A-B testing can be carried out and the best performing elements can be combined for the most relevant creative to be displayed at the time of impression.

Ultimately, we are dynamically optimizing creatives on-the-fly for each individual user that views them, so now, users are not just an impression anymore, they’re a person who has expressed intent.

 

A lot of math

These campaigns can be optimized based on various complex and proprietary algorithms, most of which calculate the probability of a user to carry out an action based on their historical information and insights from similar users. We’re really pushing the envelope here and have the top minds working on the most effective and non-intrusive ways of targeting ads to users.

 

The future

Where do we go from here? Well, there’s a lot more information out there and using this data can go a lot further towards delivering results for advertisers. When we know that a user is actively interested in certain products, there’s certainly a way to facilitate that transaction and encourage the user to purchase the products.

With advertisers passing data through to us from their users’ interactions with their sites, there is only so much optimization we can carry out. Where this becomes really exciting is when we start to include external data too.

What if you could include weather data, or sports feeds? You could then target users based on their external surroundings as well as what they are specifically interested in. What if a user we know nothing about expresses a very similar behavior to ones that we do know information about? We can then use lookalike modeling to show something that is very likely to be relevant to the user. This really is the Holy Grail of online display advertising... giving users what they want, based on where they are and when they want it.

 

What is Advertising.com doing?
Advertising.com has had a successful Dynamic Creative Optimization solution for the last five years, known as Dynamic Banners. This product has been very effective in driving results for advertisers with strong conversion growth for retail advertisers with a much improved return on investment.
Given the advances in retargeting technologies and ability, we are only just touching the surface in this space and looking at pushing the boundaries even further than ever before... Watch out for what’s coming next!

Share