Why a Souped-Up CMS Is Only Half The Battle

It might come as a surprise, but, suddenly, content management software is very sexy in online publishing circles. It’s a hot topic at publishing industry events. Publishers are investing in better CMS systems to help publish content at a rapid and efficient clip. But it remains to be seen if they’ll invest in systems to monetize that content, or if they’ll continue to look past that part of the equation.

A well-functioning CMS can be used to target ads based on specific pieces of content, marrying relevant advertising to the publisher’s most popular pieces that day, combining high traffic, exciting content, and user targeting. Publishing and advertising are still treated very differently, with direct sales focused largely on run-of-site or channel buys. The majority of ad targeting is based on audiences and broad content associations rather than on specific, well-matched pairings.

It’s great that publishers like Martha Stewart Living and Conde Nast are utilizing state-of-the-art CMS systems, but these publishers are only going half way home. To make online publishing a truly valuable enterprise, publishers need to think about the CMS system as a tool that can open the doors to targeting when used in conjunction with a first-party ad server.

CMS systems now use more data points in content identification and in content rendering than ever before. But how do you define content? Is it the article or the ad? Those two pieces can be tied much more closely than they are currently, reducing any friction behind the scenes. When the ad server is linked closely with the CMS, publishers get dynamic content generation based on several factors, along with ads that are tied more closely to the content. Not all targeting is cookie based on user profiles these days. And we all know that cookie targeting against user profiles in and of itself has its own deficiencies.

Right now, there is a major shift to programmatic buying, and the units being bought and sold are the IAB standards. This trend puts most of the inventory that is currently available into a commoditized state, unless you build in a layer of additional data (user data, contextual and geographic information, etc.). However, that’s becoming commoditized information as well. Everyone has the same data for the most part, unless the publisher has unique registration data that only they can leverage.

So, how do publishers differentiate? In some cases, they need to come up with the next “big idea.” While it’s not revolutionary, I think of the skinning on sites like Break.com and Pandora: huge ads in the background that are tied to certain content areas of the site. For others, the best way to differentiate is to integrate ads into the actual content. This looks to be the best place for a publisher to actually demonstrate an understanding of its audience and how it will interact with content. The publisher can then produce articles that tie this interest into brand messaging (We’re not talking about content farms, but long-form journalism supported by sponsor companies).

For publishers to get to this level, they need their CMS systems to facilitate rapid content production while at the same time identifying content for ad targeting. This requires tagging articles with keywords and terms that the ad server can then use for targeting based on the ad call. The ad server not only has to be content-indifferent, it also must be able to ingest multiple keywords and characters in an article to make the connection between words and actual ads. Big publishers like Gawker and Huffington Post are already doing this.

A high-powered CMS should also be able to adjust the layout of content on a particular Web page based on what ad sizes and types might perform best at a given moment.

Still, this strategy is only the first step. Publishers need to know more about what inventory they have in relation to keywords, tags and content, as well as how that inventory overlaps against other products they sell. If you can’t forecast accurately, you can’t sell accurately. Even worse, publishers can’t understand how other campaigns will compete for the same inventory. The ability to analyze the consumption of inventory once it’s sold (or about to be sold), along with the overlap and other key metrics, is something publishers need to make the best financial decisions.

Essentially, CMS and ad servers need to talk to each other — instantly sharing what tags are attached to an article, for example. And editors need to get into the mindset of creating unique ad opportunities each time they produce a piece of content. Eventually, ad inventory that was previously low value, run-of-site impressions can become high value by simply categorizing it better and sharing those characterizations with ad management systems.

There are different approaches for tying together the ad server and CMS. There are plug-in modules for popular CMS systems, like Drupal, that reduce the manual effort associated with targeting keywords and all the usual errors: fat fingers, misspellings and limited keywords.

There’s also the option of pushing in the first hundred or so characters of each article, then targeting off of those words to better leverage relevancy between ads and content. The ads still appear as IAB units, but they are typically relevant to the article. It’s not the most unique approach, but if publishers can forecast inventory against a subset of characters, they have insight into how the inventory is going to be consumed against certain words.

This has to be the way of the future for publishers. Ad units, as sold today, are undifferentiated and are losing value quickly. A system of machines talking to machines is the future for much of online publishing, whether it’s programmatic buying for performance or programmatic sharing of key information centered on user, content or brand messaging.

This is a trend we’ll see more and more as publishers begin to understand that they need to create a better experience – with both advertising and content – to get bigger dollars.