Ashu Garg

In its 10Q filing with the SEC yesterday, Google disclosed that its cost-per-click in the first quarter declined 14% year-over-year and 6% sequentially. Google asserted in the filing that "the decrease in the average cost-per-click paid by our advertisers was primarily the result of the strengthening of the U.S. dollar relative to foreign currencies," but also reflected the way advertisers managed their ad costs in response to the downturn. "Specifically, we believe that as a result of the general economic downturn, advertisers, in aggregate, have lowered their bids for keywords in response to a decrease in the sales they are able to make per paid click," the company said. During the same time period, aggregate paid clicks increased 17% year-over-year and 3% sequentially.

Google's explanation that the drop in cost-per-click is a result of a drop in conversion rates seems plausible. It also implies that advertisers were already paying the most that they could afford. That is because if advertisers had room to increase bids (assuming the same conversion rates), then they would do so to try to increase market share as the overall market declined. Therefore, while the cost-per-click will probably see a onetime increase when the economy recovers, it is less likely to continue to grow at the torrid pace that it has over the past several years.

Where does that leave Google? With market share growth being limited and overall search volume growth slowing down, it just makes Google's display and video initiatives more critical to its future.

Wendy Millard in her speech at the recent IAB Conference talked about how (wendy millard/IAB Conference) "too much data is the problem and not the solution ". She then went on to talk about how "massive reams of data still can't create compelling messages, and that (compelling messages) should be the focus of the interactive ad industry". She also said that "an over-emphasis on measurement is holding back the business." In addition, Randy Rothenberg in his blog (http://www.randallrothenberg.com/ ) re-iterates the point saying that on-line advertising, as an industry, has not focused enough on engaging the creative teams at agencies. He goes on to quote Bogart (Strategy in Advertising: Matching Media and Messages to Markets and Motivations) as saying "Advertisements may be evaluated scientifically; they cannot be created scientifically."

I think that Bogart is right. Science cannot create ads. However, it can only make evaluation so cheap (and precise) that the value associated with creation is declining, at the limit, to zero.  Dynamic ad construction tools create infinite combinations of ads combining a variety of copy, images, product information and offers at virtually no cost. With machine learning tools it is then possible to run large scale tests and figure out what ad works best in each situation. In such a process, while there is a role for the "creatives" it is far less valuable than traditional media where "creatives" rule the roost.

I share Randy and Wendy's conviction that there isn't enough brand advertising on-line and the medium is well suited to building brands. However, Wendy and Randy seem to imply that data driven advertising has less of a role to play in the context of brand advertising. I disagree. I think that data driven advertising can substantially enhance brand marketing as well. And Dave Morgan's new start-up, Simulmedia, which is applying data and analysis to improving the effectiveness of TV advertising, is a great example of that (see Dave Morgan:back to the start-up world).

Data driven advertising is here to stay. The art vs. science war is over. And the quants have won!

11:25 PM GMT  |  Read comments(2)

February 15

Notes from NASSCOM 2009

I spent Friday at NASSCOM's (India association of technology companies) annual conference in Mumbai. Like always, it was a vibrant event where entrepreneurs, students, bureaucrats (I was amazed to see teams from around the world), politicians and thought leaders came together (http://indialeadershipforum.nasscom.in). It was fun to meet so many old friends and the evening's entertainment, which celebrated Mumbai's indomitable spirit, struck a chord. I had tears in my eyes as I saw a video montage of the attacks. And proud as I saw how the city responded, and back in high spirits as the dance performance began. Below are a few observations.

1. The overall mood is best described at cautiously optimistic. The recession, the Mumbai attacks, and the Satyam scandal were on everyone's mind. At the same time, several companies I spoke to saw this as an opportunity to take market share and drive further off-shoring. Most were also optimistic about the Indian and Chinese economies and the opportunities they presented (especially for the Indian software services industry which has largely focused on developed markets).
2. CK Prahalad (wiki - C.K. Prahalad) presented extracts from his new book, The Age of Innovation, and laid out some implications for technology companies. Key takeaways were:
   1. CK emphasized that volatility (of markets, policies, etc.) was here to stay and companies need to adjust to that new reality through a combination of strategic clarity and nimble execution. He argued that clarity of direction was critical at a time where companies needed to make hard choices, and that agility would be a key competitive advantage in a volatile environment
   2. CK also shared his new mantra, "N=1 and R=G" (New-Age-Innovation/first pages). N=1 is the notion that firms need to focus on individual customer experiences even when they serve millions of customers while R=G is the idea that firms need to build an ecosystem of global resources that they influence but do not control. He went onto argue that technology companies have a unique role to play in enabling such a transformation by providing the IT infrastructure, unique analytics and facilitating flexible business processes.
3. One of the most interesting start-ups I came across at NASSCOM had nothing to do with technology. Jerry Rao (wiki/Jerry_Rao) who founded Emphasis (which he merged with BFL and then sold the combined entity to EDS) is starting a for profit low cost housing company. Jerry's aspiration is to build 1 M homes that he can sell for $ 10,000-20,000 each in the top 5 cities in India and do so while making healthy profits. What intrigued me was Jerry's consumer centric and disruptive approach. He talked about how most Indian developers view land as an appreciating asset and therefore are not incented to quickly complete construction – their business model assumes that prices will continue to rise and that construction delays are actually beneficial. On the other hand, Jerry plans to treat land as inventory with a singular focus on getting rid of it as soon as possible. In addition, through focus groups (what an idea!), Jerry had also identified the key buying criteria for his customers and was putting in place the infrastructure to deliver those – 15 years fixed rate loans (currently impossible for this target segment to access), access to an English medium school and basic amenities (sewage, water, enough electricity for a light and a fan), security, etc. Best of luck, Jerry!

I share CK's optimism. Technology has the potential to lead a new global growth wave. If there is one thing that this downturn has proven, it is that sustainable value cannot be created by financial engineering. Instead, firms need to build innovative products and services that consumers will feel compelled to buy and to do so, they will need technology more than ever before.

6:04 AM GMT  |  Read comments(2)

January 29

Will DVD players also go the way of cable boxes and LP players ?

3 years ago, watching web video meant watching grainy short clips. Over the last year, watching TV shows on the web has become common place - 50% of Americans have done so. When will we starting watching full length movies on the web ? Maybe sooner than you might think.

Over the last year, Netflix has made it possible to watch movies on your TV if you have either an X-Box or a Roku box. IN its latest earnings release, Netflix also announced plans to work with BluRay player manufacturers to embed their streaming client in as many player models as possible, starting with the newly released LG BD300. Also, Netflix continues to expand the digital content library (now 12000 titles) and improve the PC/laptop viewing experience (via Netflix.com). Also, Hulu.com has expanded from TV shows to include movies and has added 100s of movies to its on-demand service. To top it all, Viacom, MGM and Lions Gate just announced that they will launch Epix, an HBO competitor, on the web first in the fall - Epix launches on the web.

What's next ? Will BluRay die in its infancy ?

10:27 PM GMT  |  Read comments(0)

January 27

When will set top boxes go the way of LP (record) players ?

Consumers are increasingly watching video content on a variety of Internet-connected devices, including Internet TVs, laptops, mobile phone, etc. In November 2008, over 75% of the US Internet population (~ 150 M users) watched an average of 4.5 hours of video on-line, an increase of 34 % YoY. While that number is remarkable in itself, if you drill down, the statistics are even more compelling:

• A little more than 20% of the users ("the heavy users") spent ~ 14 hours/month watching video on-line. Beat Kneacht, founder of Zattoo (http://zattoo.com/), a European online TV service provided anecdotal confirmation - Zattoo users in Switzerland spend roughly 14 hours/moth using the service.
• This is not a teen or youth phenomenon anymore. The penetration of on-line video viewer-ship is largely similar to TV across age groups, except for the 65+ demographic
• Close to 15 % of the total views are for professionally produced content while another 15-20% are for "audience syndicated" content (a nice word for pirated) while the rest is UGC
• Yahoo, MSN, Viacom, Hulu, AOL and Turner all have more than 20 M UU/month, providing them with the reach/frequency required to be credible to media buyers
• In a recent survey, SRG found that 50% of all respondents had watched TV online in the past. Furthermore, 70% of 18- to 34-year-olds have watched TV online in the past. In contrast, only 36% of that group had watched a show on a TiVo or some other DVR at any time in the past.

Furthermore, audiences are fragmenting as consumers view video content via a variety of web sites ranging from video distributors (e.g. Hulu, Joost, YouTube, MYSpaceTV) to niche special interest focused web sites (e.g. Glam, Martha Stewart). For example, Hulu.com which is the leading distributor of professionally produced video individually accounts for only 1.8% of total video views (in the US), of which 67% are consumed through Hulu’s syndication network (vs. on hulu.com).

Advertisers and content owners are responding to the shift in consumer behavior. In 2008, on-line video ad spend was $ 570 M (US), and it is one of the few spend categories that is expected to grow (45% YoY) in 2009 as advertisers shift budgets from TV and prioritize video over other digital spend categories. Content owners (e.g. CBS, Viacom) are aggressively expanding distribution of their (i.e. syndicate) advertising supported content to a multitude of websites so that consumers can “watch wherever /whenever”.  Content owners are also actively mining their archives to create refreshed (or made for the web) content pieces - see  NYT: Slicing Decades of Video for New Life on the Web

And this is just the start. In 2008, 4 % of those who watched online video said that they would consider disconnecting their TV service. And, when I talk to people in their 20s in the bay area, most of them already seem to have done so - they watch TV on their PCs or through internet connected TVs. In addition, content availability could explode as networks recognize that the only way to avoid becoming irrelevant (like the music industry) is to aggressively expand and distribute online content. And in a couple of years, all TVs will come with internet connectivity out of the box. At that point, what percent of the US population will give up their cable connection to only watch TV online ? When will cable TV go away ? Or at least, when will the total number of connections start to decline ? Probably not next year, but also not 10 years from now!

11:23 PM GMT  |  Read comments(0)

December 31

Predictions for 2009: On-Line Advertising

In Q1 2008, with some trepidation, I started to write a blog on on-line advertising. What started as an occasional post is slowly morphing into a weekly post, and a few people even seem to read it regularly! And I am having a lot of fun in the process.

So, with even more trepidation, I am now taking a stab at making a few predictions for 2009.

1. On-Line advertising will continue to grow in 2009, both in absolute $ terms and in terms of share (of media spend). In the last quarter, as sentiment about the economy and therefore advertising spend has turned negative, downward revisions about on-line advertising spend are wide-spread, with an occasional doomsday (i.e. 30% drop in on-line ad spend, similar to 2001) prediction. What these predictions fail to take into account is the fundamental shift that has already happened in media consumption - more than 20% of all time spent (on media) is now on-line (vs. 5 % of ad spend). For higher income and younger audiences, those numbers are even higher.  This secular trend continues unabated and could even accelerate as consumer seek "cheap thrills". As Jeremy Liew cogently articulates in this post (http://lsvp.wordpress.com/2008/12/11/consumer-internet-predictions-for-2009/), social media/networks and on-line gaming are likely to benefit from the recession (in terms of usage). Also, the traffic growth in international markets (especially Asia) will continue in 2009. As a result, I expect global on-line ad spend to grow by at 15 % +  and to gain at least 1 percentage point in total share of media spend,in 2009.  For more detailed spend forecasts, take a look at recent eMarketer forecasts for the US - http://seekingalpha.com/article/112156-some-predictions-for-online-advertising-in-2009. 
2. Performance oriented advertising will see the most growth. Overall, advertisers will continue to shift spend to media where there is the most accountability and a clear ROI.  In particular, I expect significant spend to shift on-line from newspapers and off-line direct marketing vehicles. As a result, both search and performance oriented display will benefit. At the same time, I do expect media prices to decline. In the case of search, the recession could drive reduced conversions which in turn could drive a reduction in the marginal pricing that advertisers are willing to pay for each click. In the case of display, there is an inventory surplus which could drive down prices. Also, there is a fundamental shift in value away from media owners to data owners that is underway (see below) and will impact pricing.
3. Data driven display advertising (Display 3.0) will come of age in 2009. Historically, search advertising has been akin to a sniper approach (with advertisers buying users with a specific purchase intent) while display advertising (including performance oriented display) has been a shot gun approach where advertisers buy context or at best broad brush segments. This paradigm is beginning to change. In 2009, data driven display advertising (what I refer to as Display 3.0 in earlier posts - Display 3.0: What is it ?) will be widely adopted by leading edge performance marketers (the early adopters). Specifically,  dynamic ad serving (take a look at companies like Teracent and Tumri) will be widely used. In addition, leading edge advertisers will begin to buy media from exchanges (like Right Media and AdECN) on a fine grained basis (paying based on a profile or a specific behavior). As part of this shift, advertisers like Target and American Express will begin to combine internal (including off-line) data and segmentation models with 3P on-line data to deliver personalized ads; imagine Target combing your in-store purchase history with your Amecian Express card data and your search history to serve you personalized ads on-line. As a result, user data will begin to be bought and sold (akin to media), with large retailers, credit card companies and other data aggregators being big beneficiaries of this trend. Data exchanges like Blue Kai and start-ups who are able to capture user data and convert it into advertising related insights (see Data Liquidity for more thoughts on this front) will also see increased traction, and capture value at the expense of media owners.
4. Portals (Yahoo, AOL, MSN) will see declines in CPMs, and possibly in absolute revenues. Mainstream advertisers will begin to question the price premium they pay for portals, especially as top tier ad networks offer compelling alternatives and data driven advertising takes off (see above). This coupled with the shift in consumer time spent to UGC content will be painful for portals. Expect to see major re-structuring at all three portals as they begin to respond to the downward spiral.
5. On-line video advertising will continue to grow, while other emerging media will take a hit. Overall, on-line video will continue to grow (despite being brand focused) as advertisers look for alternatives to TV. However, pricing is likely to come under significant pressure. On the other hand, I think that the recession will negatively impact in-game advertising, mobile advertising, and advertising in applications like RockYou and Slide.  This is in part due to both the pressure on experimental budgets and their focus on brand $. The possible exception to this could be mobile advertising in Europe, India and China where mobile phones are the "new PC".
6. Ad networks: survival of the fittest. The top 10 ad networks will benefit from the recession and grow at the expense of both other ad networks and the portals. In fact, I expect at least 1 ad network to declare its intent to go public in 2010, with others lining up to follow. Times will be tough for the rest. 100s of ad networks will die in 2009. For more details, read  Ad Networks: Alive or Dead? and Follow up to Ad Networks: Dead or Alive ? 
7. Google emerges as a force in display. In 2009, Google's investments in display will finally begin to bear fruit. The re-launch of the DoubleClick ad exchange,  the scale up of Ad Manager (a free publisher ad server/revenue optimization tool), the launch of behavioral targeting and other display advertising related offerings will all come together in 2009. In addition, Google will benefit from the lack of competition - Microsoft's on-line business, Yahoo and AOL will probably continue to be rudderless in 2009.

These trends will benefit some companies while disadvantaging others. I hope that you find ways to navigate through (and even benefit from) the turbulence. Let me know if you agree or disagree with these predictions (my e-mail is agarg(at)foundationcapital.com), or if there are other trends that you think I’ve missed.

Have a great 2009!

3:38 PM GMT  |  Read comments(2)

December 23

India's Power Sector: a $ 600 B Opportunity ?

According to a recent report by McKinsey & Company (see McKinsey - Powering India: The Road to 2017 for an executive summary), if India were to grow at  8% p.a. for the next 10 years, it will require a 3 X increase in power generation capacity (to 300+ GW). Considering that India has added 5-6 GW p.a. in the recent past, this would require a 5-10 X increase in the pace of capacity expansion and an aggregate investment of $ 600 B (by 2017).

While this is clearly a multi-faceted challenge, there are 2 dimensions to it that are worth noting.

1. The shortage of peak power. The peak power deficit in India is currently 16 % and is expected to rise to 20% by 2017. On the flip side, there is already a base load surplus of 5% which is expected to rise to 14 % by 2017).
2. The need to reduce distribution losses. Average distribution losses in India are around 37-40 % rendering all distribution companies financially unviable.

While a key element of the solution is to accelerate the build out of generation capacity, there are other levers that India needs to pull, including:

1. Scaling up solar power generation. India's solar potential is amongst the highest in the world (with an annual specific solar yield that is comparable with California and almost 2 X of Germany). Furthermore, solar is a great solution for India's peak power requirements (unlike wind whose production is mostly concentrated in the months with least peak shortages). Yet, India has an insignificant amount of solar generation capacity today. India needs to launch a massive effort to scale up solar power generation capacity, possibly setting a target of 10- 20 GW of solar capacity by 2017 (even this will be only 10-20 % of the expected peak power shortage by 2017). Initially, these capacities will require subsidies to be viable, but over the next 5 years, with economies of scale and yield improvements, solar can achieve grid parity. In fact, given that significant amounts of peak power are generated via diesel generators at costs of $ 0.30 /kWh, one could argue that solar is already at grid parity in India. Fortunately, there seems to be some early signs of government support for such a scale up (india-climate-plan-summary/06-2008), but we need to see more wood behind the arrow.
2. Demand Side Management. Building 300 GW of additional capacity is both impractical (requires 5-10 X increase in pace of capacity generation) and suicidal (it will double carbon emissions on a per capita basis). India needs to reduce absolute demand (via improved energy efficiency) and shift peak demand to off-peak hours. Time-of-use pricing, energy efficient appliances/lighting, and higher efficiency standards for new construction can reduce demand from 10-15%.
3. Leveraging technology to reduce distribution losses. Technologies ranging from real-time metering, to pre-paid cards will play a pivotal role in reducing distribution losses. can help India reduce distribution losses to ~ 15%. 
4. Increasing the influence of market forces. the step changes required are only possible by dramatically increasing the role of market forces in the power sector.  India needs to privatize distribution companies, create efficient whole sale markets, facilitate the creation of merchant generation capacity, accelerate private sector investments in equipment manufacturing and incentivize private companies to improve the plant efficiencies (PLFs) of poor performing public sector plants. 

Over the next 10 years, India's power demand growth will be second only to China. Furthermore, McKinsey estimates that there the potential for industry particpants to capture upto $ 160B in EBITDA by 2017.

From my vantage point, the Indian power sector offers an incredible opportunity for value creation for start-ups - ranging from providing technology to building demand management businesses and distribution franchises. If you are working on such an opportunity, do drop me a mail at agarg(at)foundationcapital.com

Over the last year, I introduced the concept of Display 3.0 (Display 3.0: The Next Wave of Innovation in Display Advertising), and how it would all be about improving advertising effectiveness by leveraging advertiser/consumer/publisher data and algorithms. I also talked about the opportunity to build advertiser facing tools/platforms (Display 3.0: What is it ?) ranging from dynamic creative construction and optimization to landing page optimization. Furthermore, I discussed the emergence of advertising exchanges, and how the liquidity of inventory and the ability to buy inventory in a fine grained way will create a host of new opportunities for advertisers (Follow up to Ad Networks: Dead or Alive ? ).

One trend that I have not talked about to date (a miss on my part) is the increasing liquidity/portability of data, ranging from advertiser data to consumer data. The liquidity of data is critical to the success of Display 3.0 and will create a host of opportunities for both publishers and advertisers.

Historically, large and sophisticated publishers (like Yahoo) have always valued their data and leveraged it to enhance the value of their inventory. Subsequently, targeting vendors like Tacoda and Revenue Sciences offered out-sourced targeting solutions to other premium publishers who lacked the technical capabilities in-house (but in a similar proprietary data model).

What has emerged over the last 2-3 years is a range of ad targeting solutions that aggregate consumer data across publishers to create targeting products (based on behaviors, purchase intent, psychographics, demographics, etc.). Most of these offerings provide publishers with a free service in return for which they collect audience data which they then leverage for ad targeting (not all of these have launched ad targeting products to date). Examples include:

1. Free site analytics providers, e.g. Quantcast, Compete.com, Hitwise
2. Web services (e.g. rating, comments) providers, e.g. JS-Kit
3. Yield optimization platforms, e.g. Pubmatic, Admeld, Rubicon
4. Personalization/recommendation service providers, e.g. ChoiceStream, Aggregate Knowledge
5. Context extraction services, e.g. Opinmind (which extracts purchase intent from social communication)

In addition, advertisers are beginning to leverage their proprietary data to improve advertising effectiveness. The most obvious example is re-targeting where advertisers leverage behaviors exhibited when you visit their site to target you around the web. But there is more. Advertisers are beginning to think about how to link off-line data and segmentation models to users on-line. One approach is to send users e-mails with embedded cookies to associate their off-line behavior with their on-line presence and then to subsequently track them (anonymously off-course). Another approach is to drop cookies when users sign in or complete a transaction (this is especially applicable to e-tailers).

Furthermore, large data aggregators are beginning to explore ways to monetize their data. Imagine if E-bay or Amazon or Experian were to leverage their years of collected insight/data about you to enable advertisers to serve you targeted ads ? Key to enabling all this are the emerging category of providers called data exchanges like Blue Kai.

In summary, 2 things are happening:

1. Data and inventory are both becoming liquid and in the near future will be traded independently vs. being bundled by large publishers like Yahoo and MSN. This should level the playing field for small publishers and enable them to take fully capture the value of their audience. On the flip side, it reduces the relative value of large publishers that bundled inventory and data like Yahoo and MSN. One more reason that Jerry should have sold to Microsoft:-)

2. The myth of data driven adverting is slowly becoming a reality, creating a host of opportunities for advertisers ? More about this in my next post!

If you have an interesting idea or company in this space, do drop me a mail to tell me about it at agarg(at)foundationcapital.com

12:29 AM GMT  |  Read comments(2)

December 08

After Mumbai

This editorial in the Economist on the Mumbai attacks is worth a read - http://www.economist.com/after mumbai

India needs to put a stop to the violence, but doing so will require more than rhetoric and saber rattling. India needs to address some of the fundamental issues discussed in my last post. Poverty and despair are fertile  breeding grounds for terrorism, and we need to put an end to both. At home, and in Pakistan.

Even as we grieve the Mumbai attacks, we must ask what next? Terror appears to strike at will (in India), and with each instance raises the stakes. We must move forward, and we must respond. In fact, as Rahm Emanuel, White House Chief of Staff-designate, loves to say “we should never allow a crisis to go to waste".

The press and blogosphere have been full of suggestions on how to respond, ranging from attacking terrorist camps in Pakistan, to improving intelligence and adding NSG squads in multiple locations. Many of these suggestions make sense and there is clearly a need for India to improve its intelligence and rapid response capabilities. Improving intelligence is probably the only way to prevent (or at least reduce) attacks on soft targets. No amount of security can do so - after all, how many hotels and buildings can you protect? We need to improve our electronic interception and surveillance capabilities and strengthen coordination across the various agencies. We also need to improve our ability to respond - having NSG units in multiple locations in a good start. Training them for urban situations (it amazes me that 500+ men were deployed against 10 terrorists), and leveraging technology (e.g. 3-D digital models of target sites and expanded video surveillance to help plan counter-attacks) would also go a long way. There is a lot that can be done on both these fronts and many of the suggestions in the media make sense.

Yet, this is only a start. Unlike conventional warfare, there is no way to win a war against terrorists. India must strike at the root causes of terrorism if it is to win. India must arrive at a negotiated settlement on festering issues like Kashmir and Ayodhya and provide secular education for all children.

Settling the Kashmir and Ayodhya issues will strike at the heart of terrorism. The on-going violence in Kashmir and the strong emotions associated with Ayodhya are a god send for terrorist recruiters. For every person who is killed in Kashmir (irrespective of who was at fault), we are creating several new terrorists (or at best terrorist sympathizers). Every riot/pogrom fuels extremism and gives birth to future generations of terrorists. India needs to stop the vicious cycle by leveraging this crisis to force a settlement with Pakistan on Kashmir (formalizing the current borders) and negotiate a compromise to the Babri Masjid/Ayodhya issue. Doing so will prove much more effective than increasing the number of NSG commandos by a factor of 10.

Furthermore, “secular school education for all” will reduce terrorism’s oxygen supply. Multiple surveys indicate that (on an average) Muslims in India are a disadvantaged lot. And one only has to look at the Palestinian camps to see that poverty and the lack of opportunity are breeding grounds for terrorism. To add fuel to the fire, we have children being brainwashed by religious schools of all flavors. India needs to instill hope and provide more opportunities for its poor (Muslim or otherwise). And India must provide each child with a secular education that teaches them to be Indian first and Hindu or Muslim second. Religious schools (madrasas or otherwise) are not consistent with a secular democracy, and must go.

It will take a while for us to put the events of last week behind. But we can make a start by addressing some of the root causes for terrorism – Islamic or Naxalite or Hindu.

12:19 AM GMT  |  Read comments(0)

November 30

Be wary of trading analog dollars for digital pennies...

Jeff Zucker (CEO NBCU) was rightly concerned about his analog franchise when he made this comment earlier this year.  After all, the five major broadcast networks raked in $9.23 B in this year's upfront market (vs. $9.15B last year) despite their audience declining 11% (average across the five networks for the first 4 weeks of the new season). TV seems to be one of the few businesses that can demand (and get) more money while delivering less. Why is that ? Scarcity!

Brand advertisers need reach, and with media fragmentation, the broadcast networks remain amongst the few means to reach large national audiences in one fell swoop. The networks play this card to the hilt and advertisers don't want to be left out. The analogy that comes to mind is how CIOs never got fired for buying IBM. Yet, IBM feel from its perch and the same could happen to the networks. The networks will not go away (nor did IBM), but they could soon be trading their analog dollars for digital pennies.

Why is that ? Because the scarcity of ad inventory may be short lived.

1. On-line video inventory (for premium content) is like to grow dramatically In the IP based video world, content is on demand, and viewer ship is likely to explode as more premium content is available on-line, concurrent media consumption increases (according to Mindshare, millennials can cram 20 hours of media consumption into 7 hours of media time), and people come to terms with the flexibility that sites like Hulu and Fancast offer. For more on this, see earlier post - http://ashugarg.spaces.live.com/Can Judy McGrath plug a hole in the dam with her finger ? as well as http://ashugarg.spaces.live.com/On-line Video: When will the networks get it ?
2. YouTube! YouTube attracts 300 M monthly visitors (an order of magnitude more than the most popular TV show). Today that inventory is not being monetized because of concerns about context, brand, etc. But that will change. And when it does, advertisers will have to decide how much of a premium to pay for TV ads when they can reach the the same (or larger) audience and show interactive ads on the web.
3. The cost of producing content is declining exponentially and distribution (via the Internet) is getting easier. As a result there is likely to be an explosion of new age content production houses like Rev3 that will offer brand advertisers a low cost alternative to network TV, but with many of the same benefits (well defined audience, editorial control, predictability, etc.).

And so, Jeff is right to be worried about trading analog dollars for digital pennies...unfortunately, worrying about something does not mean it will not happen.

12:08 AM GMT  |  Read comments(0)

November 29

Not ye olde banners...

Like many of you, I spent most of the last 30 hours glued to the TV/computer watching the events in Mumbai unfold. Like some of you, I had family stuck in/around the attack sites (all of whom have escaped safely). Through this period, I went through a variety of emotions - surprise, fear for my family, and anger.

Anger that terrorists have defaced the city that I consider home. Anger that terrorists have killed more than a hundred people who were going about their business. And anger that these terrorists may have been supported by Pakistan.

At the same time, I could not help but think that...

1. We (India) should respond in kind. Yet, history has shown that bombing other countries is hardly a recipe for success where containing terrorism is concerned
2. The Indian intelligence agencies failed. We need an FBI like Pan-India agency to fight terrorism. But, the FBI failed to prevent 9/11 and Scotland Yard failed to prevent the London bombings
3. The police took too long to react. they should be better armed. We need NSG like (India's elite counter terrorist squad) commando teams in Mumbai ? But India had several dozen cities. How does one decide where to draw the line ?

I could go on and on. We all have lots of questions. Unfortunately, there are no easy answers. India will and must respond. Hopefully, as we do so, we will heed the lessons of history, and not over-react to these horrific attacks.

Also, while I know that Mumbai has changed for ever, Mumbai and India will march on.

10:28 PM GMT  |  Read comments(0)

November 26

Can Judy McGrath plug a hole in the dam with her finger ?

The latest issue of the Economist has an article about Judy McGrath’s valiant efforts to turn around MTV Networks (http://www.economist.com/people/displaystory.cfm?story_id=12633125). She seems to be doing all the right things – expanding internationally (where traditional media continues to grow), has bought a casual gaming site (addictinggames.com)and even lunched a hit video game (Rock Band sold 7 M copies). Yet is it enough? Can it compensate for the fact that MTV’s number of  viewers declined by 28 % in Q3 (in the 12-34 demo YoY)?

The fundamental problem is that MTV’s audience is the on-demand generation – spending time on Facebook and MySpace and watching video on-line vs. watching TV. Does all that Judy is doing amount to plugging a hole in the dam with her finger ?

If you are reading this blog, you probably already know that on-line video is exploding. The question is what does that mean for broadcast TV, especially as the MTV generation grows older? Will there be broadcast TV 20 years from now?

Just in case you are one of the skeptics about on-line video, here is some interesting data from comScore and the Leichtman Research Group. In July 2008,

• More than 142 million Americans watched an average of 80 videos each and in aggregate viewed more than 11.4 billion on-line videos for a total duration of 558 million hours. Google attracted 44 % of the views with 92.1 million people watching an average of 55 videos each on Google. By comparison, all of Viacom (not just MTV) attracted 1/20 the views of Google (2.2%)
• 31 % of adults online at home viewed video online at least weekly, compared with 25 percent that said so last year. Some 10 % of online adults said they view video online daily
• Among those who watched video online, 9% strongly agreed that they now watch TV less often and 4% strongly agreed that they would consider disconnecting their TV service to only watch video online

Furthermore, just remember that the First Law of Technology says we invariably overestimate the short-term impact of new technologies while underestimating their longer-term effects.

12:55 AM GMT  |  Read comments(1)

November 22

Follow up to Ad Networks: Dead or Alive ?

I got a lot of constructive and encouraging feedback about my post - Ad Networks: Dead or Alive? In general, there was consensus that a shake-out is likely. In order to survive, ad networks will need to determine where/how they add value beyond aggregating inventory as that function is being usurped by the exchanges.

In this and subsequent posts, I will attempt to address some of the questions/issues raised.

First, there are several questions asked about how the business model of the ad networks that survive will evolve. As indicated earlier, I think that there will be 3 broad buckets. 

1. The largest (i.e. top 10) ad networks will expand beyond display into various forms of digital media, and possibly even cable/broadcast and will begin to disintermediate agencies. In some cases, they will even begin to out-source marketing through a performance based contract. The analogy that comes to mind is the ODM business model whereby companies like Flextronics have taken over entire manufacturing functions for large consumer electronics companies. Also, many of these ad networks will begin to provide exchange like services, opening up their inventory to automated bidding systems.
2. The second category is the “hedge fund” model. This is an evolution of the current arbitrage based lead generation model whereby networks with proprietary data or algorithms begin to build a book of business buying and selling media across exchanges similar to the ways funds buy and sell stocks. Key to scaling model will be the development of more sophisticated exchanges which offer futures and options which is a few years away (IMHO).
3. In addition, I think there is still a role for boutique ad networks that can offer something unique. What does that mean ? Well, here are some possibilities.
   1. I think that a GLAM like model could survive if it can continue to offer “portal scale” reach, maintain a high level of quality control (in terms of publishers, type of content, etc.) and provide a 20-30% discount on portal prices
   2. Monetizing social media continues to be a challenge. If networks like Lookery can find ways to decipher purchase intent across social networks/blog networks and create category focused inventory bundles, I think they can build a viable business
   3. New media types will have unique requirements that exchanges will not be able to meet and individual publishers will lack scale to sell directly. In these cases, there will be role for ad networks. In game advertising, 3-D advertising, and possibly even video could fit into this bucket. The obvious question here is how big these business can become ? And how they sustainable are they ? For example, will there be stand alone video ad networks 3 years from now ?

The second broad theme in the questions was around what other business models will emerge as a result of this transformation. What is the opportunity for entrepreneurs ? I think that exchanges will create a whole new eco-system around them. Possibilities include:

1. Data networks/exchanges like Blue Kai and OpinMind
2. Buy side brokers/optimization service providers like Rubicon Projects and Pubmatic
3. Sell side brokers/advertiser facing service providers like Teracent, AdBuyer and Invite Media
4. Marketplaces that enable free lancers/boutique firms to access buyers of services such as creative construction and analytics/reporting

There are probably a host of other products and services that can be wrapped around exchange platforms. If you have any ideas, I would love to hear about them at agarg(at)foundationcapital.com

7:27 PM GMT  |  Read comments(0)

November 12

Ad Networks: Flight to Quality ?

Its funny how quickly perceptions change. A year ago, ad networks could do no wrong. Investors funded several hundred of them. Cox bought Adify, an “ad network in a box” technology provider that helped spawn many of these networks for $ 300 M in April 2008. Now, ad networks are suddenly considered the living dead, a sentiment cogently articulated by none other than the WSJ a few weeks ago in this article - http://online.wsj.com/article/SB122514803617173825.html

Clearly the climate has changed and advertising spend is expected to decline in 2009. At Web 2.0, Mary Meeker predicted that a flat GDP would result in a 4 % YoY decline in ad spend. And on-line spend cannot be isolated from that, especially since financial services, automotive and retail are the 3 largest spenders on-line. 

Yet, lets not forget that ad networks serve a critical purpose. They aggregate inventory across 1000s of publishers and re-package that inventory to create “advertising products” (as distinct from engineering products) that target specific advertiser needs. They enable the more than 900 B (yes, billion) impressions that are generated in the US alone to be monetized, especially in a world where the premium publishers (Yahoo, MSN, etc.) are loosing share of voice (see chart below for global minutes share data). Without ad networks, the web as we know it could not be sustained - all those publishers do need a way to be paid for their labor, and ad networks provide that. And while on-line ad spend growth is slowing, absolute spend is not declining. Advertisers continue to need help in navigating the complexity of buying on-line media, and their existing agencies lack the skills to help them.  Hopefully, by now, you are convinced that ad networks have a rightful role in the world. If not, take a minute to read a post by Chris Weiss (of Lucid Media) that makes the case for continued proliferation of ad networks - http://www.imediaconnection.com/content/20779.asp

Source: Mary Meeker's Web 2.0 presentation

Well, having hopefully established that ad networks have a rightful role in the world, I wish I could say that all is well in the world of ad networks. In fact, I do think that ad networks are under threat, but not from the economy. Rather, the advent of ad exchanges like Right Media (now owned by Yahoo) pose substantial (and more fundamental) challenges for the business model of ad networks - by increasing transparency and liquidity in the marketplace, they raise the bar for networks, in terms of the value they need to add to raw inventory to justify their margins. 

I do think that the networks which lack differentiated data driven targeting capabilities face a tough future. At the same time, there is role for ad networks in this new world (of ad exchanges), and many (but not all) ad networks will cross the chasm.

1. The line between large agencies and ad networks will blur with the top 10 ad networks evolving into full service advertising solution providers that will deliver to advertisers leads, sales or what ever business goal they seek to achieve at a pre-negotiated price. Doing do will require the networks to substantially expand their account management/client servicing capabilities, build deep integrations into the exchange to tap into large pools of inventory, and build structured finance like skills to create specialized ad products (where networks assume more risk in return). At the same time, marketers will need to learn to quantify their needs and to set and manage relationships where they out-source more and more of their responsibilities. Maybe, marketers can learn a thing or two from their supply chain counterparts! Do take a minute to read this post by Tom Hepos that also makes the case for ad networks becoming more like agencies - http://www.imediaconnection.com/content/20912.asp
2. Some of the smaller tech savvy ad networks will morph into hedge fund like businesses whereby they apply unique algorithms to capture arbitrage opportunities. These are likely to be boutique businesses as markets become more efficient, and the next generation of PhDs set up their own advertising hedge funds.
3. There is a role for networks that can offer advertisers with portal like reach for targetted audiences AND contexual quality control. If  such networks can  aggregate high quality niche targetted sites/blogs, build quality control content filters/tracking and provide advertisers with transparency into where their ads are being posted, they could emerge as worthy competitors to the portals.

For the rest, well...it was a character building experience.

Equally importantly, this evolution will create new opportunities for start-ups. If you have ideas, do drop me a mail at agarg(at)foundationcapital.com

I just spent 10 days in India across Mumbai, Bangalore and Delhi, where I also attended TiEcon. It was an exciting, fun and productive trip - I met with entrepreneurs, budding entrepreneurs, investors, and investment bankers.

• While I was in India for only 10 days, the overall mood changed quite a bit during that period. The change was (I think) mostly driven by the fact that the  stock market dropped by 20%, and the SENSEX dropped below 9000 from a peak of 21000 at the start of the year. In a country where even the cab drivers wager money in the stock markets, that changes the overall mood quite a bit. The newspapers were echoing the same themes as in the US, the credit crunch, the collapse of the stock market, money market funds being under water, etc. The bad press coupled with SMS based rumor mongering about banks failing (which is a uniquely Indian phenomena) had created a panic like situation amongst the middle class, driving huge redemptions of mutual funds (in line with the "buy high and sell low" philosophy), and a run on several private banks including ICICI, which was the darling of the stock market till recently. Yet, the very same newspapers talked about how the economy would grow as 6-7 % for the next 2 years instead of 7-9 %. And everyone talked about how the worst case scenario for India is 5.5 % growth next year. Seems like a disconnect to me!
• The overall air travel experience was significantly better than recent times. The privatization of the airports is paying off. The new Bangalore airport is very nice and seems quite spacious while both Delhi and Mumbai are a step up though the continue to be relatively small, and in comparison to Shanghai feel tiny. I also tried the recently launched Jet Airways SFO-Shanghai-Mumbai flight which was quite a treat: good food, great service, even landed 15-30 mins. early! Prior to taking the flight, I was quite skeptical about the market opportunity for Indian airlines seeking to build a global business. After all, why does the world need another airlines! Post the flight, I am a convert. And then, both Jet and Kingfisher (another India airlines) canceled their SFO flights and so I am back to flying in Lufthansa.
• The traffic keeps getting worse. Mumbai is virtually impossible to move around in it, taking 2.5-3.5 hours in the evening to drive from South Mumbai to Bandra, which is a less than 10 miles. The new Bangalore airport takes several hours to get to from most parts of the city and Delhi traffic has taken a turn for the worse, especially wherever there is metro construction underway, which is most of South Delhi.
• Health consciousness on the rise. At Gold's gym in Bandra (a well to do and fashionable neighborhood), I saw the usual suspects - incredibly good looking young men and women pumping iron, and the occasional westerner (Bandra is the hub for expats living in Mumbai). What was a pleasant surprise though was that I also saw several middle class housewives pounding away on the elliptical, and boxing with their trainers.That was a first for me! In addition, a couple of restaurants had low calorie meals and quite a few people I met seemed to be running in the upcoming Delhi half marathon.
• Having your cake and eating it too. The day I landed, Jet Airway announced that it was laying off 850 flight attendants to reduce its bleeding, and there was an uproar. Local politicians threatened to prevent Jet flights from taking off in protest, the media was full of crying air hostesses, and several ministers came to the rescue. Within 24 hours, Jet re-hired then all back. Ministers publicly warned Jet and Kingfisher not to lay off employees or be prepared to face the consequences. We (Indians) want the benefits of capitalism, but are we willing to live with the consequences ? 

All in all, the energy of India continues to be intoxicating. As a friend said one evening, "growth is a drug like cocaine, once you are addicted you need more and more and..."And it was clear that everyone is still on a high...

11:45 PM GMT  |  Read comments(1)

September 24

A Proud PC User...

I recently left Microsoft after having worked there for more than 4 years and returned to the valley (to join Foundation Capital where I invest in both the US and in India related opportunities ). One of the hardest adjustments was to see how "uncool" it had become to be a PC user, especially one that thinks that Vista is OK and actually likes Office 2007! I often found myself being defensive. Now, thanks to the new PC ads (not the Jerry Seinfeld ads), its feels OK to be a PC user again. The "Proud to be a PC." ads celebrate being  ordinary...ordinary people doing extraordinary things...

I think the ads are brilliant. The Apple ads will be etched in my mind for ever. They struck a chord. And so do these. In one simple 30-second spot, they remind us of why we use a PC -- and that the majority of PCs are powered by Microsoft. Positioning a technical product as widely used at Windows is a challenge and after spending years and several hundred million dollars, Microsoft seems to be getting there...If you haven't seen the ads, do take a look at http://www.youtube.com/watch?v=cRg_Uh1AO2A and tell me what you think.