I See Dead People… Responding to my Marketing!

ImageI was recently at a Digital Marketing Association awards dinner where data legend Charles Stryker was being honored. After accepting his award, he told a famous story about data that every digital marketer should know.

A long time ago, the US Postal Service discovered they were paying a ton of money to deliver mail to deceased people. Charles was hired to help them get a handle on their records and create a sort of “Do Not Mail” list. Part of doing the work involved considerable A/B testing to ensure he was making the correct assumptions about the data. Direct response mailers were being sent to groups of dead people and similar groups of folks who were still alive. Something astonishing happened when the results came in.

The dead people responded at nearly twice the rate of the living.

Of course everyone at the dinner (about a hundred senior direct marketing executives) laughed uproariously. They have seen all kinds of unpredictable results with direct marketing. In today’s digital age, marketing is moving faster than ever. The velocity of data is increasing in orders of magnitude, and attribution is going to get even trickier.

What happened with the “dead” people was pretty interesting. It turns out that the successful mailers went to households where the husband of the family died, and the elderly spouses were taking great care to go through and read the mail of their deceased partners. The wives wanted to make sure there was nothing important in those letters—and probably were connecting with their husbands through that simple, daily task. Those widows made a great mailing list “select” since they actually opened and read the mail!

In today’s digital marketing, where we seem increasingly dependent on algorithms and attribution models for targeting and measurement, I wonder if we are too deep in the weeds. Are we forgetting the real, human element of marketing? Do we really understand how success and failure happen with our campaigns? At a recent iMedia agency conference, a lot of the talk was about trying not to forget that advertising is first and foremost about storytelling. Leading with emotion is so important. The marketer has to make an emotional connection with his audience and get them to care.

That struck me when watching a joint town hall workshop with Google and Kellogg’s about dynamic creative. Yes, changing the background color or “call to action” on a 300×250 ad in real time can bump the lift of a campaign incrementally—but are we tweaking a  broken process?

Can we really tell great stories on standardized banner ads?

With the rapid rise of programmatic, a lot of platforms and data companies are fully committed to a standardized industry where scale is king. Display, video, and mobile—biddable and accessible at full scale—is the mandate. Kellogg’s wants inexpensive access to a large electronic palette where Frosted Flakes ads can be constantly tweaked to get optimal performance. Nothing wrong with that.  American Express recently announced an “100%“programmatic” initiative for digital marketing. Why not? Both companies spend tons of money on TV, and optimizing the bottom of the funnel makes complete and utter sense.

But that’s all we are talking about: Optimizing the bottom of the funnel with standardized ads. Sorry, but we are not creating new customers with dynamic 300×250 ads that get a .05% click-through rate. If you are in this business, working for a venture backed startup or newly public adtech company whose value proposition is around driving audience targeting at scale, then you are not “creating stories” online.

As an industry, we need to create digital campaigns that get people to “open the mail.” This is incredibly hard to do with standard display banners, today’s woeful “native” executions, and interruptive social ads. Video has promise, but scale still eludes marketers, and low video completion rates erode available reach considerably.

So, how do you leverage programmatic technology to get great creative out at scale? The only real answer is to automate the workflow behind securing premium inventory and custom programs. That’s where the promise of programmatic direct comes in. Marketers want great ideas from publishers, access to the best inventory they have, and non-standardized units. They just don’t want to pay 10% of their media budgets for planners to cut and paste data into spreadsheets.

Innovation in the space is not just limited to programmatic direct companies with API connections into the publisher side (iSocket, ShinyAds, AdSlot) and workflow automation players (Centro, MediaOcean, and Bionic Advertising)—but also includes RTB players like Rubicon and MediaMath who are building new automation capabilities to augment their RTB stacks. In other words, it’s all about automating the deal right now.

Do they want access to evergreen programmatic campaigns that drive their most likely customers through the bottom of the sales funnel? Of course, and that’s a great job for programmatic RTB. But it does not, cannot, and will never replace the kind of media you can secure through a guaranteed transaction. Also, speaking of dead people responding better—that kind of sounds familiar. In programmatic RTB, some of the best click-through rates come from the dead—fake visitors created by robots.

That said, I think more and more digital advertising will go programmatic, and that programmatic RTB will command the lion’s share of performance budgets. But, when it comes to building brands, bringing automation to the process of securing quality inventory will win.

[This post originally appeared in AdExchanger on 5.13.14]

The Battle for Workflow Automation: What’s Next for “Programmatic Direct”

ImageEven though programmatic RTB has seen the lion’s share of venture capital funding and an enormous amount of innovation, RTB buying only accounts for 20%-30% of all digital media dollars. The real money still flows through the direct buying process, with agencies spending up to 400 hours and $50,000 to create the typical campaign, and publishers burning through 1,600 hours a month and 18% of their revenue responding to RFPs. What a mess….and an opportunity.

Everybody’s battling for a slice of that direct sales pie, and the game is all about helping buyers and sellers automate the manual processes that drive almost 80% of transactional value.

The Holy Grail for both sides is a web based, connected platform that will enable planners and sellers to thrust aside Excel, and start to transact business in the cloud. Although a number of companies have tried and failed to deliver on the promise of workflow automation, the time seems ripe for true adoption, as agencies are being challenged by their clients to create the same programmatic efficiencies across all media channels that they have embraced with RTB. As we speak, winners and losers are being selected, so let’s look at the landscape.

When you look at all of the companies providing a slice of the end-to-end workflow just in digital media execution, it’s hard to imagine that there can be “one system to rule them all” or a true “OS” for digital media. Yet, the dream is just that: An end-to-end comprehensive “stack” that handles media from research through to billing, and eliminates the many manual tasks and man hours involved in connecting the dots. But what are the realities? Let’s saddle up this unicorn and take a ride:

The End of the End-to-End Stack?

The notion of a single end-to-end “stack” for the digital marketer is a tough vision to execute upon. Build a system that has every little feature that a huge agency needs and you have effectively built something no one else can use. The flip side is building something so standardized that individual organizations find little value in it. The “operating systems” of the future that will win should enable agencies and marketers to leverage a standard operating system, but customize it with their own pricing, performance, and vendor data. This enables the efficiency of standardization while enabling data to provide the “secret sauce” that media shops need to justify their fees.  More importantly, the modern operating system for media must be extensible, to allow for a wide variety of point solutions to integrate seamlessly. The right system will certainly eliminate a few logins, but must not limit the numbers of tools that can be accessed through it. That concept necessitates a highly modern, scalable, API-driven, web-based platform. It will be interesting to see how today’s legacy systems (which are exactly the opposite of what I have described) adapt.

Hegemon Your Bets

Several years ago, I wrote that the merger between Mediabank and Donovan may actually be a good thing—provided it offered more choice, flexibility, and open standards. Looking some three years later, I am not sure agencies have any more of that today. Like any other near monopoly, Mediaocean has a disincentive to open up its ecosystem because it invites competition. So time will tell whether their nascent “Connect” effort will become a way for agencies to quickly consolidate their “stack” around a flexible operating system—or if it’s just an integration tax for vendors (a revenue strategy quickly becoming known as the “Lumascrape”). After an IPO, the company will face enormous quarterly pressure for growth. It will be hard to raise prices on already stretched agencies, so publishers will be in the crosshairs. I smell “marketplace” and some monetization strategies around “programmatic direct” enablement for guaranteed media. And what about open standards? Despite years of work by the IAB, the standards and protocols for creating electronic ordering and invoicing are still very much in flux.

Connecting the Dots

More than anything else, the most exciting thing happening in digital media is seeing real programmatic connections between buyers and sellers for guaranteed media. After so much innovation in programmatic RTB (hundreds of vendors, billions in venture capital), we now have some amazing pipes that impressions can flow through. Unfortunately, this has largely been limited to lower classes of inventory and focused almost exclusively on the DR space. Creating the same programmatic efficiencies for “premium” brand-safe inventory is now starting to happen. Whether it comes from new “programmatic direct” pure play technologies, or happens through the RTB pipes, it will not happen successfully without transparency. That means giving publishers control over their inventory, pricing, and what demand partners can access their marketplaces. Will these connections thrive? Not if vendors charge network-like fees, arbitrage media, or don’t provide transparency. Will the endemic fraud in programmatic RTB push more transactions outside the RTB pipes? I think so, and a lot of publishers (see Yahoo/AOL/Microsoft deal) are betting that there are better ways for buyers to access their inventory.

Time for Real Time

Look at all the RTB players who want a piece of the guaranteed action. Three of them (Rubicon, Appnexus, and Pubmatic) will IPO soon, and be under tremendous pressure to increase revenue, margins, and continue to innovate and find new markets. When international expansion stops providing double-digit growth increases, then it’s time to look toward new streams of demand generation—namely, the 80% of deals not currently flowing through their pipes. Those pipes have been engineered for real-time bidding, but guaranteed deals are neither real-time nor bidded. Can they innovate fast enough to provide real value between buyers and sellers? Can they apply years of innovation in DSP and SSP tech to the more prosaic problem of workflow automation? Probably, but there are still business model issues to work out. Most of these companies have put a stake in the ground for either publishers or marketers, and a transactional platform must be agnostic to sit in the middle. It will be interesting to see how new offerings are received in the marketplace.

As the Chinese curse says, “may you live in interesting times.” Indeed, the past several years of ad tech has been nothing but interesting, but the real action is just starting—and it’s taking place in what was the most uninteresting field of workflow automation.

[This post originally appeared in AdExchanger on 3.12.14]

Complexity is the Digital Agency’s Best Friend

Agencies are afraid of change, but change always happens. Is your manual workflow a "red stapler?"

Agencies are afraid of change, but change always happens. Is your manual workflow a “red stapler?”

But Solving the Right Problems are the Key to the Future

I once heard Terence Kawaja remark that “complexity is the agency’s best friend.” It’s hard to argue with that. Early digital agencies were necessary because doing things like running e-mail campaigns, building websites, and buying banner ads were really complicated. You needed nerdy guys who knew how to write HTML and understood what “Atlas” did. Companies like Operative grew admirable services businesses that took advantage of the fact that trafficking banner ads really sucked, and large publishers couldn’t be bothered to build those capabilities internally. The early days were great times for digital agencies. They were solving real problems.

Fast forward 13 years. Digital agencies are still thriving, mostly by unpacking other types of complexity. “Social media experts” were created to consult marketers on the new social marketing channel, “trading desks” launched to leverage the explosion of incomprehensible RTB systems, and terms like “paid, owned, and earned” were coined to complexify digital options. It’s hard being a marketer. So much easier to hand the digital keys over to an agency, and have them figure it all out.

Some of that complexity is dying, though.

Have you ever done any advertising on Google? It’s not that hard. You can get pretty good at search engine marketing quickly, and it doesn’t take anything more than common sense, an internet connection, and a credit card to start. Facebook advertising? Also dead easy. Facebook’s self-service platform is so intuitive that even the most hopeless Luddite can target to levels of granularity so minute that you can use it to reach a single individual. Today’s platforms leverage data and offer great user interfaces and user experience mechanisms to make the complex simple.

This has created the OpenTable effect. Remember when you had to call 8 different restaurants to get a Valentine’s Day reservation? What a pain in the ass. I used to always get to it late, and usually spend a few hours getting rejected before finding a table somewhere. Today, I log into OpenTable, type in “11743” and see all the available 8:30 reservations for two in Huntington. A few clicks, and I am locked in. Would I ever go back to doing it the old way? Sure, why not? Call my beeper if you need me. Please “911” me if it’s important.

So, with all of this innovation making the complex simple, and all of these platforms democratizing access to advertising inventory, analytics, and reporting, why are digital agencies still making a living off of the lowly banner ad? Is there a good business left in planning and buying digital display media?

Programmatic RTB is coming on strong, now representing the way almost a quarter of banner inventory is purchased. That’s a good thing. Platforms like Rubicon Project and Appnexus are making it easy to build great businesses on top of their complicated infrastructure. Marketers can hire an agency to trade for them, or maybe just build their own little team of smart people who can leverage technology. That seems to be happening more and more, making managing RTB either a specialist’s game, or not an option for the independent agency.

Really complicated, multi-channel, tentpole campaigns and sponsorships can never be automated. They represent about 5% of overall display spend, and that’s really where a digital agency’s firepower can be leveraged: the intersection of creativity and technology. That sector of digital involves a lot of what’s being called “native” today. Working with content owners and marketers to build great, branded experiences across the Web is where the smartest agencies should be right now.

How about the rest of the money spend on digital display—the 70% of money that goes through the transactional RFP space? A lot of agencies are still making their money buying reserved media, trafficking ad tags, and doing the dreaded billing and reconciliation. Marketers who pay on a cost-plus basis are starting to wonder whether spending money to have expensive agency personnel create and compare spreadsheets all day long is a good use of their money. Agencies that do not get paid for such work are seeing their margins shrink considerably, as they grind away money paying for low value tasks like ad operations. Clients don’t care how long it took you to get the click tag working on their 728×90. Just saying.

A lot of this viscosity within the guaranteed space is being solved by great “programmatic direct” technologies. Right now, you can use web-based systems to plan complex campaigns without using Excel or e-mail, and you can leverage web-based tools to buy premium inventory directly from great publishers—the kind of stuff not found inside RTB systems. Protocols and standards are being written that will, in a few short months, make the electronic IO a reality. Systems are being built with APIs that can enable trafficking to go away completely. Yes, you heard me. People should not have to ever touch JavaScript tags. That’s work for machines.

This future (“programmatic direct”) has been coming for a long time, but it is still met with resistance by agencies, some of whom are continue to benefit from complexity—and others who are (rightfully) scared of change and what it means for their business. Looking at legacy workflow systems, you wonder why they are so hesitant to leave them, but the cost of switching to new systems is high in terms of emotion and workplace disruption—and previous attempts to “simplify” agencies’ lives didn’t really work out that way.

So, how can digital agencies start to change, and embrace the new world of programmatic direct tools, so they can turn their energy to strategy and client care, rather than be an “expert” in processes that will eventually die?

Part of that is learning to recognize if you have a “wizard” on staff. The Wizard is the guy that has truly embraced complexity within the agency. He is the “systems guy” who knows how to pull complicated reports out of legacy workflow platforms. He probably knows who to write the occasional SQL query, and he knows where all the bodies (spreadsheets) are buried. When a web-based technology salesperson comes calling on the agency, and shows the CEO or VP of Media what web-based programmatic direct buying looks like, they are showing an agency a world where a lot of complexity is suddenly made simple. That demo shows the future of digital media buying: a directory-driven, centralized, web-based method of planning, buying, and serving inventory. Just like search! C-level agency executives and media people want it. They want their employees focused on strategy and analytics…not ad trafficking. But to get it, they invariably tell you to go see the Wizard. “Fred is our ‘systems guy.’ He’ll know whether this can work for us from a technical standpoint.”

That’s when innovation dies. Fred, the Wizard of the legacy systems, will shut down any innovation that comes his way. Complexity is Fred’s best friend. When you are the only guy that can pull a SQL query from your data warehouse, or reconcile numbers between SAP and your agency’s order management system, then you are a God. Fred is God…and he doesn’t want a downgrade. Complexity is the reason great digital agencies were built, and continue to thrive. Tomorrow’s big challenges are going to come from complexities in cross-channel delivery and attribution, and keeping up with new platforms that are delivering amazing native marketing opportunities, not being the next at reconciling ad delivery numbers from servers.

When innovation comes knocking on your door, don’t let Fred answer it.

[This post was originally published in AdExchanger on 6.3.13]

Does Driving Efficiency Drive Profit (A Contrarian View of RTB)

unicorn

Online display would be like this, if branding metrics took profit into account.

I’ve always loved the notion of programmatic RTB. As a data hound and an early adopter of Appnexus , the notion that advertisers can achieve highly granular levels of targeting and utilize algorithms to impact performance is right in my wheelhouse. Today’s ad tech, replete with 300 companies that enable data-driven audience segmentation, targeting, and analytics is testament to the efficiency of buying ads one impression at a time.

But what if driving efficiency in display actually does more harm than good?

Today’s RTB practitioners have become extremely relentless in pursuit of the perfect audience. It starts with retargeting, which uses first party data to serve ads only to people who are already deeply within the customer funnel. No waste there. The next tactic is to target behavioral “intenders” who, according to their cookies, have done everything BUT purchase something. Guess what? If I have searched 4 times in the last three hours for a flight from JFK to SFO, eventually you will get last view attribution for my ticket purchase if you serve me enough ads. Next? Finding “lookalike” audiences that closely resemble past purchasers. Data companies, each of whom sell a variety of segments that can be mixed to create a 35 year-old suburban woman, do a great job of delivering audiences with a predilection to purchase.

But what if we are serving ads to people that are already going to buy? Is efficiency really driving new sales, or are we just helping marketers save money on marketing?

It seems like online display wants to be more and more like television. Television is simple to buy, it works, and it drives tons of top funnel awareness that leads to bottom funnel results. We know branding works, and even those who didn’t necessarily believe in online branding need look no further than Facebook for proof. With their Datalogix offline data partnership, Facebook conclusively proved that people exposed to lots of Facebook ads tended to grab more items off of store shelves. It just makes sense. So why are we frequency capping audiences at 3—or even 10? I can’t remember the last time I watched network television and didn’t see the same car ad about 20 times.

The other thing that RTB misses out on is profit. RTB drives advertising towards lowering the overall cost of media needed to drive a sale. Even if today’s attribution models were capable of taking into account all of the top-funnel activity that eventually creates an online shopping cart purchase (a ludicrous notion), we are still just measuring those things that are measureable. TV ads, billboard ads, and word of mouth never get online credit—yet I believe they drive most of the online sales. Sorry, but I believe the RTB industry creates attribution models that favor RTB buying. Shocking, I know.

So, what is true performance and what really drives it? For most businesses, performance is more profit. In other words, the notion that a sales territory that has 100 sales a day can generate 120 sales a day. That’s called profit optimization. If I can use advertising to create those additional 20 sales, and still make a profit after expenses, than that’s a winner. RTB makes it cheaper to get the 100 sales you already have, but doesn’t necessarily get the next twenty. Getting the next batch of customers requires spending more on media, and driving more top-funnel activity.

The other thing RTB tends to fumble is how real life sales actually happen. Sure, audience buying knows what type of audience tends to buy, and where to find them online, but misses with frequency capping and a lack of contextual relevance. Let me explain. In real life, people live in neighborhoods. The houses in those neighborhoods are roughly the same price, the kids go to the same school district, the people have similar jobs, and their kids do similar activities and play the same sports. The Smiths drive similar cars to the Joneses, they eat at the same restaurants, and shop at the same stores. If the Smiths get a new BMW, then it’s likely the Joneses will keep up with a new Audi or Lexus in the near future. When neighbors get together, they ask each other what they did on February Break, and they get their vacation ideas from each other. That’s how life works.

What media most closely supports this real-life model, where we are influenced most by our neighbors?  Is it serving the Jones family a few carefully selected banners on cheap exchange inventory, which is highly targeted and cost effective? Or is it jamming the Smiths and Joneses with top-funnel brand impressions across the web? The latter not only gets Smith, the BMW owner, to keep his car top-of-mind and be more likely to recommend it—but also predisposes Jones to regard his neighbor’s vehicle in a more desirable light. That takes a lot of impressions of various types of media. You can’t do that and remain efficient. The thing is—you can do that and create incremental profit.

Isn’t that what marketers really want?

[This post originally appeared in AdExchanger on 5/20/13]

Is Programmatic Premium?

Will "buy it now" buttons control display media?

Will “buy it now” buttons control display media?

As the bloated Display LUMAscape shifts, more and more companies focused on real time bidding are turning their venture-funded ships in the direction of “programmatic premium” and trying to pivot towards an area where nearly 80% of display media budgets are spent. This has been called the “Sutton Pivot,” referring to the notion of robbing banks, because “that’s where the money is.”

The fact that that 80%—over $6 billion—is largely transacted using e-mail, Microsoft Excel, and fax machines is staggering in a world in which Facebook is becoming passé. The larger question is whether or not publishers are going to enable truly premium inventory to be purchased in a way that lessens their control. At a recent industry conference, publishers including Gannett and Turner completely rejected RTB and “programmatic” notions. In a world of ever growing inventory, the premium stuff is ever shrinking as a percentage—and that means scarcity, which is the publisher’s best friend. Selling less of a higher margin product is business 101.

As I wrote recently, at the same conference, Forbes’ Meredith Levien laid out the three principle chunks of inventory a super-premium publisher controls, and I want to examine the programmatic premium notion against each of these:

  • Super Premium: Big publishers love big “tent pole” branding campaigns, and are busy building mini-agencies within their sales groups, which bring together custom sponsorship packages that go beyond IAB standard banners. A big tent pole effort might involve a homepage takeover, custom rich media units, a dedicated video player, and branded social elements within a site. While some of the display elements within such a campaign can be purchased through a buying platform, this type of complex sale will never scale with technology, and is the very antithesis of “programmatic.” For many publishers, this type of sale may comprise up to 50% of their revenue. Today’s existing buying and selling platforms will be hard pressed to bring “programmatic” efficiencies here.
  • Transactional: Many super-premium (and most premium) publishers spend a lot of their time in the RFP mill, churning out 10 proposals and winning 2   or 3 of them. This “transactional RFP” business is begging for reform, and great companies like AdSlot, iSocket, Operative, and ShinyAds are starting to offer ways to make selling premium inventory such as this as programmatic as possible. Companies such as Centro, Facilitate, MediaOcean, and NextMark (disclosure: my company) are starting to offer ways to make discovering and buying premium inventory such as this as programmatic as possible. Much of the RFP process is driven by advertisers looking for information that doesn’t need to be offered by a human being: How much inventory do you have, when do you have it, and how much does it cost? This information is being increasingly found within platforms—which also enable, via tight pub-side ad server integrations, the ability to “buy it now.” 100% of this business will eventually happen programmatically. Whether or not today’s big RTB players can pivot their demand- and supply-side technologies to handle this distinct type of transaction (not very “real time” and not very “bidded”) remains to be seen.
  • Programmatic: There will always be a place for programmatic buying in display—and there has to be, with the sheer amount of inventory available. Let’s face it: the reason the LUMAscape is so crowded is that it takes a LOT of technology to find the “premium” needle in a haystack that consists of over 5 trillion impressions per month. If the super-premium inventory publishers have to sell is spoken for, and the “transactional” premium inventory publishers sell is increasingly going to other (non-RTB) platforms, then it follows that there is very little “premium” inventory available to be bought in the programmatic channel.

The middle layer—deals that are currently being done via the RFP process, is where “programmatic premium” is going to take place. In this type of buy, a demand-side platform will create efficiencies that eliminate the cutting and pasting of Excel and faxing and e-mailing of document-based orders, and a supply-side platform will help publishers expose their premium inventory to buyers with pricing and availability details. That sort of system sounds more like a “systematic guaranteed” platform for premium inventory.

So, is programmatic premium? Not the type of programmatic buying happening today.

[This post originally appeared in ClickZ on 2/18/2013]

Programmatic Premium is not about Bidding

bid-nobid-article “A market is never saturated with a good product, but it is quickly saturated with a bad one” – Henry Ford

When it comes to digital publishing sales, it seems like many publishers are questioning whether the product they have—the standard banner ad—is what they should be selling. Last month, I wrote that 2013 would be the year of “premium programmatic,” where LUMA map companies who make their living in real time bidding turn towards the guaranteed space, where 80% of digital marketing dollars are being spent. My recent experience at Digiday Exchange Summit convinced me that this meme continues—with an important distinction: “premium programmatic” is not about bidding on  quality inventory through exchanges. Rather, it is about using technology to enable premium guaranteed buys at scale. Here is what I heard:

The Era of the Transactional RFP is Over

Forbes’ Meredith Levien currently gets 10% of her display revenue from programmatic buying, up from 2% in 2011. The rest of her revenue is comprised of 45% premium programs , and 45% from what she calls the “transactional RFP” business. The latter is the type that comes from continually responding to agency RFPs for standard IAB banner programs, with little customization. Levien questioned whether that type of transactional business was completely on its way to becoming driven exclusively by technology.

Are publishers really going to be able to abandon the  relentless RFP treadmill where countless hours are spent reacting to agency RFPs—many of which are sent to over 100 publishers, despite the fact that an average of 5 find themselves on the campaign? In order for that to happen, Levien said, the very language we are using must change. The language of the transactional RFP (“GRP,” “CPM,” “Impressions”) must change to the language of premium (“Social Shares,” “Influence,” and “Engagement”). Ultimately, Levien sees a world where there are fewer people managing  RFP response and more multi-disciplinary teams that create super premium tentpole programs for large brands. Forbes’ teams feature copywriters, developers, and creatives who don’t talk about the “buy details” of a campaign, but more about the social and cultural implications a great advertising program can create.

To paraphrase Federated’s CEO Deanna Brown, publishers “really have to question whether sending 100 e-mails to win a $50,000 RFP is worth it.”

Create Scarcity

Gannett’s Steve Ahlberg was even more forceful in his rejection of programmatic buying, and the transactional nature of guaranteed buying. After living in a world of their own creation (pages festooned Nascar-like with low-CPM banners) USAToday.com took the draconian move of removing all below-the-fold ads from its site, stripped every network and exchange tag from its pages, and decided to have one large ad placement per page. The experiment—revenue neutral in the 4th quarter of 2012—has thus far proven that publishers can get off “set it and forget it” SSP revenue by creating the type of scarcity that drives up both rates and demand. According to Ahlberg, the publication is talking to quality brand clients that were not on the radar just months before.

Part of the equation is getting away from standardized IAB units and trying to create a “television-like” experience for brand advertisers. Like Forbes, that means getting RFP response teams away from transactional duties, and leveraging cross-disciplinary teams that think like wealth managers, rather than salespeople. Instead of asking about reach and frequency, USAToday.com asks brands what they want to accomplish, and works with them to craft campaigns that work towards a different set of KPIs.

“If the Premium Publishers’ Product is the Banner Ad, then they are in Trouble”

For Walker Jacobs, who oversees Turner’s digital ad sales, the recent leaps and bounds in programmatic technology has done nothing but “accelerate the bifurcation of the ecosystem” which is divided been the good inventory and the bad. Like Gannett, Turner takes a jaundiced view of the programmatic ad economy.  “Our RTB strategy is ‘no’,” as Jacobs concisely put it.

Going further, Jacobs suggested that “there is no such thing as programmatic premium” in a world saturated with banner ad units, many of which go unseen. Standard banners, therefore, are a “flawed currency.”  It is hard to argue with Jacobs in a world that sees 5 trillion impressions every month.

It is clear that, despite the massive strides being made in programmatic buying technology, there is a very large gap between publishers who control super premium inventory, and those that do not. Publishers in the former want to find more streamlined and efficient ways to respond to RFPs, and ultimately turn more of their efforts into selling creative, multi-tiered, tentpole solutions to major brands. It doesn’t sound like many premium publishers are implementing private exchanges either, despite all of the hype in 2012. As Dan Mosher of Brightroll Exchange remarked, a private exchange “is just a blocklist feature of a larger platform.”

In 2013, it seems like premium publishers are not embracing private exchanges, not because of the technology, but rather because they are rejecting the notion of commoditization of their inventory in general.  For most premium publishers, there are the types of sales: Super-premium programs, that will continue to be handled primarily by their direct sales force; “transactional RFP” business for standard IAB display units, which most see being streamlined by “systemic” reserved platform technologies; and programmatic RTB sales of lower class inventory.

So, is 2013 the year of “programmatic premium?” Yes—but only if that means that publishers embrace technologies that help them streamline the way they hand-sell their top-tier inventory.

[This post originally appearded in the EConsultancy blog on 2/5/13]

Managing Data in [real] Real-Time

A Conversation with Srini Srinivasan, Founder and VP Operations of Aerospike

Even today, the notion that a consumer can go to a website, be identified, trigger a live auction involving as many as a dozen or more advertisers, and be served an ad in real-time, seems like a marvel of technology. It takes a tremendous amount of hardware and, even more than ever, a tremendous amount of lightning-fast software to accomplish. What has been driving the trend towards ever faster computing within ad technology are new no-SQL database technologies, specifically designed to read and write data in millisecond frameworks. We talked with one of the creators of this evolving type of database software, who has been quietly powering companies including BlueKai, AppNexus, and [x+1], and got his perspective on data science, what “real time” really means, and “the cloud.”

Data is growing exponentially, and becoming easier and cheaper to store and access. Does more data always equal more results for marketers?

Srini Srinivasan: Big Data is data that cannot be managed by traditional relational databases because it is unstructured or semi-structured and the most important big data is hot data, data you can act on it in real-time. It’s not so much the size of the data but rather the rate at which data is changing. It is about the ability to adapt applications to react to the fast changes in large amounts of data that are happening constantly on the Web.

Let’s consider a consumer who is visiting a Web page, or buying something online, or viewing an ad. The data associated with each of these interactions is small. However, when these interactions are multiplied by the millions of people online at any moment, they generate a huge amount of data. AppNexus, which uses our Aerospike NoSQL database to power its real-time bidding platform, handles more than 30 billion transactions per day.

The other aspect is that real-time online consumer data has a very short half life. It is extremely valuable the moment it arrives, but as the consumer continues to move around the Web it quickly loses relevance. In short, if you can’t act on it in real-time, it’s not that useful. That is why our customers demand a database that handles reads and writes in milliseconds with sub-millisecond latency.

Let me give you a couple examples. [x+1] uses our database to analyze thousands of attributes and return a response within 4 milliseconds. LiveRail uses our database to reliably handle 200,000 transactions per second (TPS) while making data accessible within 5 milliseconds at least 99% of the time.

This leads into the last dimension, which is predictable high performance. Because so much of consumer-driven big data loses value almost immediately, downtime is not an option. Moreover, a 5-millisecond response has to be consistent, whether a marketing platform is processing 50,000 TPS or 300,000 TPS.

What are some of the meta-trends you see that is making data management easier (standardization around a platform such as Hadoop? The emergence of No-SQL systems? The accessibility of cloud-hosting?

SS: Today, with consumers engaged more with Web applications, social media sites like Facebook, and mobile devices, marketers need to do a tremendous amount of analysis against data to make sure that they are drawing the right conclusions. They need data management platforms that can absorb terabytes of data—structured and unstructured—while enabling more flexible queries on flexible schema.

In my opinion, classical data systems have completely failed to meet these needs over the last 10 years. That is why we are seeing an explosion of new products, so called NoSQL databases that work on individual use cases. Going forward, I think we’ll see a consolidation as databases and other data management platforms extend their capabilities to handle multiple use cases. There will still be batch analysis platforms like Hadoop, real-time transactional systems, and some databases like Aerospike that combine the two. Additionally, there will be a role for a few special-purpose platforms, just like in the old days we had OLTP, OLAP and special purpose platforms like IBM IMS. However, you won’t see 10 different types of systems trying to solve different pieces of the puzzle.

The fact is we are beginning to see the creation of a whole new market to address the question, “How do you produce insights and do so at scale?”

One of the biggest challenges for marketers has been that useful data is often in silos and not shared. What are some of the new techniques and technologies making data collection and integration easier and more accessible for today’s marketer?

SS: Many of our customers are in the ad-tech space, which is generally at the front-end of technology trends adopted by the broader marketing sector. We are just beginning to see a new trend among some of these customers, who are using Aerospike as a streaming database. They are eliminating the ETL (extract, transformation, load) process. By removing the multi-stage processing pipeline, these companies are making big data usable, faster than ever.

The ability to achieve real-time speed at Web-scale, is making it possible to rethink how companies approach processing their data. Traditional relational databases haven’t provided this speed at scale. However, new technology developments in clustering and SSD optimization are enabling much greater amounts of data to be stored in a cluster—and for that data to be processed in milliseconds.

This is just one new way that real-time is changing how marketers capitalize on their big data. I think we’ll continue to see other innovative new approaches that we wouldn’t have imagined just a couple years ago.

Storing lots of data and making it accessible quickly requires lots of expensive hardware and database software. The trend has been rapidly shifting from legacy models (hosted Oracle or Neteeza solutions) to cloud-based hosting through Rackspace or Amazon, among others. Open source database software solutions such as Hadoop are also shifting the paradigm. Where does this end up? What are the advantages of cloud vs. hosted solutions? How should companies be thinking about storing their marketing-specific data for the next 5-10 years?

SS: A couple years ago nearly everyone was looking at the cloud. While some applications are well suited for the cloud, those built around real-time responses require bare metal performance. Fundamentally it depends on the SLA of the applications. If you need response times in the milliseconds, you can’t afford the cloud’s lack of predictable performance. The demand for efficient scalability is also driving more people back from the cloud. We’re even seeing this with implementations of Hadoop, which is used for batch processing. If a company can run a 100-server cluster locally versus having to depend on a 1,000-server cluster in the cloud, the local 100-server option will win out because efficiency and predictability matter in performance.

What are top companies doing right now to leverage disparate data sets? Are the hardware and software technology available today adequate to build global, integrated marketing “stacks?”

SS: Many of the companies we work with today have two, four, sometimes more data centers in order to get as close to their customers as possible. Ad-tech companies in particular tell us they have about 100 milliseconds—just one-tenth of a second—to receive data, analyze it, and deliver a response. Shortening the physical distance to the customer helps to minimize the time that information travels the network.

Many of these firms take advantage of cross data center replication to include partial or full copies of their data at each location. This gives marketers more information on which to make decisions. It also addresses the demand for their systems to deliver 100% uptime. Our live link approach to replication makes it possible to copy data from one data center to another with no impact on performance and ensures high availability.

Over the last year, we’ve have had customers experience a power failure at one data center due to severe weather, but with one or more data centers available to immediately pick up the workload, they were able to continue business as usual. It comes back to the earlier discussion. Data has the highest value when marketers can act on it in real-time, 100% of the time.

This interview, among many others, appears in EConsultancy’s recently published Best Practices in Data Management by Chris O’Hara. Chris is an ad technology executive, the author of Best Practices in Digital Display Media, a frequent contributor to a number of trade publications, and a blogger.

This post also appeared on the iMediaConnection Blog 1/11/12.