David Evans, Feb 14, 2013
There’s an old saying that nothing’s certain but death and taxes. I want to add something else to the list. There are only 168 hours in a week. Everything you do from sleeping to working to watching television to eating and romance you need to do in those 168 hours. That’s one tough constraint.
Now you might ask, what does this have to do with the purported subject of this column—the antitrust economics of multi-sided platforms? Plenty it turns out, as I explain in my recent paper on attention rivals.
There are many multi-sided platforms that are in the business of trying to get a piece of your 168-a-week time allotment hours. Most of them turn around and sell access to the time you are spending on their platform to various businesses that would like to reach you—like advertisers, merchants that want to sell to you directly, and companies that want you to use their applications.
Most of the online platforms are attention rivals and those are the businesses I focus on in the paper and in this column.
Facebook has secured a big chunk of people’s time. And they make money from that attention by selling space to advertisers on Facebook pages. Those advertisers hope that you’ll give them a piece of the attention you are giving to Facebook. Amazon on the other hand gets a lot of people to go to its pages to read reviews, browse merchandise, and then buy products. They share that attention with merchants who sell through Amazon. Google draws in people who are searching for things. And it auctions off real estate on its pages to companies that also hope to get people to click on ads. Apple is a bit more complicated but it is looking for attention too. It has a lot of ways of making money from the time that people spend using their phones with online advertising just being one of them. If you march down the list of online firms most of them are harvesting attention so they can sell it off to other companies for which attention is a valuable input.
These online attention rivals have to compete pretty aggressively for attention. In part that’s because that 168-hour constraint is actually much tougher. Although a AAA-personality entrepreneur I know claims to sleep only 90 minutes a day (and 6 hours a day on the weekend) most of us need our shut-eye. Many of us need to work and there are lots of other activities that are really hard to cut back on. You still can’t do everything online! So, in economic jargon, the supply of time for attention-grabbing activities is likely to be pretty inelastic.
It turns out that over the 2000s the amount of time that people spent online (this doesn’t include mobile now, read on….) doubled from about 10 hours a week to about 20 hours a week in the US. That’s not a lot when you consider the vast increase in things to do online. And in the last few years that number hasn’t increased much. That means that when some new activity is developed—say Tweeting, Pinning, or playing Farmville—a big chunk of the time people are spending on that activity likely has to come from other stuff they were doing before.
My paper on attention rivals presents at least a circumstantial case that that is exactly what’s happened over the last decade. New categories took attention from old categories, new players took from old players.
In the paper I document that there has been a considerable amount of churn of online platforms. Of the top fifty web domains in 2002 only about a quarter were still in the top 50 in 2012. 10 percent of those sites weren’t even in the top 15,000 websites. The same kind of churn takes place over shorter periods of time like 2007-2012.
This competition results in websites losing attention. Of the top 15,000 websites in 2002, 85 percent had fewer hours spent on them in 2012 than in 2002. That is despite the fact that the amount of time on the web was increasing dramatically during this period as more people got online.
It also turns out, not surprisingly, that the growth of one category of online activity ends up reducing the time that goes to another category.
This view of competition is consistent with how the companies themselves see things. I went through the 10ks of the companies that operated one of the top 50 websites in 2003, 2007, and 2011 to find out who the companies said their websites competed with. What’s interesting is that they usually identify as competitors websites that offer very different products and services. As Yahoo! put it “[w]e also compete with social media and networking sites which are attracting a substantial and increasing share of users and users’ online time, and may continue to attract an increasing share of online advertising dollars.”
Basically, what’s going on with these firms is that they are competing to get attention. Everyone is trying to come up with new features, products, or services to attract consumers and take eyeballs away from someone else. Sometimes that happens through drastic innovation where they come up with totally new things for people to do, but not always – so competition is sometimes for the market, more often it is in the market.
The implication for antitrust is that at least when it comes to seeking attention we need to focus on the competition for attention (the thing they are trying to harvest) rather than the particular products they are using to attract that attention (the tools they are using to harvest that attention). At the end of the nanosecond they are selling attention to, for example, advertisers. How they got that attention doesn’t necessarily matter.
Now, of course, you see where I’m going with this. Just to be clear, I’m not arguing that online attention rivals are all in one big happy antitrust market. They could be differentiated in the eyes of the seeker enough for it to matter, or be differentiated in the eye of the buyer of that attention for it to matter. But there really isn’t any reason to believe that, from the standpoint of analyzing competitive constraints, that competition is limited to firms that are offering ostensibly similar, or functionally interchangeable, products and services. That’s clearly too narrow a view of the market given how the competition for scarce attention actually works. This has pretty important implications for how we should analyze market definition and market power in online industries.
The paper goes into a lot more detail for those who find this interesting. I cover a few other issues:
• The relationship between online attention rivals and offline attention rivals under this analysis. I show that differentiation is important but the general principles still apply. It seems obvious now that online and offline attention rivals come together to compete with each other to some degree.
• The competition for providing attention. I show especially that with new technologies, which many advertisers (or their agencies) are using, there is likely to be significant competition between seemingly very different online attention platforms. These new technologies dynamically move advertising spending between online platforms, and in some cases between offline ones as well, based on estimated ROI.
• The implication of smart mobile platforms. There has been a recent discrete jump in time available for online activities. Many people are carrying smart phones, which enable them to engage in essentially online activities during a much larger part of the day. There’s been a huge increase in the supply of attention although much of that is being consumed by new kinds of activities. This is imposing competitive constraints on many traditional online players.
I think the paper also helps understand how online platforms and those who follow them see competition. It is widely understood that Amazon, Apple, Facebook, and Google are significant rivals. That’s mainly because one’s success comes at another’s expense, to a large degree, even though they are doing very different things.