Google: “Programmatic” is all about the buy-side

Did anyone else think yesterday’s virtual “hang out” with Google’s Scott Spencer was fascinating?

First of all, where was he?  Was that a jungle or a conference room?

More importantly though (much more importantly), did anyone else find Google’s definition of programmatic a little lopsided?  According to Scott:

  • “We define [programmatic buying] based on the fact that the buyer gets to define the targeting”
  • “The ability to not buy is a critical attribute to programmatic buying”

This is a fair characterization of RTB as it exists today, but defining the entire Programmatic market this way isn’t exactly seller-friendly.  The real eye-opener for me though was the response to a question from the audience on delivery guarantees (good question, Tom!).  Scott made it clear that in the context of Programmatic Reserve, any delivery guarantees should only go one way.  So for Google, a programmatic delivery guarantee means that the seller guarantees they will provide the inventory, but the buyer doesn’t guarantee they will buy it.

Huh?  You have to admire the frankness, but if I was a publisher I’d be very concerned about that perspective.  I think for programmatic reserve to “really happen” it’s going to take a much more balanced approach that addresses both buy-side and sell-side needs.

What do you think?

How do we make Programmatic Reserve really “happen”?

Interesting article last week on Programmatic Reserve by Jay Sears in Ad Age (call it “‘guaranteed” if you want – just don’t call it “premium”).  Jay’s a smart guy and right on the money with respect to the potential for a broader definition of “programmatic” to fundamentally change the online ad landscape and dramatically expand the pie.  So from that angle I was in violent agreement, but two things still kind of grated on me as I read.

First was the shifting tense and implied timing.

Seemed to me that we went freely back and forth between present and future tense as the article unfolded.  We were talking about a ripe opportunity for Programmatic Reserve, then something that was happening now, then sorting out the technology to make it happen.  Those of you who know me (and/or read this page) know this is a topic near and dear to my heart.  It’s most certainly a ripe opportunity.  But it isn’t happening now to any meaningful degree.  Lots of talk.  Maybe almost that much work going on too, but not a lot of real money moving yet.  Perhaps what he means is that this is the year that the work bears fruit and Programmatic Reserve becomes real.  Possible, but we really have our work cut out for us.

Which brings me to my next issue – what needs to happen to make it real?

Jay outlined a vision including structured electronic RFPs on the buy side, structured electronic catalogs on the sell-side, even index funds.  This was a good synthesis of many of the ideas that have been discussed over the years.  So I love the aspirations, but I think the harder part is how to get there.  That’s what folks haven’t been able to figure out yet (yes, that includes me when I was at Brand.net).

Exactly what problem are we trying to solve?  Are we (as this article’s subheader suggests) trying to get up front budgets running through RTB?  Not really.  Are we trying to compete with excel as some others have suggested?   I’m not sure that’s exactly right either.  Is this primarily a buy-side problem or a sell-side problem?  Seems to depend on who you talk to.

I think we’d be most productive as an industry to answer these questions once and for all, rather than spilling more (virtual) ink on the high-level vision for the future. So I am going to do my best this year to do just that.  I.e. figure out what’s the right path to get us to this future we all see and frankly must create to keep growing long term.

I’ll keep you posted, and I’m very much looking forward to your input!

Can we please retire “Premium”?

I spent some time on Wednesday talking with John Ramey over a couple of the delicious finger sandwiches we were serving after our latest AppNexus Summit.  During that conversation, he and I agreed that we hated the term “Programmatic Premium” to describe the process of using technology to enable media buyers to buy, in an efficient way, specific packages of media from specific sellers that guarantee delivery.

He and I are not exactly spectators in this arena, so I think this convergence is worth noting.

Unfortunately, the label “Programmatic Premium” is gaining traction as the broader online ad industry’s attention turns to the simply huge opportunity presented by [whatever we call it].  I think it’s important we get the terminology right before we’re stuck with the current kludge by default.

Here’s the problem: nobody agrees on what half of this unfortunate handle – “premium” – even means.

A recent Digiday article is typical of the debate.  Group of informed grown-ups, simple word, complete lack of convergence on definition.  If we can’t agree on what this term means, we’re not going to make the progress we need to in this area.  As I mentioned in a post a few years back, I think a big part of the problem is that folks are conflating the “quality” dimension and the “reservation” dimension.  These dimensions are correlated, but distinct.  The very fact that we are still having this confusion/discussion 3 years later pretty much proves that this term is a problem.

I believe when folks talk today about “Programmatic Premium”, they are essentially talking about automating the planning and guaranteed allocation process that today most often occurs through a direct sales interaction (reservation dimension).  Indeed, this process often deals in higher quality inventory at higher prices (quality dimension), but that is secondary and using the “premium” term invites unnecessary confusion precisely because of its subjectivity.

This is too important to keep getting confused about.   85% of the money in the display ad ecosystem is still transacted, inefficiently, “the old fashioned way”.  If we use a different term, we can avoid muddying the the water by grafting the irrelevant “what is ‘premium’?” debate onto a critical, urgent conversation about how technology can be used to streamline the process between buy side and sell side for media that cannot, should not or simply will not be traded via RTB.

Can we please all just say “Programmatic Reserve” instead?  “Reserve” focuses on the notion of a delivery guarantee or reservation, while retaining the positive associations with quality and supply-side control (think “reserve wine”).  And it gets us away from the all-too-squishy “premium”.   At least give me “Programmatic Guaranteed”.

Anyone care to second?

Consolidation curve?

An insightful post from investo-blogger Jerry Neumann yesterday on Ad Exchanger.  I like what he’s thinking about in the post and agree with much of it, but there’s an important meta-point that he didn’t mention.

Jerry’s first point was that there is a huge shortage of experienced talent in the online ad industry and what does exists is primarily clustered within the myriad tech vendors in the ecosystem.  Agree.  His second point was that even as the exchange ecosystem (which at its core promises increased efficiency through a common set of pipes) grows, we see continued fragmentation of supply / demand relationships.  Agree.

But I would also argue that these two observations are causally related.  The reason things continue to fragment is largely that there are too many tech companies making too many pitches to too many media buyers and sellers that are still coming up the learning curve.  Tech company convinces still-learning buyer or seller to participate in “private market” promising some advantage in terms of functionality or monetization.  Careful A/B testing is hard to do without committing even more limited time/resources (hence it’s rarely done at all).  Whatever advantage was expected may or (more likely) may not actually be delivered, but such decisions are infrequently revisited.  As a practical matter, once the sale is made the arrangement has tremendous inertia, regardless of relative value add.

So Jerry’s “thin exchange standards” may well become necessary, but I think that would have much more to do with folks not thoughtfully using the tools that already exist rather than a “real” need.

“Private markets” are rarely the most efficient alternative.  The more participants in the market the better, assuming careful thought is given to structure and business rules.  I saw frequent examples of the private market dynamic in my time at Yahoo!.  Some enterprising salesperson would convince a content group GM to dedicate a placement to a particular advertiser.  Such arrangements almost always under-monetized relative to an open, competitive market for the same placement.  There was just an article last week in the ‘Journal offering up some more evidence from Goldman’s experiment with private markets.    Or coming at it from another angle, have you ever tried to sell anything locally on craigslist, failed, then posted on eBay?  eBay’s national market with huge liquidity almost always closes the deal at a fair price.

The faster we collectively get up the learning curve, the faster things will consolidate so we can actually realize some of the efficiency gains we’ve all been chasing.

An Inconvenient Truth

An interesting piece yesterday from Adam Cahill of Hill Holliday, with some great thinking on how to address the quality challenges posed by the evolving real-time digital media landscape.

As Adam correctly points out, for most Brand campaigns delivering results is about more than just protecting a Brand from objectionable content.  That itself is very important (and we’re very good at it, by the way), but it’s only the beginning – “necessary, but not sufficient” as they would say back at MIT.  Media quality involves not just the the text of a page, but the editorial environment in which it exists.  That second bit makes this an even harder technical problem, particularly when you consider that quality is a page-level issue.  So we’re currently left with the false choice between audience and content that Adam correctly suggests we reject.

Let’s push a bit further though.

Without taking too much license (Adam, please feel free to chime in), I think I can safely say that “Audiences vs. content” is essentially a compact way of describing the choice between two different operational approaches.  “Audiences” is shorthand for scalable, efficient, automated buying via RTB on exchanges.  “Content” is shorthand for manual, site-by-site buying.  In the rush for operational efficiency, “audience” buying has grown very quickly over the last 2 years while “content” buying has stagnated, resulting in well-documented challenges for many high-quality publishers.

Audience buying works great when fast, high-resolution feedback on a financial goal metric is possible.

For example, let’s assume Netflix’s goal for a big chunk of its marketing spend is profitable subscriber acquisition and they have conversion value and attribution models that they trust.  Then, just as long as they have a scalable way of keeping their ads away from damaging content (porn, profanity, etc), they can pretty much ignore the editorial quality / “shades of goodness” issue Adam focuses on in his piece.  The tie between editorial quality and performance will show up in the CPA numbers and cause money to move appropriately.   So, for this block of DR money, Netflix can optimize based on their conversion metrics and they’re done.

For a brand campaign, the situation is different.

Brand metrics (e.g., awareness, consideration, intent) take longer to measure, they take longer to translate into financial value and that financial value is most often (95% of the time) realized in an offline transaction.  This means there is no fast, high-resolution feedback on a financial goal metric for Branding, but the push for enhanced efficiency of audience buying is no less acute.  What to do?

Unfortunately, today’s “solution” most often involves substituting for the meaningful data that is lacking some mix of a) meaningless, but conveniently accessible metrics like CTR or b) nice-sounding audience descriptions (like “peanut butter bakers”).   Once these substitutions are made, Brand campaigns can smoothly run through the DR-tuned “audience” infrastructure.  The problem is that these simplifying substitutions require a huge leap of faith at best and are very often detrimental to performance against the metrics that really matter.

The right way to leverage the new real-time online ad infrastructure for Branding is first to carefully test and measure the impact of different scalable, repeatable targeting criteria on *meaningful* metrics (like purchase intent or offline sales).

This process is conceptually similar to the Netflix example I detailed above; i.e., test, measure, optimize.  However, because Brand measurements involve longer time lags and lower resolution, there will need to be some manual effort applied to the process itself before intelligent instructions can be fed into the real-time execution machine.  The machine can’t do all the work itself.

It’s an inconvenient truth, but it’s the truth nonetheless.

Unfortunately, these “meaningful, but harder to get” metrics are too often not even gathered today, so the convenient lie persists.

Reading Adam’s article in this context, the richer standards for quality that he’s calling for essentially represent another set of scalable, repeatable targeting criteria added to the mix, one that he expects to have high correlation with results for brand marketers.  I wholeheartedly agree there would be a lot of value there.  We’ve certainly seen the impact of media quality in our own results.

But I also think it’s important to underscore the higher-level point raised here.  In order for the real-time digital ad infrastructure to be complete, it needs native support for branding that is sadly lacking today.

Data Ownership is the Krux

Just read Tom Chavez’ AdExchanger post on his new company, Krux Digital.

Tom and I go way back.  Tom ran Rapt, an innovative yield management technology company.  As part of my role at Yahoo! in early 2003, I helped Rapt make the transition from the computer hardware market to the online display media market.  This transition ended well, as Rapt’s market leading online media capabilities resulted in an acquisition by Microsoft 5 years later for a figure rumored to be nearly $200M.

Tom’s a smart guy and seems like he’s onto another smart concept here.

As we have written, there are really two separate and distinct burners under the increasingly bubbly cauldron of online media data issues:  privacy and data ownership.  Starting on the data ownership side, as it appears Krux is doing, seems like the right answer.  Strikes me that it’s tough to have a market driven mechanism for managing privacy without any scalable way of enforcing property rights in the relevant data.  Most in the BT industry seem to really want the former, but don’t seem to want to talk about the latter.

Kudos to Tom and Krux for giving publishers the tools to keep the dialog honest.  I wish them luck!

Of Taxes & Chaos…

Interesting blog entry from Eric Porres of Lotame earlier this week, about the various and increasing “taxes” levied on online advertising.  I particularly liked his closing thought on 3rd party verification providers.  In a very early stage market with multiple competing vendors and approaches, the real potential exists for the scenario he outlines – i.e., several different parties to a media transaction winding up with different, conflicting “verification” reports and creating chaos. We have seen examples of this in our own business already on campaigns where 3rd party verification was used.

Let’s also keep in mind this problem only gets worse if the verification market moves (as is claimed to be the goal) from passive verification (reporting) to active prevention (blocking).   It’s one thing to have a set of false positives in a reporting package, creating a media provider / agency discussion that is pure friction for both sides.  It would be an entirely different thing to have “underdelivery” (and thus underpayment) due to those false positives.  What is friction squared?

I tend to agree with Eric that the right answer is for the incumbent ad servers to develop their own standards for verification, much as they have done with impression counting.  Personally, I think this could actually be done as either a feature within the demand-side ad serving platforms (DFA, Atlas) or the exchange platforms (AdX, RM, AdECN) themselves.

There are arguments for both, but I think the exchange platform might actually be the right integration point, particularly if blocking is the ultimate objective.   It seems like this approach would also facilitate additional active services.  On the demand side, perhaps focusing more broadly than just objectionable content.  On the supply side, perhaps helping publishers to avoid instances of “blocking” by preventing problems before they occur.

friction2

CBS’ Decision

Some quick comments on this morning’s Ad Age article on CBS stepping away from networks. The “publishers vs. networks” issue has ebbed and flowed pretty consistently since I started in this business at Yahoo! in 2002.  It seems to ebb when revenue is scarce and flow when demand picks back up, with clear evidence of both trend and seasonality.  There is obviously some rationality to this pattern, but I have always thought that the “turn ‘em on, turn ‘em off” approach is a blunt instrument that doesn’t serve publishers, particularly in the long term.

For example, in this article, CBS draws a distinction between the third party networks they are turning off and agency-owned networks (e.g., Vivaki) with whom they will continue to do business. As Michael Zimbalist of NY Times points out in another recent article, from a publisher perspective these agency-owned entities have a lot in common with third party networks.  So it’s unclear how leaving them “on” makes sense if the best solution for third party networks is “off”.

Apart from this inconsistency, two other big issues with the on/off approach are lack of resolution and poor responsiveness to dynamic market conditions.  While networks overall may monetize at a lower rate than direct sales efforts, certain networks will be more or less competitive for certain inventory (resolution) and at different times (dynamics).  RTB was designed to address these two “hard coding” issues (amongst others), but neither AdX 2.0 nor Right Media are close to ready to be relied upon as sole indirect demand channels.  Internal agency network efforts are still nascent as well.  The bottom line is that vastly more demand still flows through third party networks than through of any of these channels.

So rather than bowing out of a significant majority of the quickly evolving ad ecosystem, I think the right publisher solution is a framework that coordinates direct and indirect sales efforts to create the competition for inventory that drives maximum revenue for the publisher. Based on my long experience at Yahoo!, I laid out the broad strokes of such a framework in an article for MediaPost earlier this year. Publishers that learn fastest and best how to apply such a framework in their particular circumstances will achieve levels of monetization that increasingly distinguish them from their more isolationist peers.

None of us is as smart as all of us; the key to staying on the cutting edge of monetization is coordinating the best efforts of both direct and indirect channels on a dynamic basis.  Today and for the foreseeable future, third party ad networks are an important part of that picture.

In Search of Exchange 3.0

I thought readers of this blog may also be interested in my guest post for Ad Age, where I give a brief history of the evolution of the display advertising exchange ecosystem and suggest what I believe is the next step.  This post for Ad Age follows up on my previous post here.

As always, let me know what you think!

A very smart publisher

I just read a great piece by Michael Zimbalist of the New York Times Company on paidcontent .org. He clearly has a deep and precise understanding of the substantive issues at play in the channel conflict debate.  The more other major publishers and the market at large understand the distinctions he’s helping to clarify here, the better off we’ll all be. We’ve certainly done our part to help clarify terminology and to help publishers prevent channel conflict, so it’s great to see smart publishers joining that effort.