David Pakman's Blog www.pakman.com

Why AdTech is Back (It Never Left)

advertising expand
Oct 30

One of the most valuable characteristics of venture investing is that sectors go in and out of favor. Certain sectors, no matter the investment climate, have perennial long-term value. At least, that is my view. And I hold that view strongly about the AdTech sector.

More than 60% of the enterprise value created by internet companies comes from companies whose business model is primarily the selling of ads. Since the internet is both a communications medium and a transactions platform, I believe it will always create massive value through advertising. The internet, unlike most traditional media, is inherently a performance-oriented medium and it delivers on the promise to make advertising and marketing more accountable and more efficient. Underlying the delivery of better ad performance, in a world filled with big, quantifiable data, is an ever-increasing slate of sophisticated technology operating on massive datasets in real time. If you advertise on the internet, and eventually, every brand and service in the world will, you need exposure to these technologies or you will underperform your competitors. The AdTech sector, fundamentally, is the delivery of these advanced advertising technologies to all advertisers.

In my previous posts on the evolution of online advertising, I painted a picture, like so many observers of the space, of a world where all impressions are traded on exchanges. That inevitable transition is happening at light speed now. More than 17% of display impressions on the web are traded on exchanges and the forecasts are bullish on this trend. In that world, online advertising looks much more like trading stocks than the buying of ads over lunch meetings. In 2008, we believed that significant value would be built in this exchange layer. It was this thesis that supported our Series B investment in AppNexus. That company continues its incredible run and is one of the largest global ad exchanges in the universe. We believe AppNexus will remain one of the most important companies on the internet.

While a huge amount of buying has moved to the exchanges, the level of sophistication of many advertisers taking advantage of these RTB platforms is still rudimentary. It turns out, just like outperforming the stock market year in and year out, it’s hard to do it well. The amount of data available to buyers is enormous. The number of parameters available in tuning and targeting your audience is almost limitless. And, most importantly, there are always better data scientists down the road doing a better job than you can at building proprietary targeting models. For all of these reasons, in our opinion, the second-most valuable layer in AdTech is the data-driven ad network layer. (Not to be confused with the inventory driven ad networks.) Data-driven ad networks employ either large proprietary data sets or proprietary targeting models on top of very large data sets. The sophistication of the data scientists within these companies delivers a sustainable performance advantage over their less well-equipped peers. Two examples of these companies in our portfolio are Dstillery (formerly Media6Degrees) and Bizo. Both Rocket Fuel and Criteo are two additional companies in the space. Rocket Fuel’s recent IPO fetched it a market cap of more than $1.7B and Criteo is now over $2B. Many are asking themselves, “Why?”

[As an aside, Zach Coelius, the CEO of Triggit, points out that these companies should no longer be called ad networks because they no longer amass large amounts of inventory. They are instead more accurately "algorithmic media buying" companies or "data-driven targeting" companies...not really sure, but they aren't traditional ad networks.] 

The reason these companies are so valuable is that buyers on the exchanges are dominated by performance-oriented marketers today. Their dollars seek the best performance. The data-driven ad network layer is increasingly a case of the haves and have-nots. The better you perform relative to your peers, the more ad dollars you receive. These four companies significantly out-perform their peers, and their incredible revenue growth (and enviable media margins) indicate this.

The reason these companies have bright long-term futures is that this layer is increasingly necessary, hard to replicate, and experiences tremendous network effects. In the early days of AdTech, some believed the traditional media buyers would be able to build their own technology stacks and deliver better performance and value than independent companies in the market. This has not turned out to be true. The best performance can be found elsewhere, largely within technology companies, and so that is where the dollars are flowing. This presents enormous long-term challenges for the incumbent media buyers and will continue to pressure them to flow more and more of their client’s dollars to the better performing AdTech companies.

I believe this layer will eventually see tens of billions of dollars of media buying flowing through it. Of course, the exchange layer benefits from all of this too. For these reasons, there will be additional public AdTech companies which will fetch multi-billion dollar valuations coming to market. AdTech is back. Except it never left.


The Pressure on TV Networks, Ari Emmanuel and Cable Companies

Jun 06

Lots of recent discussion on TV and Hollywood. Ari Emmanuel accuses Google (again) of aiding and abetting pirates. Henry Blodget writes a nice piece on the changing TV viewing habits of consumers. Dan Frommer says those changing habits won’t really affect the MSOs and Networks anytime soon. And Jeremie Allaire seems to claim that Apple’s next move in TV will be to emulate TiVo’s (largely failed) box/cable-card strategy (but correctly points out the disruptive power of AirPlay). Oh, and Sean Parker launched version one of AirTime.

I wanted to add a few points to the discussion about the pressures on the TV industry. First, some basic observations:

TV programming is not homogenous

The uber-bright Hunter Walk provided me with a fascinating view into his opinion of the real tiers of TV programming. Out of the 4-5 hours of TV the average household watches each day, there are essentially three tiers:

  • Hour 1 (No Substitute) – This is the never-miss-an-episode, live-sports, must-see-TV that exists across many networks. The Sopranos, Mad Men, Yankees/Red Sox, French Open, Homeland, etc. When we watch TV, this is the first hour we watch. We watch this stuff live or DVR it and try never to miss it. We will pay for it any way we can and even endure roadblocks to watch it (like when networks won’t make it available on our preferred viewing device, or expire old episodes, etc.) While the networks believe 80% of their content fits this description, it is probably more like 20% of all shows currently on the air, at most.
  • Hours 2-3 (Nice to see) – This is stuff that we have an allegiance to, but are comfortable missing an episode and won’t really endure friction to see it. Many comedies fit this category, from 30 Rock to The Simpsons, as well as the countless procedural crime dramas like CSI, etc. The networks think all of this content is in the category above, but it really isn’t. And probably another 30% of all shows on the air fit this category.
  • Hours 4-5 (Filler) – This is the low-budget, mostly reality show programming that networks use to fill the time between their one or two hit shows. Think Kate Plus 8 or Let’s Make a Deal re-runs. The only time you watch this stuff is when you are couch-surfing. This is probably 50% of all programming on air.

When Ari insists that Facebook, Google, Twitter and everyone else in tech will have to “pay for Aaron Sorkin”, he is really talking about the “Hour 1″ category of programming. That stuff is really high-value and is not in a lot of danger of being disrupted any time soon (although the rising production costs and off-the-charts no-risk fees paid to talent are surely to be reconsidered in the future.) But as for the other two categories…

Our attention is shifting away from TV

All media operates in an attention economy. They compete for our attention against the backdrop of thousands of choices as to how we spend our time: email, video games, Facebook, Twitter, Flipboard, Instagram, etc. The latest numbers show those choices are finally catching up with TV; we are watching less of it, whether DVR’d or not. We aren’t watching less of that incredible No Substitute programming, but we are watching less of the 80% of the other stuff. And by the way, those big “hit” shows that Ari talks about have relatively small audiences. Only about 3 million people “tune in” for an episode of “Game of Thrones” and over the course of a week about 9 million people have seen it through various means. Same for Mad Men (3 million)Desperate Housewives (9 million) and The Good Wife (9 million). That’s a pretty small audience compared to, say, the 450 million on Facebook every day, the 800 million who watch YouTube videos every month, or the more than 100M people who watched the final episode of M*A*S*H. As TV and other entertainment choices proliferate, “hit” audience sizes have decreased. So, one of the immediate threats to network/cable television is that we are likely to watch less and less of the “Hours 2-5″ programming that fills so much of their programming grids. (The smart production companies know this and are already producing much lower-cost, quality programming for YouTube and other online-only outlets.) And where will that lead us?

The pressure will first come from the advertisers

If Nielsen didn’t lie and try to convince TV advertisers that the 50% of people with DVRs still watch commercials (hint: that is utterly ridiculous. We don’t watch any commercials anymore unless we watch a live sports event), I believe advertisers would appreciate that we aren’t seeing their commercials anymore. While the PC and mobile web still don’t offer nearly the great story-telling opportunities for advertisers as TV commercials do, it just doesn’t make sense to continue to buy very expensive TV media when no one sees your commercials. Certainly live sports TV CPMs will go up, but the rest has to fall as advertisers figure this out. And reports detailing that we are watching less TV has to start to sink in. Advertisers would love to try to buy only the hit stuff, but networks are good at bundling to force them to buy the filler programming too. But the whole bundle will start to feel more and more pressure.

The dual revenue stream model of the cable networks provides lots of air cover against decreased ad revenue. The affiliate fees they get for carriage will sustain them for a while. Brand advertisers are looking elsewhere to find places to tell their stories and to reach their audience. And online, we can target viewers and assemble audiences with drastically better efficiency (and reliability) than on TV. Online video is becoming so performance-based, that advertisers now can pay only when someone has actually watched the commercial and not pressed the “skip this ad” button. If you really care about making sure someone sees your commercial, online is the only place to show it. And more and more, we just aren’t seeing the ad on TV anymore.

What’s This Mean?

  • Advertisers will begin to spend less on TV and that will be the canary in the coal mine that big changes are afoot
  • We will continue our shift away from viewing traditional TV and towards IP-delivered unbundled shows, some which will have migrated from traditional TV, but many that will be organic and native to internet programming (the made for YouTube stuff is a prime example here.)
  • Ari will continue to demand high prices for the “Hour 1″ shows created by his elite clients, but the audiences for those shows will grow smaller and smaller.
  • As a result, networks will begin to feel the pinch of decreased advertiser spending, and they will try to raise carriage prices to the MSOs more aggressively
  • MSOs will keep trying to push our bundled TV prices up higher as a result of this, pushing more and more of us away and into other IP-delivered options
  • Finally, I believe the as more of us watch IP-delivered programming, the lure of certainty that the audience you really care about is seeing your ads will prove appealing to more and more advertisers, and online video ad revenues will continue a dramatic ascent
  • And so the cycle will go

(Update: this report from Pivotal Research refutes all of Henry’s points…but bases all of its observations on data provided from a single and biased vendor: Nielsen – a panel-based research method that looks at activities of only 25,000 households – and has concluded, for one, that those of us with DVRs still watch ads. Go figure. Oh, they make their money from the TV industry.)



Wither the Giants? The Arrogance of Aging Incumbents

Jan 25

My friend and former colleague Greg Scholl sent me an article this week and a provocative quote jumped out of it. Here is the view of Irwin Gotlieb, CEO of one the largest global advertising agencies on the planet, as he shared his view on this year’s CES. Given last week’s SOPA/PIPA debate, I thought Mr. Gotlieb’s observations were worth elevating, as they effectively capture a way of thinking that ultimately undermines incumbent media companies and the businesses that serve them:


Much of what we saw at CES relates to things we’ll be seeing 24 months out. In my mind, it’s all good: we’ll be able to target better, we’ll be able to segment better. The ads will be delivered on screens that are sharper, look better, larger, which ultimately provides more effective communication. There’s one last element: in the role that we [media buyers] play, we have a responsibility to ensure that technology develops in a manner that doesn’t shake up the supply-and-demand equation of our business, doesn’t destroy the content amortization business, isn’t disruptive simply for the sake of being disruptive.

If it does alter the supply-and-demand equation, it needs to do so positively, not negatively. When you have the share of the deal volume that we do, you can’t just be passive about it. You have to try and influence it. The technologies and devices that begin to get manifested at a trade show like this needs to be guided, so that it all works out in the best interests of our clients.

- Irwin Gotlieb, Global CEO, GroupM, originally appeared at TVExchanger.

We have a responsibility to ensure that technology develops in a manner that doesn’t shake up the supply-and-demand equation of our business.

A bold statement and, it seems, a common mindset for many incumbent business giants in their respective industries; a mistaken belief that they can somehow coax disrupting forces (be they new companies, or larger macro consumer trends) into conforming to their legacy business models and cost structures. As we have seen countless times, the actions of incumbents, when faced with technology disruption, often is to turn to litigation, legislation or other non-market strategies (i.e., anti-trust investigations, artificial price barriers) in an attempt to delay or block the challenging technology or companies. This perhaps work as a delaying tactic in the short term (Rio MP3 player case, Napster, book publishing agency pricing model with Amazon) but fails in the long term.

Mr. Gotlieb’s apparent belief that he and other advertising agency leaders can “ensure that technology develops in a manner that doesn’t shake up the supply-and-demand equation of our business” is futile in the long run but perhaps more pernicious is the implicit arrogance of thinking the market force of the web can be channeled into their bank accounts by sheer force of will. Of the many problems with this way of thinking, paramount is the ability to rationalize away making the hard choices and decisive actions to ensure the Group M’s of the world play a vital role in the new economy as they have done in the legacy one. (Cue Scotty from Star Trek…) “You cannot change the laws of physics.” For Group M and other incumbents, it’s almost difficult to fathom, given how entrenched and advantaged they are, that they could drop the ball. But, many will, as history has shown over and over again in times of market transformation.

Technology forces which bring greater efficiency and transparency to markets simply don’t care about privilege, access, and rolodexes. They disrupt predecessor markets because of structural problems like price opaqueness and false scarcity that no longer “work” in the new market. Look at Google: their entire approach to advertising is to ultimately remove the middle man just as increasingly, the media buying side of traditional agencies is the inefficient middle man, marketing up the cost of media to provide their services. Google is now selling $40B of media every year, the majority of it without a middle man (or at least with different sort of middle man … and in any case, getting far lower margins than traditional media bought by agencies.)

We watched as the music industry delayed their demise by suing Rio, Napster, and literally hundreds of others, delaying adoption of new business models not based on scarcity. We listen to Jeff Bewkes decry Netflix as the Albanian Army as he feverishly works to reduce their influence with his content. We observe the movie industry fight with everything they have to protect the windowing strategy and defend limited access to content instead of move towards open and immediate paid access to their movies. (Fantastic post on this from Rich Greenfield here, “Innovate Don’t Legislate”.)

And, as a microcosm of this larger conversation, we watched, over a very short period of time in the SOPA/PIPA debate, as the web demonstrated the disruptive advantages of network effects and scale, as over a period of weeks, legislation that appeared all but ratified was shuttered, up to and including an implied Presidential veto. Heady stuff. Granted, if we extend the metaphor and use SOPA/PIPA as a microscope, there are extremes on both sides, and it will be messy and require compromise if the big media incumbents and new technology disruptors are to learn how to co-exist. For big media companies and the service businesses that cater to them, this means recognizing the practical realities of changed business models – probably mostly that their cost of production needs to drop dramatically and they need fundamentally to re-think distribution and customer relationship management to remain profitable and relevant. On the tech side, it means recognizing that progress requires some level of institutional engagement and political compromise – because like it or not, this is the way our system of government works and how laws get written. This won’t be easy or natural, as it’s anathema to the culture of how new media tech and the startups that encompass it conceptualize and operate in our worlds. Facing reality and then demonstrating a bit more collaboration and compromise, however, would go a long way and be better for the customer, who, like our democracy, these industries ultimately serve. Because it’s the customer who is in the driver’s seat, and increasingly, they know it.

Perhaps it’s Pollyanna, but if so, my chips go on technology. Big media has the most to lose because after decades of the game being rigged in their favor, increasingly, it’s the opposite. Of course it is difficult and painful for media incumbents to embrace digital markets considering these markets ultimately are smaller and have less attractive economics. That’s presumably why big media executives are so well compensated – if it was easy, anyone could do it. The alternative, however, is to be disrupted by new entrants which don’t have any allegiance to aging business models and who could care less how out of whack someone else’s cost structure is. Coming back to Mr. Gotlieb’s view, I offer these thoughts. First, incumbents won’t be able to meaningfully guide the technology juggernaut of more efficient advertising mechanisms, so it’s perhaps better for them to focus their energies and advantages towards thoughtful reinvention. New technologies are bringing actual measurable performance and more efficient means of buying to a large share of advertisers. The challenge for incumbents is to adapt their enterprise to embrace this chaos and profit from it. The good news is, it’s doable. However, to think they can bluster their way out of this disruption is a fool’s errand.

This work is licensed under a Creative Commons Attribution 3.0 Unported License.

In-Stream Ads: What Twitter and Facebook are Missing

Feb 17

I’m a huge user of Twitter and spend more time staring at a Twitter feed than pretty much any other online/PC/mobile utility. Collectively, literally more than an hour or two a day, my eyes glance at a feed delivered to me on top of Twitter. In terms of time spent, maybe my email client is close, but that’s about it. Like most of us, I do this increasingly on a mobile device. Thanks to the excellent Twitter iPad,  iPhone and more-than-adequate (but not perfect) desktop apps, I rarely use Twitter’s web service. Because of all this, I have never seen a “promoted trend”, “promoted tweet”, or “promoted account”. Twitter and Facebook largely place advertising around the stream — in the chrome to the right of the feed. I know Twitter is experimenting with in-stream ads, but they are largely absent from Twitter clients today.

At Facebook, there is great resistance internally (and from some users) to the idea of putting ads into the newsfeed. Most believe that it would be a huge turnoff to their users. So Facebook is inventing some really cool social ad products that appear in the ad gutter to the right of the feed. But you never see these ads on Facebook’s mobile apps. And increasingly, that is where all our eyeballs are going, right? Even Facebook has admitted that a large amount of their future growth is expected to come from mobile products. So, where do the ads go?

I think they are going in the feed. They need to be highly targeted, socially influenced, and tastefully produced. They should solicit feedback from us and operate like content (contain links, be informative, utilize the conventions of the platform like hashtags, etc.)

I feel like it is 1995 again and we have the opportunity to invent the banner ad. There really is very little traction around what in-stream ads will look like. I am a firm believer that more and more of our time will be spent consuming information and media in streams. And if our eyeballs go there, advertisers will follow.

Author David Pakman
Category Ad Tech
Comments 23 Comments

How Facebook Reaches $20B in Revenues

blog-facebook-logo expand
Dec 16

Facebook has transformed the web, and indeed the planet, by connecting us all in ways we weren’t before. At their current growth rates, they will finish this year at around 600M users and will be larger than Google by the end of next year (if they aren’t already, see here). (As Facebook’s head of growth once told me, the nice thing about growth being truly viral is that it becomes very easy to predict your future growth. Facebook actually employs several epidemiologists to measure and predict growth.) Once Facebook has a true webwide view (i.e., is connected with nearly all 1.1B people on the web), their scale becomes completely un-ignorable by all major advertisers. They are already at this point today in many markets, but this will become true next year in essentially all online geographic markets.

Today, we understand that Facebook generates about $2 – $3 per user per year in revenues. Google, however, generates about $25 per user per year (more than $25B in revenue from about 1B users). The gap is considerable, but Facebook is just getting started with their monetization efforts. Their main forms of revenue today are engagement ads and self-service ads. The non self-service ads are sold by a large and competent sales staff around the world who cater to brands and agencies. They are selling these ads largely on a CPM basis. That is, they are not positioned as performance-oriented ads, but sold more like TV. Reach and frequency are the main measurements and selling points. Self-service ads are sold more to the mid-tail of advertisers and are sold more like Google AdWords on a CPC basis. These ads allow for extremely rudimentary targeting. The targeting criteria is based on the info you, the user, have put in about yourself. I am a male in my early 40′s in New York City and I like tennis. So, I see ads targeted at these keywords.

There are pros and cons to these forms of advertising. First, Facebook’s reach is undeniable and advertisers love the idea of appearing on everyone’s Facebook page for an entire day. The CPM ads are targeted more at brand advertisers who are less interested in demand fulfillment and more in awareness generation and demand creation. As Sheryl Sandberg has pointed out many times, this “brand advertising” market is many times larger than search/display “performance” marketing. They have their eyes set on TV.

I think the main reason they are focused on this is because, quite frankly, ads on Facebook don’t perform very well. We know that the effective CPM of these ads on Facebook is well under $1. Advertisers understand that social media has proven, thus far, to be a place where people don’t seem to want to be interrupted to click an ad. In addition, the performance of rudimentary ad targeting doesn’t nearly beat the efficacy of search targeting or good display behavioral targeting for that matter. That makes it less attractive to advertisers who measure performance by clicks, and also to a company who expects to be paid per click. Hence, sell CPMs! Facebook has lots of growth in these two ad strategies. But that is not what will get them to $20B in revenue quickly.

What will? Two things: payments and off-site social targeting.

Facebook payments, with Facebook getting 30% of all virtual good sales, will generate several billion dollars a year once it is ramped. This can become a $4B revenue line for Facebook within three years. But the other often underdiscussed opportunity is an off-site socially targeted ad network.

We know from our investment in Media6Degrees that the most effective form of ad targeting after search is social targeting. These are ads that are targeted essentially at the friends of a brand’s existing customers. Your customer’s friends show extremely similar brand affinities as you. Of course, these prospects don’t have to be the actual friends of your customers…they just have to have similar social signatures to your customers. And Media6Degrees is the leading company pioneering this form of targeting. Other companies exist trying to do the very same thing. Frankly, the results are staggeringly good.

The issue for Facebook is that applying this advanced form of targeting on their site would be uninteresting, owing to the lack of performance of advertising on social media. This is where Facebook Connect comes in.

Already, more than 2M sites have implemented Facebook connect. I believe what Facebook will do is to offer to every one of those sites essentially an AdSense replacement. Pull out your AdSense ads and replace them with socially-targeted Facebook ads. When a friend of a brand’s existing customer appears on a publisher’s site, they can see ads for that brand. I believe these results will perform considerably better than AdSense’s contextual/semantic targeting and hence provide publishers with larger payouts than Google provides them. Facebook already has the large social data set to understand the connections necessary for this targeting. Of course, this will further encourage more sites to implement FB connect and Like buttons, since FB can make these requirements to be in the social targeting ad network.

By the way, this form of targeting is proving to work equally well for video. Video is likely to be the most important form of ad creative deployed for brand advertisers. But to whom do you show the commercials? The friends of your existing customers.

At scale, with, say, 5M sites in this network, FB could operate the largest off-site ad network (display and video) that outperforms all others. At this scale, I can see this generating $15B a year in revenues.

What’s stopping them today from doing this? Well, first, they don’t have to. They have plenty of headroom left in growing the existing businesses. In addition, the data science necessary to pull this off at scale is profoundly challenging. It’s not as easy as it sounds, particularly if performance matters — there is a cost to showing an ad to someone who may be connected to a customer, but not close enough. It takes years to perfect this form of targeting. Finally, the current privacy scrum needs to further settle out. I think we could see Facebook launch products like this in the second half of next year. This will get them to $20B in revenue and worth at least $100B in market cap. Stay tuned.

Author David Pakman
Category Ad Tech
Comments 6 Comments

The Brand Social Graph

Apr 28

I very much agree with Fred Wilson and Albert Wenger that we construct various social networks, not just one, to suit various needs. One’s Facebook social graph is often the group of people with whom you are comfortable sharing light-hearted info and photos. One’s Foursquare graph is often more limited: the group of people whom you want to know your physical location. Your Twitter graph is really an interest graph and I think, over time, less about people and more about information sources. Your Venmo graph is even more limited: those people with whom you exchange money and trust. Facebook, I think, could have developed additional functionality sooner to provide impetus to construct these additional graphs (often a subset of your FB graph) directly with them, but I think they missed this. They have been too slow to recognize the importance of physical location, were late in allowing asynchronous following, and are now getting around to payments. None of this is likely to impact their continued massive success, but it opened up opportunities to others.

Missing from this discussion is what I find most interesting about social media: the graph most important to monetization. There are interesting businesses being built on top of the graphs discussed above. However when you really want to build an enormous media business on top of the ridonkulous amount of social data available on the web, you find a new graph emerges: the brand social graph.

At Media6Degrees, the company has pioneered the ability to assemble custom audiences for brands from the connections between their customers and those customers’ friends. For any particular brand, the company actually computes (all anonymously, of course — no personally identifiable info is involved) a brand social graph, finding the friends of a brands’ customers most likely to be interested in actually buying the product. It turns out this is much harder than simply advertising a product to everyone in someone’s social graph. We have multiple connections between us as people, and some are more predictive of mutual brand affinity than others. The brand graph can change by the day, ad campaign, product or offer. This is a hard data problem, but one whose rewards are enormous.

The brand social graph is a big idea. I think, as social media matures, friend targeting via a brand social graph will be a significantly large business model for the largest media companies. Perhaps one qualification: the search targeting model dominated by Google is still likely the most effective form of advertising ever created: the user’s intent is entirely clear, so advertising works really well. But it isn’t great, as we all know, in building brands, finding audiences and creating demand. Brand spending and demand creation is a much larger part of ad spending than demand fulfillment is. I think these needs will largely be served by friend targeting in the future, and one which Facebook is in the position to dominate (but it not yet doing). Google, Yahoo and Microsoft all have enormous amounts of relevant social connection data but do not yet show evidence of using it. Many of Facebook’s announcements last week put it in the pole position of delivering web-wide friend targeting when they deploy an ad network on top of FB Connect one day. And Twitter’s interest graph also gives it a great position in this emerging ecosystem.

Author David Pakman
Category Ad Tech, Startups
Comments 5 Comments

Areas of Focus

Dec 28

With one-quarter of a cup of coffee left, my fingers are jumpy as my mind turns to next year. In my first full year as a VC, I have been humbled by the great entrepreneurs I have met (and those I haven’t!), the brilliant colleagues of mine who help guide me, and the powerful teams of people who make startups work. I have started to hone my thinking about investment areas and although this is a never-ending process, here are some areas about which I am continuously excited:


All advertising is asymptotic to performance-based buying. No matter what those who have sold CPM TV advertising for fifty years tell you, advertisers want performance metrics with meaning. Google proved it was possible on the demand fulfillment side of the equation. The next wave of innovation, fueled by fantastic availability of data coupled with network and social theory algorithms, is the science of assembling audiences ripe for your brand. These are audiences who may not yet know about your brand or product. This is the demand creation side of the marketing coin. But advertisers will get real honest-to-goodness performance-based buying here too. And this, not just the availability of online video advertising at scale, is what I think will help alter the “only 8% of brand dollars are spent online yet 35% of our time is spent online” imbalance. The entrepreneurial ecosystem  focused on this opportunity is quite literally filled with some of the smartest computer science people around, and the density of startups is greatest in NYC.

Video Ad Tech

When you apply the innovations coming out of the effort above with the right video ad units online, you have the future of TV advertising (except it won’t take place over cable wires and terrestrial broadcasts). That’s a big market. Really big. And I like it.

The Disruption of Big Data

As I have discussed before, the challenges introduced by the amount of data we emit as humans attached to so many gadgets is disruptive to the systems we have used to process and serve web data and applications for 15 years. Turns out these are hard computer science problems and require different thinking. (In fact, it even requires a different engineering degree! Apply here and see if you can get in to Penn’s new Market and Social Systems Engineering elite program.) You see this first in mobile, social, real-time and network applications. As more and more sites incorporate both feedback and content from the real-time web, their system architectures will change, their approaches to databases will change, and this creates opportunities for startups servicing these areas. One such startup, Cloudera, has commercialized Hadoop to address the needs to operate and distribute large data sets. Here is a nice overview of Hadoop’s architecture.

Social Media and the Enterprise

The impact of social media on our societies cannot be understated and has been so well-expressed by so many (see this brief, popular prez as one example.) The impact of social media on how we run our companies, make decisions in the enterprise and manage teams is an area of focus for me. We need new decision tools, collaboration tools (Facebook doesn’t attempt to solve this, see this post) and new ways of communicating with our teams, partners, and most importantly, our customers (who become our partners). (Actually, customers always were our partners, but we never really treated them that way. Now we have to, thanks to the power of social media.)

The New Marketing Department: Our Customers

Well, good marketers know that customers always were the best marketers for a brand, product or service. With social media, the power has fundamentally shifted from company-has-megaphone to customer-has-megaphone. Some companies still choose to ignore this (often at their own peril, see “United Breaks Guitars“) but many don’t and are clamoring for tools, methods, and best-practices about how to engage their customers across social platforms. I think it requires a structural change in our approach to marketing and communications. And with this change comes new value creation opportunities for startups. Think about all the newswire companies and how likely they are to be disrupted here. Press releases? The only function they serve in the new world is regulatory (must get info to the market simultaneously) and even that will change.

Real-time, Baby!

I am fascinated by the real-time nature of the new web. In some ways, I always have been. In 1993, I had this game concept of “flight simulator meets the real world”. The idea was that if you took a flight sim game and plugged in real-world, real-time data, everything changes. Real weather, real traffic, real news events. Imagine flying over NYC and seeing the real world conditions mirrored in your game environment. In my small head, that was the beginning of this fascination. Couple that with the well-understood appreciation for the value of crowd-sourcing and you have the real-time web. Soon, this will be billions of people acting as the world’s best information and entertainment curators, filtered through your social graph. It dwarfs and upends every form of editorial filter that has come before. Apply this to business information and Bloomberg is disrupted. What media company isn’t challenged by this notion? And every communication company? And every company that guards a monopoly over its distribution channels? With a mobile, social, real-time web, we can always find a way around every barrier. It’s so powerful. And I love it.

So Happy New Year everyone! I am excited to work with you on the inevitable progress and disruption coming our way in 2010. Time for another cup of coffee.

How to Monetize Social Networks

Oct 13

As comScore and others have reported, we are spending far more time on social networking sites than on any other sites or web activities today. From eMarketer today:

Time spent on social networking sites is increasing steadily, taking users’ attention away from sites such as portals. According to comScore, the top 20% of social network users visit networking sites 2.4 times per day, on average, and spend 31 minutes on them—twice as long as the same users spend with e-mail or instant messaging. And that, in turn, means more time with ads.

The problem is, this inventory is sub-optimal for display and text advertising. The same article discussed the challenges of low CPM inventory. Users are too engaged in social communications and games to click away on ads. Also, when we are reading about friends, we are not demonstrating intent like we do in a search query, so ad targeting is harder.

The truth is that we do demonstrate intent, authority and endorsement while on social networks. We demonstrate who our closest friends are, we share links to goods and services we like, we click on links shared to us by friends we trust. We throw off all sorts of useful data for ad targeting. But social networking site inventory is not the inventory on which to display such targeted ads.

The solution here is to use social networking data and re-target existing customers, prospects, and friends of prospects. This allows highly relevant ads to be delivered to customers on inventory less likely to be interrupting task-based or conversational behavior. Media6Degrees, a Venrock portfolio company, is doing exactly this. And they are not dependent on a single social network, but instead have amassed a meta view of our connections on the web. They are building custom audiences for brands that are highly scaleable. The 50 or so advertisers who have used them have seen dramatic results. There are others working on the same problem. 33Across and Lotome are focused on a similar problem. And I wouldn’t be surprised to see Facebook launch social ads on an AdSense-type distributed platform in the future.

Author David Pakman
Category Ad Tech, Startups
Comments No Comments

Got a Ph.D. in Behavioral Psychology, Comp Sci and Statistics? We’re hiring!

Jun 24

My friend and fellow VC Raj Kapoor at Mayfield Fund has an insightful post about monetizing social networks. He believes social networks’ value is in the data, not in the inventory and I completely agree. In fact, this realization is the basis for our investment in Media 6 Degrees, who, so far, is the only company who has figured out how to do exactly this.

As Raj mentions, a new economy is developing around targeting data. As advertisers demand more performance out of the media they buy, the need for more precise targeting in the display space emerges. Data companies like Blue Kai and Exelate can sell you very specific audience based on a host of particular targeting criteria. Often times, the more precise your targeting, the smaller the audience you can reach. Media 6 figured out that social data is the key to both precise targeting and audience expansion. It is helpful to be able to target people who are interested in your product. It is better if you can also target their closest friends who are most likely to be interested in the same product based on the concept of homophily.

Today, the social nets have a treasure trove of data about their users and can put that to good use both in the service of advertisers and in making advertising more relevant for the end-user. It’s important to note that doing so in a way that actually produces great ROI for the advertiser is pretty hard. Media 6′s data scientists have been at it for years. It’s a new field (or at least a new application of a well-understood concept) and I look forward to seeing years of innovation in this space. It is fun to see interactive advertising technologists now spending lots of time with social scientists and behavioral psychologists.

Author David Pakman
Category Ad Tech
Comments No Comments

The Ad Ecosystem: It’s All Going Towards Performance

Apr 07

ROI. Get used to it, Madison Avenue.

In the past 15 years, the hyper-growth of interactive media has presented the advertising world with some of its most pressing challenges; since it is so easy to measure performance online, advertisers would like to better measure the performance of non-interactive media (print, radio, outdoor, TV). Spurred on by John Wanamaker’s legendary quote, “Half the money I spend on advertising is wasted; the trouble is I don’t know which half,” advertisers have been asking big media to justify $30 – $100 for much of the last five years. The pressures of the current economic conditions have only added to the urgency.

My view of the prices of media advertising in general is somewhat controversial. I believe the reason print and TV have been able to demand the rates they have is largely because there were so few ways to measure the efficacy of the buy; they are largely opaque non-performance-based media. We all know TV viewing rates are “measured” by a measly 5,000 Nielsen Peoplemeters and rates have been artificially boosted by the entire “Sweeps” concept and the Up-front sales process. When you look at the TiVo data, you get a more clear understanding of how poorly correlated the Nielsen viewing numbers really are with reality (when DVR usage is taken into the picture.)

Why talk about this now? Because online, every click is known. We know people try their best to ignore ads. The heatmaps of eyeball viewership online demonstrate that people try their best to avoid looking at banners. But we have also learned that search is the best indicator of intent and that when highly relevant (text) ads are shown, users click in great numbers. And when we combine intent with relevancy and interactivity, the real value of reaching a customer is known. Today, Google’s adwords market is the best indication of what a customer is worth. And it is nowhere near $100 CPM.

Why are online CPMs $5 – $20 for text and banner ads and TV is 5-10 times that? Because TV (and most traditional media) has been significantly over-priced for decades and online media is priced closer to the value to an advertiser of reaching a consumer. In addition, there is a growing supply and demand problem. What percentage of total available web inventory is unsold? 60%? 80%? That drives down prices. And even so-called premium inventory (Yahoo home page, nytimes.com) is sold at half that of TV CPMs.

So, where is this all going? Traditional media CPMs are coming down. And as those media become interactive and can be sold on a performance basis, their effective CPMs will fall dramatically as the true value of reaching a customer emerges.

This means the race is on for making all media as hyper-targeted, efficient, and as performance-based as it can become. One of my favorite digital media sectors is the advertising technology space. There are a plethora of companies focused on exactly this. By using every type of math around applied to every piece of data thrown off by our surfing behavior, we can make advertising more relevant and measure its efficacy. This satisfies the advertiser’s need to buy advertising on a performance basis and, in theory, makes advertising that much more relevant to the consumer. (I am skeptical that, aside from search marketing, we have yet seen any advertising actually appreciated by a consumer, but that is the subject for a different blog post…)

So, stay tuned for more thoughts on this space. I’d like to explore some specific sectors of ad tech like media buying, behavioral targeting and media exchanges.

Author David Pakman
Category Ad Tech
Comments 7 Comments