There is a short entry in today's Financial Times about Facebook and the uncertainty surrounding its IPO. Essentially, people worry that Facebook may not have that much upside potential. Sure enough, everyone sees the spectacular growth of membership: now around 800 million users and forecasted to be one billion by August 2012. The doubt is more about profitability. At nearly a billion users the market expects Facebook to be close to scale and, therefore, hugely profitable. Yet the company's profitability is average. According to the article, its revenues were 4-4,5 billion dollars in 2011. A useful benchmark here is Google of course, Facebook's biggest rival. Google's revenues stand at over 40 billion dollars, an order of magnitude higher. As an investor, there is reason to worry.
These simple heuristics hide a more profound consideration however, which is related to the rate of innovation in advertising technology. The reason Google is so profitable is because of it's killer search advertising technology. This is a huge innovation compared to classic advertising formats. A similar potential innovation may happen soon for social networks, where the opportunity consists in harnessing word-of-mouth on a network of connected friends and relatives. If this problem is cracked, Facebook has tremendous advantage given its comparative size (just as Google benefits from its size in search advertising). The uncertainty is about when such innovation in 'network marketing' will happen? If a 100 years of media history is any guide, it is wise to remain patient.
Monday, 30 January 2012
Thursday, 26 January 2012
Media Bias and Advertising
Media bias is a very interesting topic. Much recent research tries to figure out what causes media bias (slanting) and what is responsible for large differences across countries. There are many theories, some obvious (e.g. the media is owned by a mogul with strong political interests - such as the case of Silvio Berlusconi or Rupert Murdoch), some more subtle (e.g. the idea that people do not consume media to discover the truth but rather they look for confirmation of their existing political biases - see Chapter 4 of my book). A recent paper in the Journal of Marketing Research by three researchers at the University of Pittsburgh brings an interesting perspective to this debate by showing how the nature of advertising influences media bias.
Imagine a scenario where a substantial part of media revenues come from advertising. Assume further that consumers' product preferences are correlated to their political views (e.g. republicans might be more likely to buy Chevy trucks and Dell computers while Democrats may be more inclined to buy Toyota and a Mac, to just mention two categories). For many product categories this is a plausible scenario. If this correlation is high then advertisers are inclined to advertise in media outlets that have bias correlating with their customers' preferences because, in this way they can better target their messages. Such "differentiation" in the placement of ads (called single-homing by the authors) is likely to soften price competition. In response to this incentive, the media will become biased because they can have monopoly power over the advertisers. In the opposite case, when such correlation is negligible, firms tend to place their advertising in all media (multi-homing), which then provides no incentive for media firms not to differentiate on content (as it doesn't help attracting advertisers) and as a result makes media firms less likely to slant.
Of course a lot depends on what we assume is the underlying need of consumers, what proportion of media revenues come from advertising vs. subscriptions, what is the nature of competition for eyeballs across product categories, the cost of slanting and so on. Nevertheless, this is an original idea and shows how complex the media industry is with its vast ecosystem of industry players and two-sided markets.
Imagine a scenario where a substantial part of media revenues come from advertising. Assume further that consumers' product preferences are correlated to their political views (e.g. republicans might be more likely to buy Chevy trucks and Dell computers while Democrats may be more inclined to buy Toyota and a Mac, to just mention two categories). For many product categories this is a plausible scenario. If this correlation is high then advertisers are inclined to advertise in media outlets that have bias correlating with their customers' preferences because, in this way they can better target their messages. Such "differentiation" in the placement of ads (called single-homing by the authors) is likely to soften price competition. In response to this incentive, the media will become biased because they can have monopoly power over the advertisers. In the opposite case, when such correlation is negligible, firms tend to place their advertising in all media (multi-homing), which then provides no incentive for media firms not to differentiate on content (as it doesn't help attracting advertisers) and as a result makes media firms less likely to slant.
Of course a lot depends on what we assume is the underlying need of consumers, what proportion of media revenues come from advertising vs. subscriptions, what is the nature of competition for eyeballs across product categories, the cost of slanting and so on. Nevertheless, this is an original idea and shows how complex the media industry is with its vast ecosystem of industry players and two-sided markets.
Friday, 13 January 2012
Academic literature on new media
Recently a colleague of mine has asked me what would be the key academic papers on the topic of "new media". It took me a few hours to think through the request not because I didn't remember the papers but rather because I had a hard time to categorize them around themes that are relevant for the topic.
Finally, I have settled for the following list of papers and categories. No wonder, the list is heavily biased towards my research. Sorry for this.
1. The first category deals with the topic of Word-of-mouth (WOM). The important papers here are marked by Dina Mayzlin, a colleague at Yale. She has a few publications in Marketing Science (a recent one alone, one with Chevalier in 2006 and another with Dave Godes 2004).
Btw, this is a very technical empirical paper but quite profound.
Finally, I have settled for the following list of papers and categories. No wonder, the list is heavily biased towards my research. Sorry for this.
1. The first category deals with the topic of Word-of-mouth (WOM). The important papers here are marked by Dina Mayzlin, a colleague at Yale. She has a few publications in Marketing Science (a recent one alone, one with Chevalier in 2006 and another with Dave Godes 2004).
2. Then, I'd have another set of papers about links, link formation and how this relates to content and content building. Important papers here are:
- Katona and Sarvary 2008 (Marketing Science)
- Mayzlin and Yoganarasimhan (forthcoming in Marketing Science)
- Stephen and Toubia 2009 (JMR).
- Also the job market paper of Liye Ma from CMU. Here is a link:
http://www.rotman.utoronto.ca/userfiles/departments/marketing/File/Ma%20PRESENTATION%20PAPER.pdf. Btw, this is a very technical empirical paper but quite profound.
3. The third category would consist of a set of papers about networks and social communities. Here, the Economics and Computer Science literatures are quite important to lay down the foundations. The classics are:
- Brin and Page 1999.
- Bala and Goyal 2000, (Econometrica).
- Jackson has a good review in a book Chapter in 'Group formation in economics' 2004, (Cambridge U. Press).
- I also have a paper on social network formation with Katona and Zubcsek (2010) in JMR.
4. Finally, I'd have papers about content in old and new media. Here I have some research but again it is a bit too biased (our recent working paper with Kaifu Zhang, and a previous paper in Marketing Science with Yi Xiang, 2007). Also relevant is Tuba Pinar Yildirim's job market paper on user-generated content in new media.
I would probably omit the literature on search advertising and position auctions. This is a big body of work both in Economics (e.g. Varian's paper) and marketing (Katona and Sarvary, and Yao and Mela) but not really about new media (rather specifically about search advertising).
Subscribe to:
Posts (Atom)