Given Data And Media's De-Coupling, Should A Data Company Be Selling Media, Too? - Day 4 Reactions

Data-MediaThis is the fourth and final day of the data-media series. If you'd like to begin at the start of the 4-day, reaction series, click here.

Click "read more" to read the participant's answer to the AdExchanger.com reaction question on the decoupling of data and media:

Andy Atherton, COO, Brand.net: “...it’s fine for data companies to sell media as long as they have appropriate legal authority to do so and all relevant parties are aware of potential for conflicts. Practically though, this is rarely the case today.” Read more.

Ari Buchalter, COO, MediaMath: “Regardless of the type of Data and who is selling it, the only thing that is “inappropriate” is being opaque about what you are selling.” Read more.

Zach Coelius, CEO, Triggit:It is really just a business model question about what the advertiser prefers.Read more.

Jeff Hirsch, CEO, AudienceScience: “Data companies will play in the media game.  The questions that arise will be around data rights, and true optimization.” Read more.

Michael Katz, President, InterCLICK: "...its actually not important whether or not a data company sells media, what is important is whether or not buyers actually understand the value of the data that they are buying." Read more.

John Nardone, CEO, [x+1]: “The question that clients will need to answer as they look across the landscape of choices is, is the pairing effective.” Read more.

Andrew Pancer, COO, Media6Degrees: “Companies are trying to offer everything, but by doing so, they're failing to provide the basic services that a more specialized provider can offer.” Read more.

Alan Schanzer, CSO, Undertone Networks: “Companies that acquire and resell or rent data for targeting purposes should remain independent of the media channel.” Read more.

Philip Smolin, GM, Turn: “The question really comes down to what are the core competencies of any given vendor -- Do they understand data, media, and strategy all together, and can they assimilate the data in real-time and at scale to truly maximize its potential?” Read more.


Andy Atherton, Brand.net

Theoretically, I think it’s fine for data companies to sell media as long as they have appropriate legal authority to do so and all relevant parties are aware of potential for conflicts.

Practically though, this is rarely the case today.

Let’s start with the reality that the line between what is a data company and what is a media company has been blurry for some time and certainly isn’t getting any more clear (probably the reason for the question).  For example, is an ad network that sells media based on detailed behavioral data about users a media company or a data company?  Actually, I think this is a great example to help illustrate the concepts at play.

Assume a hypothetical network runs a campaign for an eyeliner product from a large CPG advertiser on a group of womens’ content sites.  The network collects data on which users it saw on those sites or site sections (e.g., beauty tips, product reviews), who clicked on and/or engaged with the eyeliner ad, and possibly some advertiser-specific site activity data.  In the context of this example, I think it would be OK for the network to sell media based on that data – say by a) retargeting those users on other sites or b) offering those users  / look-alike users to other advertisers or c) some combination of both – if they had appropriate legal permission in place from both the publishers and the advertiser to do so.  The activities I outline are commonplace, but the appropriate legal permissions are almost never explicitly granted today.  In fact, I think many advertisers and publishers would be surprised (shocked?) at how their data is being used.  This is a pretty big disconnect if you ask me, about which all ecosystem participants should be concerned.

In addition to data ownership concerns (and simmering privacy issues), there is also the potential for conflict if the data company in question is providing audience or content “verification” in addition to media – the proverbial fox guarding the henhouse.  This issue seems simpler to address.  I would submit that just as long as the media buyer knows that the potential for conflict exists, she can make up her own mind.  If she has serious concerns about conflict, she can simply buy media and data from different sources.

Back To Top

Ari Buchalter, Chief Operating Officer - MediaMath

Let’s start with a definition: “Data” (capital D) is everything that an agency or advertiser can analyze in order to decide how to deploy dollars against a marketing goal, be it awareness, engagement, conversion, ROI, whatever. That Data includes the media, which can be characterized by variables such as the content category, publisher, page URL, placement size, location on the page, session frequency, etc. (some of that media Data is also used as a proxy to describe the audience, typically through panel-based measurements). The Data may also include any third-party information about the individual user, such as demographic, behavioral, social, transactional, etc., typically sourced from data providers such as BlueKai, Exelate, Targus, or many others (which by the way, often get their data from publishers). In this respect, publishers, networks, exchanges, and the “data providers” are really all Data providers (capital D), offering information to enable the buyer to determine whether and what to buy.

So if media is Data, the question then translates to “is it appropriate for a data company to be selling Data?” Why, of course it is! Let the media companies sell (non-PII) information about the user. Let the data companies sell media. As long as they are transparent about the granular elements of the Data, rather than obfuscating them. Virtually all Data has some value, but it tends to be different Data that is of value in different contexts and against different goals. The more inputs there are, the better armed the buyer is to optimize against their goals. As long as the sellers are respecting user privacy and compliant with industry regulations, buyers should be agnostic and absolutely voracious when it comes to Data, of both the media and user variety. Because then you can ask the real question: how do you know what all that Data is really worth to you?

I’m going to skip over the non-trivial issues of integration and execution. In this vapor-filled landscape where most buying platforms/trading desks/DSPs/whatever-you-want-to-call-them cannot actually deliver on their Powerpoint promises, it is absolutely critical to choose a technology partner that can bring disparate sources of media data and user data (not to mention other capabilities like dynamic creative, ad verification, and brand measurement) to the point of decisioning – real-time or otherwise. But let’s pretend that part is easy. You’ve got it all at your fingertips – now what?

Now you need to quantify the value of the Data. You need a world-beating algorithm that can analyze thousands of variables, some numerical, some categorical, many of which are spotty, dynamic, and correlated. And if you’re using it for RTB, you need to do it in milliseconds, billions of times of a day.

The good news is that if you can do that, it works really well. You can take in all the Data from all the sellers, and figure out exactly what elements in what combination drive performance against your goal, and how much to pay for those. You’ll find yourself delighted to be paying high CPMs to “corner the market” on those precious media+user Data combinations, secure in the quantitative knowledge of the value and scale created. Interestingly, it’s often not the obvious Data that matters. Sure, in-market auto consumers drive high response rates for auto ads (and by the way, are also heavily competed for by all the auto advertisers and therefore expensive). But can you also name 26 other, non-automotive, BlueKai segments you’d never have guessed have comparably high response rates for auto ads? Can tell you how those response rates vary when combined with different media characteristics and put a precise value on the user Data, the media Data, and the combination of the two, against a stated GRP, CPC, CPA, or ROI goal? If you can, then you know what we know, and how well it works.

The bad news is that if you can’t do that, you’re at risk of allowing the seller, whether it’s a media company, a data company or someone else, to package up the media and data at your expense. If you’re seduced by words like “social graph,” “search retargeting,” or “look-alikes” and really want to buy that stuff, but can’t put an accurate value on it yourself, then someone is going to assign an inaccurate value to it for you. They are going to bundle it all up into a high CPM product with a sexy name and make you pay through the nose for your ignorance. Heck, they may ask you to place a pixel on your site, and then sell remarketed customers back to you, but under a different, catchier label. And it may even perform decently against your goal, even after the seller has taken out their hefty margins. So much for transparency. Regardless of the type of Data and who is selling it, the only thing that is “inappropriate” is being opaque about what you are selling.

Will buyers take a step forward, by partnering with best-in-class platforms, opening the Data aperture, and using sophisticated algorithms to identify the valuable media and user Data elements that drive scale and performance against their objectives? Or will they take a step back, allowing networks, data providers, or whoever, to re-obfuscate those elements into over-priced, monolithic bundles? Fortunately, the tools to take the step forward exist today.

Back To Top

Zach Coelius, CEO, Triggit

The bundling of data and media makes perfect sense where an advertiser only wants to work with one company. In the past, data companies had to partner with large publishers or ad networks in a painful, cumbersome process to get distribution of their data and where often forced to play a secondary role with the advertiser. Now, demand side platform (DSP) technology and real time bidding makes getting access to media scalable, transparent and simple. A data company can work with a DSP to buy impressions against their data and provide the complete package to the advertiser. It is really just a business model question about what the advertiser prefers.

Back To Top

Jeff Hirsch, CEO, AudienceScience

As an industry, we have taken on the task of breaking up the life of a campaign into many pieces, including data acquisition, data management, media buying, creative generation, inventory access, and verification. In this new world of single point solutions, those managing data have a unique position in that they know more than the rest of the chain about the target audience, and they protect one of the most valuable assets in the space.  As we start putting the pieces back together, re-integrating the buying process, the data companies are in a unique position to act as systems integrators, leveraging the core asset they have access to.

Data companies will play in the media game.  The questions that arise will be around data rights, and true optimization.  Optimization must take into account all of the components of a campaign as a bias toward data ONLY will not allow the full potential of a campaign to be met. Data companies must also resist the pull to leverage data from sources where data rights might be a point of contention.

Back To Top

Michael Katz, President, InterCLICK

First, it depends on what your business model is. If you are creating a free market like BlueKai, you are hoping that there are enough rational buyers, where the information used to purchase data is indicative of the true value of data, and that laws of supply and demand ultimately dictate price accordingly.  Conversely, if your business is delivering audiences to advertisers, you must understand the impact that the data has relative to specific sets of inventory in order to be able to deliver the most responsive audiences.

Second, if you are a buyer of data, how do you buy data without understanding the impact it has on media? Unless you are buying irrationally, the value derived has to be based on running actual campaigns with specific objectives and restrictions, on various sources of inventory. This gives you the true value of the data you are buying.

So, its actually not important whether or not a data company sells media, what is important is whether or not buyers actually understand the value of the data that they are buying.

Back To Top

John Nardone, CEO, [x+1]

Well, it’s not *IN*appropriate.

Let’s look at this question from two perspectives; first that of marketers. The question that clients will need to answer as they look across the landscape of choices is, is the pairing effective? For pure-play, direct response campaigns, where knowledge of the data is a key driver of success, it might be a solid choice. To do this, and to do it well, however, data firms would need to attract top quality media buying talent.

Beyond that, clients will also demand strategic insights and broader perspectives from the data companies, while much of their business is about the quantitative aspects of marketing. Lastly, how likely is it that data firms will be able to scale campaigns beyond their own data set?

I asked Ken Rona, our VP of Analytics & Data Strategy, to think about the question from the data companies' perspective. Here is what Ken had to say: It's really a business model question. On one hand, it is difficult for a data company to unbundle the data it generates, in effect, giving away its competitive advantage. If a data company makes its data available to the market, it enables direct competitors who are going to go take market share. Here, the natural response is to bundle media. On the other hand, no data company has been able to maintain its exclusive use of its data. Once people see that a data source generates incremental lift or can find a specific audience for a brand marketer, someone will figure out how to create a competitive product. So, trying to keep the data proprietary works for a little while, but eventually competitors will enter the market and take share. Therefore, data providers should get the data out there and get first mover advantage. Their natural response in this case is to unbundled media. Back to the business model issue, if the data company insists on coupling data with media, they don’t maximize the value of their data, but receive some extra revenue from the media. Which strategy maximizes value? It depends. Each company needs to value the different business strategies and come up with its own answer.

Back To Top

Andrew Pancer, COO, Media6Degrees

I don't think that this is too big of an issue. If a data company has the ability to sell media then by all means they should be able to do so. There's no issue of conflict of interest with clients. This speaks to the larger trend of streamlining advertising buys.  We're seeing it with demand-side platforms and now with data companies selling media.  Everyone is trying to provide the best ROI for advertisers, while capturing as much of the value chain as possible. Part of the value proposition to advertisers is providing fewer moving parts. It's a good idea in theory, but the issue is execution. Media optimization has different challenges and requires different competencies than what a data company will be typically used to. Some companies will succeed at doing both, but I expect many to fail in their execution. Companies are trying to offer everything, but by doing so, they're failing to provide the basic services that a more specialized provider can offer.

Back To Top

Alan Schanzer, CSO, Undertone Networks

The online advertising landscape is constantly evolving. Today, the marketplace is filled with ad networks, media exchanges, demand-side platforms, media trading desks, verification businesses, and now independent third-party data providers and data exchanges are part of the mix.

As a media professional, I am excited to have access to data that may provide incremental value to advertisers. Third party data can be very valuable and the data exchange model provides an efficient way to access lots of data from multiple sources. In the end, this type of access will provide the testing required to truly understand what data driven targeting can do.

Data capture and data retransmission are very specialized, and potentially highly regulated businesses. Companies that acquire and resell or rent data for targeting purposes should remain independent of the media channel. The separation of data collection and data usage will provide the safest environment for the consumer, and the industry will be best positioned to self-regulate. The most valuable data is often found directly on a publisher site. Media companies that have access to this primary data should continue to find ways to monetize it directly or through partnerships that extend targeting and reach opportunities.

As the industry evolves and regulation looms, it is increasingly critical that data companies maintain transparent business practices. At Undertone, we build partnerships with the highest quality data providers available that will fulfill our client demand. We focus our expertise on using, understanding and collecting data that drives results and adds value. Then, we leverage that expertise to develop in-demand solutions for the marketplace; including products that are designed to help publishers better monetize their own data.

Whatever your stance may be on this topic, it’s going to continue to spark conversation, debate and possibly even regulation in 2010.

Back To Top

Philip Smolin, GM Platform Solutions, Turn

There's no inherent conflict for marketers if a single business entity sells both data and media; traditional media companies have already indicated that they will likely sell data to complement their inventory offerings. What is critical for marketers is receiving the highest level of transparency for both the data and media.

Successful campaigns most often begin with a well-developed campaign strategy that guides the combination data and media. I think technology is key to accomplishing this effectively and efficiently. The question really comes down to what are the core competencies of any given vendor -- Do they understand data, media, and strategy all together, and can they assimilate the data in real-time and at scale to truly maximize its potential? For instance, from an end-to-end DSP, a media buyer can research and build a campaign strategy, and then centrally negotiate their data and media deals themselves from multiple sources. Alternatively, that buyer could depend on a single service provider (media or data provider) to bring both data and media together for them.

Back To Top


Read: Day 1, Day 2, Day 3, Day 4

2 Comments

  1. DSP Jockey

    I think any data company that closes their sales pitch with a media offer pigeonholes themselves as just another network. A network on steroids, really. A true data provider should do just that - provide data that can be portable and implemented across a variety of supply sources.

    Reply

Add a comment

XHTML: You can use these tags: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>