An Open Letter to Lotame et al.
04.14.2011 - Krux Digital
When it comes to market dialogue, Krux prides itself on being an active participant through blog posts, tweets, thought leadership, and editorial contributions to market publications. We welcome dissenting viewpoints. When we see something we disagree with, we go out of our way to avoid sniping from the sidelines out of respect for the author and the reader because we believe that different opinions almost always lead to clearer thinking and stronger solutions.
But there are exceptions that prove every rule. Recently, we came across a post from Lotame regarding incentive conflicts for players who offer both data management platforms and a competitive network or exchange. The article was, sadly, wide of the mark, which is why we feel compelled to address its many logical errors. (You can find the post in its entirety here: “How Pure Is Your Pure Platform?”)
The basic premise of the Lotame piece was that publishers should not be concerned about a platform provider who also runs a marketplace. Rather than introducing incentive conflict or business risk for publishers, Lotame asserts that such a scenario is actually beneficial. By Lotame’s reckoning, fire-testing their technology on their own media and data business enables them to deliver a more robust solution for publisher clients.
With pride, they report:
“At Lotame we eat our own dog food... and it sure is tasty.”
We recognize that Lotame is swinging a cat, and as much as we’d like to play along, we just can’t. When Oracle uses the exact same HR software it sells customers to manage its internal HR function, that’s eating your own dog food. Lotame’s claim is an extreme abuse of the old eat-your-own-dog-food technology dictum precisely because the software required to run a data marketplace is totally different from the software used to manage data.
If you’re running a marketplace, you seek to efficiently match buyers and sellers, typically with a view towards optimizing your own piece of the spread. By comparison, publishers seeking data management seek to earn greater revenue from data for their own bank account. They don’t care who’s on the other side of the transaction or the overall efficiency of the match. Attempting to use the same code for both means that you’re doing neither terribly well, which is why Lotame’s claim doesn’t make sense even on its own terms. It’s akin to the fox telling the farmer he is perfectly suited to guard the henhouse because he, of all creatures, understands how truly delicious chicken can be.
Or if you prefer less agrarian analogies, it’s a bit like a merchant bank asserting that their own first party trading operation gives them an edge in delivering faster and better brokerage services to individual investors. Not even Goldman Sachs, with all their polish and chutzpah, ever tries to mount an argument as nutty as this. What they do is try to conform to an increasingly stiff set of regulations (cf. the Dodd-Frank bill) that prohibit line-crossing between brokerage services and trading operations, regulations that we don't yet have in digital media, but the premise for which is well-understood by savvy publisher principals (and the FTC, for that matter). It’s a safe bet that, as the advertising industry continues to adopt transaction models more like those of financial markets, we’ll see a similar regulatory push.
Analogies work when one thing has something to do with the other. Lotame’s analogies are an interesting exercise in sophistry – argumentation that sounds good but doesn’t make sense when you actually try to unpack it.
"…I'd be hesitant to ever recommend someone use a product offered by a company that they themselves do not use and rely on. Would the iPhone be so successful if Steve Jobs walked around with a Blackberry? Would Salesforce be the force that it is if they internally managed their own relationships using their email address book and spreadsheets?"
Huh? Salesforce is successful because they saw the SaaS opportunity, built a good CRM product, and effectively scaled their operation before the competition did. The lead architects at Salesforce will happily report that they learned 100x more about their product from their first enterprise-scale implementations than through their own internal testing. The iPhone is successful because it’s a great product that the users love, it delivers huge hardware margins, and it’s vertically integrated with Apple’s music and app marketplaces. (Vertical integration is key here – we can guess why Lotame shied away from that part of the Apple story.)
Lotame has its causes and consequences mixed up. The fact that Steve Jobs uses an iPhone is an interesting byproduct of his team’s commitment to building a great product that lots of people want to buy. Salesforce started with the objective to build on-demand CRM software for thousands of companies. The fact that Salesforce and Steve Jobs use their products reflects – but doesn’t cause or explain – their success.
More important, to suggest even indirectly that secondary channels for data monetization pose no risk to publishers doesn’t jibe with the historical record. One need only look at the pricing and yield dynamics introduced by ad networks over the previous decade. Between 2005 and 2008, a number of us on the Krux team observed several leading publishers aggressively shift huge swaths of their inventory (30-40%) into secondary channels. While they saw marginal improvements for their remnant operations, they saw an attendant decline in their direct sales eCPMs (20-30%). This came as the direct result of the market competition their remnant inventory had energized – in particular, the ability of networks to segment, package, and sell inventory to the same advertisers and agencies the publishers were calling on. If Lotame has a clear conception as to why this won’t happen with publisher data, or why their marketplace precludes this from happening via its fundamental design, then it’s on them to shed light as to how.
Few publishers we have spoken to are eager to move blindly into secondary data channels right now (which of course explains the expeditious re-potting of Lotame and other marketplace players into DMPs over the last nine months). Further, they believe in the value of their audiences and don’t need an intermediary to show them that it’s so. Publishers have all the evidence they need in watching others build very healthy data businesses with zero COGS by piggybacking on their decades-long consumer and editorial investments. (See here and here for examples of what we mean.)
Let’s boil this down and look at it in the simplest terms – a single transaction. So Lotame has, and we're paraphrasing here, ‘un-siloed’ the platform from the network. Suspending disbelief for a moment, it is a model that is good for publishers because of the well-tested platform and ready access to demand this provides. Imagine a single piece of demand comes in, looking for data associated with a particular audience segment. Given the inherently closed nature of network-based transaction models, how do you ensure you’re getting the best price for your publisher? In Lotame’s un-siloed scheme, your audience segment has one chance to be monetized – in their marketplace.
Everything we know about microeconomics tells us that fewer bidders for an asset means less value to the seller. Conversely, the more bidders one has for an asset, the higher the value of that asset will become. That kind of vibrant exchange environment is difficult to ensure if you’re working both ends against the middle, presumably managing margins for yourself. From the post in question:
“Before I get started, and in the interest of full disclosure, our company is in the business of providing a data management platform to publishers, advertisers, and just about anyone else that may benefit from one. Our product, Crowd Control, is a powerful online data management system that has served as the infrastructure for our brand focused online media and data business for over four years."
Many exchanges all aim to become the ‘eBay for data,’ where one can be assured of lots of bidders because that’s where most of the supply is already. That eBay itself pulled back from its own plan to be, well, eBay for data, is instructive.
Why? Because they recognize, like many others, that it’s not where the data market is headed. You simply don’t need a central juggernaut of an exchange because there are already dozens of RTB-enabled DSPs and trading desks and agencies who are expressing plenty of demand for data, independent of any one clearinghouse.
With the right level of direct connectivity, publishers can have ready access to as much or as little exchange-driven data demand as their business strategy dictates. In short, they can dispense with middlemen and meet the market on the terms, times, and channels of their choosing.
In closing, let’s consider that dog food line one last time:
“At Lotame we eat our own dog food... and it sure is tasty.”
Lotame started this analogy mess, so don’t mind us while we bring it to its absurd conclusion: If the platform is dog food, their network business is, in that case… a horse farm? Which would mean that publishers are horses, well on their way to becoming dog food apparently.
To all the Lotame publishers out there: if you’re feeling put out to pasture, feel free to give Krux a ring.
A bit about us: Krux delivers a market-neutral platform that helps anyone with a website safeguard, manage, and responsibly monetize consumer data signatures across devices, sources, and formats. Many solution providers’ primary business is the buying and selling of media or data – often in competition with the publishers they serve. Krux doesn’t sell ads. Krux doesn’t run a network or an exchange of any kind. Krux just does data.