The Display LUMAScape Explained

Media Planning and Attribution

Another more niche area of digital, media planning and attribution are typically sophisticated analytics platforms that help marketers built a better picture of their consumers, but also where their media campaigns are most successful.  There’s been an ongoing argument in digital media for a long time about how to allocate credit for online conversions if the consumer saw an ad in more than one place.  A simple model is called “last-touch attribution” which gives 100% credit to the last ad a consumer saw, which is often a search ad; but many media companies (especially those that serve the non-search market) would say that the only reason a consumer was searching in the first place was because they were aware of the company or product, which was most likely driven by a display ad.

Dividing the credit in some fractional manner (called “multi-touch attribution” to many different channels is complicated, but theoretically leads the marketer towards smarter campaign optimizations.  C3 Metrics, Korrelate, and Clearsaleing (now part of eBay) are all in this business, while others like Pulsepoint can handle both the analytics as well as the optimized media buying, and still others like Networked Insights specialize in merging data from multiple sources to feed predictive models and provide broader consumer insights.

Tag Management

Another Ad Ops heavy part of the digital space, I’ve always thought of tag management technology as an ironic joke within the industry.  The landscape has gotten so fragmented, and publishers and advertisers are both working with so many technology vendors that they often need a technology to manage the technology(!), and that’s the service tag management provides.  To be a little more specific, technology partners often require some kind of tag on page to work – that’s true of data management platforms, retargeting platforms, analytics vendors, and more.  But making changes to the source code of a webpage is often a non-trivial exercise at a big company – it can take a long time and involve a lot of people.

Not only that, but many tags only need to fire on certain pages or certain parts of the site, and it’s not easy to do that.  So, tag management acts like an ad server for vendor tags, and allow both publishers and marketers to control when vendor tags serve, how often they should serve, in what order of priority, and for how long.  In this way the tag management system is the only technology that the IT team actually has to place in the source code, and from there the ad operations group within the company can place or remove whatever they want.

Measurement and Analytics

There are really two distinct sets of companies in this box, and most in the digital space don’t think of Analytics platforms as part of the ad tech industry, even though they often can play a role for marketers.  Omniture, CoreMetrics, and Webtrends are historically the system of record for tracking engagement with a site – what pages are users visiting most, how much time do they spend on site, how do they enter the site?
Where the rubber hits the road with ad technology is typically when a marketer embeds a pixel tracker from their analytics software into their ad tag to create a unique cohort they can analyze.  Omniture can also feed site engagement data back into creative optimization for marketers.  Far more common in the industry is interaction with comScore and Nielsen, which are measurement behemoths with a long heritage in traditional media like TV.  Those companies provide a host of solutions, from broad rankings of publisher reach and frequency to assist marketers in the media buying process, to more specialized services like calculating demographic trends of who saw their ads, especially in the video space.


Ad exchanges, like DSPs and SSPs, are one of the core components of the programmatic advertising space.  Exchanges facilitate transactions and provide liquidity to buyers and sellers, so that everyone doesn’t have to integrate with everyone else, but can simply connect to a handful of large clearing houses that are connected to everyone.  Importantly, exchanges do not sell media and do not try to optimize the auction for the benefit of either the buy or sell side.  The example I typically provide is that DSPs and SSPs are like high frequency trading hedge funds, while the ad exchanges are like the NYSE or NASDAQ marketplaces.


As you might notice, Appnexus sits in its own special category on the LUMAscape that’s somewhere between DSPs and Ad Exchanges.  And it’s true that Appnexus occupies a unique spot in the industry because it serves as an open platform for both buyers and sellers who can use it as a DSP or an SSP, and yet also connects with other DSPs and SSPs to provide agnostic exchange services.  You might say that Google is the only other company that acts this way in the market, but Google occupies quite a number of categories in the market (it’s one of the largest publishers online, for one), and far more than Appnexus plays in.  It seems like LUMA Partners thought it made more sense to carve out Appnexus as its own category than represent it in each category as they did with Google / DoubleClick.
For sake of full disclosure, I work for Appnexus since they acquired Yieldex in April 2015.

DMPs (Data Management Platforms) and Data Aggregators

Outside of Ad Servers, DSPs, SSPs, and Ad Exchanges, the data space is the next most important one right now in the digital space.  Like many categories we’ve covered so far, there is a wide variety of companies represented on the LUMAscape.  The two major services these companies provide is data management software and data products.  The software piece allows both publishers and marketers to tag users (usually via a cookie, but many DMPs are ID agnostic at this point and support any arbitrary way to identify a user) and assign them to specific audience when they demonstrate a particular behavior, like navigating to a particular page or perhaps declare information about themselves through a form.

Data products are cohorts, or segments of users the vendor creates on their own by observing user behavior across a network of sites that host their tag, or by combining multiple 3rd party datasets they might purchase into a proprietary product, which they can then sell to anyone whether they are on their platform or not.  Some, like BlueKai and Lotame, provide both data management platform software as well as data products, while others like Krux, Demdex (now part of Adobe), and X+1 strictly provide the technology.  Some like Exelate (now part of Nielsen) strictly provide data products and do not offer the platform itself as a service.  Still others like Peer39 and Proximic specialize in semantic data vs. behavioral data, and companies like Upfront (now part of Undertone), were really more like smart retargeters and a bit mis-categorized in my own humble opinion.

Data Suppliers

These companies also cover a range, but tend to be the old guard of the data business from direct mail days who have modernized their businesses for the digital age.  You might think of them as data aggregators, but with data that doesn’t come from an online source.  Many distribute their data through 3rd party DMPs, though in recent years many have also built their own in-house DMP technology as well. The value these companies bring is in aggregating, scoring, and normalizing large public and private datasets that cover core demographic traits.  Things like age, car and home ownership, presence of children, political affiliation are sourced from strictly offline datasets and might require tying multiple datasets together to get to the truth.  For example, if you buy a house, a data supplier could know that through public property registration records, which may or may not be digitized. Often, data suppliers have to merge many different sources for a single data point to get national coverage.

The other major dataset these large suppliers have access to is loyalty card data, which they aggregate and anonymize.  If you wanted to target people who are brand loyal to Tide, that data is coming from grocery loyalty cards, which the data suppliers tie to cookies through an offline data match process.  Acxiom, Experian, V12, and SymphonyIRI are in that business, while folks like BlueCava are more specialized in providing identity maps across devices.  They don’t actually own data, but activate others’ data across channels.  For example, if a marketers wanted to target data they had connected to a cookie over in a mobile environment, they could use BlueCava to transfer their cookie-based dataset to a non-cookie identifier in a privacy compliant way.

Ad Networks

There are a LOT of ad network categories on the LUMAscape, and rightfully so as networks still comprise a very large component of the industry, despite how quickly marketers and publishers rushed to embrace RTB channels.  Today, ad networks are programatic players in virtually all cases, so there are lots of blurred lines in this category in particular.

Ad Networks – Horizontal

In the horizontal category it’s all about players that can offer massive reach across a wide range of content.  The big portals like Yahoo!, Microsoft, AOL, and Google are here, as are a smattering of other networks that I could just as easily see in the Rich Media or Performance categories.  This is going to be a tough space to survive in given the competition from programmatic, so I think you’ll see most companies in this space moving toward another bucket; the big players will either become more like SSPs (like Casale), or migrate more toward the performance category (like Tribal Fusion) or rich media category (like Undertone).


  1. Hi Ben,

    I don’t usually leave comments, but I can’t help how worthwhile your articles are.
    I was recently hired into an ad ops role and I’m still in the training stage. Unfortunately, there isn’t much material in place for me to learn except when I directly ask our senior ad ops person and oftentimes, they’re too busy to help me.

    Please don’t stop sharing your knowledge, it means a lot to us.

  2. Thanks Melanie! I’m glad you’ve found the site to be a resource – feel free to ask any questions you might have in the comments section in the future.

  3. Hi Ben,

    Just a quick comment to echo the sentiment Melanie shared. I’ve been a loyal reader of this blog for years and have found it to be incredibly useful and valuable in understanding the Ad Op/Tech side of the business. I appreciate your time and efforts in sharing your experience and knowledge on these topics. Thanks and keep it coming!


  4. Hey Ben,

    Im currently working as an intern in a Video+Display Programmatic Advertising Agency. I have to come up with pricing metrics for the different companies within the industry. I’ve found so little information since then. Any suggestions ?

  5. Hm, what do you mean by ‘pricing metrics’ – like, what the typical rates are for various services, or the going rates for specific types of media, or something else?

Leave a Reply

Your email address will not be published. Required fields are marked *