An improvement on the LUMA-scape
Mediaplex recently published a new map of the display ad landscape. It’s like a cleaned-up version of the famous LUMA-scapes.
It’s a great graphic, and this post isn’t intended to diminish its value. If Mediaplex neglected to include ad creation, I’d like to explore why, and then figure out what to do about it.
PDF version on mediaplex.com
Where does creative fit in?
Mediaplex seems to have left out any mention of display ad creative, or more specifically, how ad creative originates. It’s as if ad creative appears out of thin air; it’s even difficult to say which ring said ad creative would choose to materialize in. But given the fractured and inefficient way that the display ad ecosystem deals with ad creative, it’s completely understandable why Mediaplex omits it. LUMA is guilty of the exact same thing.
There are, however, sections of the Mediaplex graphic that imply the existence of ad creative:
- Rich Media (a broadly-defined creative format)
- Creative Optimization (the process of algorithmically perfecting an existing creative concept)
- Ad Servers (serving up existing creative files or ad tags)
Ad creative, go stand in the corner.
If Mediaplex and LUMA are attempting to encapsulate the entirety of the display ad landscape, it follows that anything left out must not be part of the landscape.
However, I don’t think that Mediaplex or LUMA are intending to imply anything of the sort; they know as well as anyone that you could remove large swaths of both their existing landscapes (MediaPlex | LUMA) and still have a functioning system where advertisers can run display campaigns. But remove ad creative and the whole thing grinds to a halt.
So why does ad creation get the short shrift? A few theories:
- The industry pundits making the graphics don’t fully understand how or where ad creative originates.
- Many industry participants still buy into the fable that ad agencies are like storks that deliver babies. Agencies are where ads come from. End of story. Now stop asking silly questions that will just embarrass everyone.
- The creation of ads is seen as a given, therefore it’s taken for granted and not scrutinized.
- Failure to grasp the distinction and relationship between the creation of display ads and the trafficking/tracking/optimization of display ads.
- Shallow, uninformed, or misguided assumptions about the creative process.
- An overall bias towards engineering in the display ad ecosystem.
To illustrate just how much is being omitted, I created a rough overlay of where ad creative comes from (below). It’s literally all over the map (and my overlay doesn’t even chart the complex paths that a piece of creative might take after it’s been generated).
You’re a mess, get your act together
As I’ve pointed out, even if the Terry Kawajas of the world wanted to add a “Creative” section to their ecosystem maps, where would they put it? What companies would it include? What would “Creative” be adjascent to and how would it interact with its neighbors? There are literally hundreds of answers to those questions.
Display ad creation is currently a free-for-all. The only real standards come from the IAB, and those mostly boil down to a ever-growing array of width x height pairings. Beyond that, it’s whatever you can cobble together and successfully sneak past ad ops.
What all this implies is that ad creation is sorely in need of standardization. Ad creation isn’t just a necessary evil, it’s one of the lynch-pins of the whole ecosystem, so making it more efficient deserves some serious thought.
Let me suggest a handful of opportunities that would arise if digital display ad creative shared a common set of standards that went beyond basic attributes like width x height or file size:
- No more wasted time haggling with ad ops over Flash versioning and clickTAG minutia (seriously, this alone probably costs the online ad industry hundreds of millions of dollars every year in wasted time).
- A single ad can more easily re-render and re-size across a burgeoning and increasingly unpredictable array of browsers, devices, screen types, and ad unit sizes (to help solidify the implications of this, think where the web would be if HTML/CSS wasn’t a universal standard and every company needed to create dozens of separate but nearly identical websites to accommodate the different screen sizes & devices used by viewers).
- Ads can be generated much more easily. No longer does an agency need field a platoon of designers to spend weeks creating ads in 20 sizes and 6 idiosyncratic creative formats…all for a single campaign.
Need display ad creative? Currently you’ve got some amazing options. Have fun.
Why are display ads still like fax machines?
The standardization of display ad creative also enables a tectonic shift from unstructured to structured data. Let me stress: this will be huh-yooge when it happens.
To illustrate the chasm between unstructured and structured data, think of insertion orders. A faxed insertion order sitting on a desk is unstructured data. It’s impossible for a machine to parse or manipulate whatever information is contained in a paper fax. Any potential scale, efficiency, or automation is lost, locked up in a piece of paper.
On the other hand, RTB transactions are like billions of mini insertion orders, except they’re executed at massive scale via billions of points of structured data.
- faxed insertion orders = unstructured data = vastly inefficient and old-fashioned
- RTB = structured data = highly automated and scalable
Just think of display’s savior, retargeting. How would retargeting even be possible if media buys were all still executed via fax?
Want more on structured versus unstructured data on the web? Here’s an oldie but a goodie.
In a display ad ecosystem that’s increasingly data-driven, automated, and scalable, display ad creative seems to be getting a pass. This is a huge mistake, but rest assured it will change eventually. There’s way too much pent-up efficiency waiting to be unlocked.
I’ll leave you with a final hypothetical to illustrate the potential value of structured data as it pertains to display ad creative:
Imagine the Travelzoo ad above right…the structured-data version that’s machine-readable. If, for every impression, a publisher site can understand the content of all available ads bidding for that impression, the publisher can reliably contribute to the decision of which ads to show. After all, the publisher and advertiser are basically after the same thing: the most relevant & compelling match between viewer and ad.
However, that decision is currently mostly a one-way street. Advertisers have a great deal of control over which publishers/networks on which to run ads, but publishers have less and less control over exactly which ads get shown on their sites, in part because there is no existing way for a publisher to parse the content of an ad in the split-second it takes to complete an RTB transaction or choose an available ad from a server.
This would be akin to Netflix taking down their website and then blindly sending me DVDs based only on their own recommendation algorithm. Sure, some of the movies would be great, but without my direct input into choosing movies, Netflix would send me a lot of unwanted stuff.
In conclusion (finally)
Sorry if it seems like we’ve ended up miles from where we started. Let’s recap our journey:
- Display ad ecosystem maps seem to overlook ad creation.
- This isn’t surprising, ad creation is stuck in the dark ages.
- How would ad creation have to change in order to become better integrated into an otherwise highly data-driven ecosystem?
And there you have it…love to hear some thoughts in the comments.