Advertisement as Signaling
Advertising is the communication of a signal (information) with the express purpose of gaining advantage over one’s rivals in a competitive environment. That’s my definition of advertising in the broadest terms. In those terms, advertising is not limited just to the world that is human made. All sorts of life forms over evolutionary time scales have been in the business of advertising. In the natural world, advertising serves a very useful function, the end result of which is the astonishing appearance of beauty and diversity. A beautiful human face is not just an advertisement but also the result of successful advertising. Little wonder that beautiful people appear in advertising pictures.
A competitive environment is a must for there to be a need for advertising. If you are the only game in town, you don’t have to waste resources signaling that you are the best because you are the only one and therefore necessarily the best. It is only when you have competitors—whether in mating or in selling soap—that you have to signal that you have better genes than the others or that your soap cleans up the best. As the environment becomes more crowded, greater competition pushes up the need for more effective and efficient methods of signaling.
I have been thinking about advertising a bit lately. The world economy is growing rapidly and competition is heating up. Since advertising involves the communication of information, the sophisticated information communication technology available has made signaling cheaper. But that has the downside of too much information being pushed towards the receiver and instead of signal, the aggregated signal degrading to noise and the receiver blocking all signals. Information overload is going to be a major problem faced by people.
Perhaps that is not entirely true. One can never have too much information if information is defined as valuable data. It is an overload of irrelevant data—with respect to the receiver—or noise, that is the problem. Because information is lost if there is too much noise, and irrelevant data is noise, that we have a problem. Our challenge is to block the noise so that the signal can get through.
The economics of advertising is fairly simple. Consider any product, with its associated cost of production. That cost defines its supply. Then there is a perceived benefit from the use of the product. That benefit defines its demand. Since price is jointly determined by the interaction of supply and demand, if demand were increased, a higher price can be obtained and this means a larger profit for the producer. We must also keep in mind that profit is the difference between the price and the cost. Advertising increases the perceived benefits of a product, and therefore its demand. Firms compete to increase their profits through increased revenues and decreased costs. Competition has two distinct effects. On the demand side, competing firms induce an increase in demand with the goal of pushing up the price. On the supply side, firms try to lower their costs to below their competitors level. The interaction of these two drives in a competitive environment both prices and costs down.
This is plain Econ 101. Competition makes the world go around. If there were no competition in the marketplace, you would get only shoddy goods and services. If there were no competition in the natural world, we would not be here. Evolution is powered by competition between selfish genes. We, and everything that lives, are the product of that competition in the natural world. The diversity of goods and services we take for granted and enjoy is the product of competition in the commercial world.
Back to the economics of advertising. Part of the profit that a firm earns, it earmarks for advertising. That is a cost. So if the cost of achieving an additional sale through advertising is lower than the additional profit arising from that sale, it makes sense to advertise. In concrete terms, if it costs me Rs X to advertise to get an additional sale of my super-duper widget, and my revenue from that sale is Rs X+Y, then my net profit is Rs Y, then I should go for it. Advertising costs are critical in the profit calculus. If that cost can be lowered, profits go up.
Information is communicated over a medium, or a channel. That channel can be cheap or it can be expensive; it can be broad or it can be narrow; it could have a high fixed cost and a low marginal cost; or it could have a low fixed cost and a high marginal cost. Take the town crier, for instance. It is cheap where labor is cheap; it is narrow, as voice can only reach a limited number of people who are within ear-shot; it has low fixed costs but high marginal cost (for each lot of people, he has to put in the same effort in delivering the same message). A TV channel, on the other hand, is broad (reaches a lot of people); high fixed cost (very expensive to build a TV station); low marginal cost (it does not cost much more to reach more people).
Let’s keep in mind two facts. First, advertising is valuable information and serves a crucial function in a market economy. Second, people pay for information. Advertising inherits all the properties of information goods. Since the channels for the transmission of information are becoming broader (think broadband and cell phones, not just print, TV, and radio), we are getting overwhelmed by advertising and the challenge is to reduce noise and increase the signal.
Increasing the signal implies that you as an advertiser have to figure out the characteristics of the receiver so that only those bits reach the receiver that is meaningful. Traditionally, in the print media, the information is broadcast and the receiver searches through the pile and takes the bits that matter. For example, the classifieds in a newspaper: you ignore all the ads except the ones listing apartments when you are looking for an apartment.
Advertising on print is expensive because the “real-estate” is limited. What happens when that limitation is removed, as in the case of web pages? The cost of advertising comes down, and therefore the volume of advertising goes up. To get the attention of the receiver, then, you have to “profile” the receiver and only send those bits that are significant. The receiver’s attention, rather than printed space, is the scarce resource.
The main point so far is that in a world where advertising is plentiful because transporting bits is cheap, and where there are zillions of firms competing to sell their wares to you with your limited attention, the problem of how to get your signal across all the noise needs to be solved.
Now we address a different matter. Advertising not only uses a channel but in many instances actually makes the channel commercially sustainable. The cost of production of a newspaper is much higher than the price it sells at. The difference—and a bit more for profit—is made up by advertising. In the case of commercial TV and radio, all of the costs are met by advertising and you receive the content for “free.” I write “free” because actually you do pay for the advertising by paying a bit more for the product than you would have had if the firm had not incurred advertising costs. There is nothing really free in the universe. Everything has a cost.
Up until recently, the main channels of advertising have been (in chronological order) print (newspapers and magazines), radio, and TV. The economics of radio and TV have been such that they can be entirely supported by advertising revenues. Briefly, these have high fixed costs. The infrastructure investment required is large. But once the fixed costs have been paid for, the marginal costs of serving an additional customer is very close to zero. In any large market, such as you would find in any urban area, the average cost of transmitting information is very low. It is low enough that the marginal revenue from advertising is lower than the marginal cost of the advertising. In the case of radio and TV, the total advertising spend exceeds the total cost of providing the channel and therefore the user gets content delivered free.
This is what we call marginal analysis. Around late 90s, another channel was added to the mix: the World Wide Web. made possible by the internet. The economics was similar: high fixed costs but very very low marginal costs. Because of very low costs of transmitting information, advertising again supported the cost of content delivery. Not just content, a plethora of services became commercially sustainable without the use of user fees. What began with free email (hotmail comes to mind) has now graduated to be a host of services which includes everything from word processing to social networking services. Google is the biggest game in town among other giants such as Yahoo! Brought to you by the advertising dollars of a gazillion sellers.
The newest kid on the block is mobile phones. Phones are essentially one-to-one communications devices. They are not a broadcast channel unlike newspapers, magazines, radio, TV, and the internet. Therefore, the question is whether the economics of phones—which basically means mobile phones—render an advertising supported free mobile service possible like free TV and radio services.
The answer depends on the cost of mobile communications. That is, if the cost per bit of information communicated on that channel drops below the marginal revenues from advertising, then advertising will make mobile communications free and the users will only have to pay with their attention, not money.
The economics of mobile phones is simple to state. The infrastructure required has very high fixed costs (although lower than landline infrastructure) but have very low marginal costs. In other words, the technology exhibits large scale economies due to very low marginal costs. Therefore the larger the market served, the lower is the average cost. Which partly explains why it is possible to price mobile communication so low in economies such as India. As the installed mobile user base grows (which is right now growing at an incredible 6 million subscribers a month in India), the average costs will continue to decrease. At some point, the average cost will become low enough that given a suitable advertising model, mobile communications will break free of the need for user fees altogether.
Free Mobile Communications
So far I have been painting the landscape. Now to place the object of my attention in there, which is, the matter of how and under what conditions will mobile phone use become fully advertising supported just like radio and TV.
The conditions first.
- The average costs have to come down below the average revenues from advertising. Technological progress will lower the fixed and marginal costs. Together with that, increased installed base will reduce average costs. This is inevitable.
- Average revenues from advertising have to increase. This will happen only if advertising information is delivered more efficiently. Merely broadcasting more advertising over the mobile channel will degrade the channel with too much noise and very little signal for the user. Targeted advertising is required.
Which brings us to the how. In the traditional broadcast media, the advertisers choose the content around which they will push their advertising message. The users then select from the available content and thus self-select which advertising messages they receive. Thus, Home Depot chooses to advertise on a home improvement show, and Nike advertises around the Super Bowl. Depending upon the choice of the show, the user reveals his profile and thus receives the advertising message. The user thus self-selects the message he or she wants to receive. Both the advertiser and the users meet around the chosen content.
In the case of a non-broadcast channel such as the mobile phone, the advertiser does not have the freedom to choose the content since the content is not known to the advertiser. The advertiser therefore has to somehow choose the user to whom they will send the message. For this, the advertiser has to know the profile of the user. Getting the user profile is therefore the most critical bit when it comes to advertising on the mobile.
There are broadly two ways of getting the user profile. One is to somehow give an incentive to the user to reveal his or her profile. This is difficult and costly as the users have an incentive to misrepresent who they are. The other is to opportunistically build the user profile by keeping track of what matters are of interest to the user. This is more accurate and I will call this method the “revealed preference” method of user profiling.
The revealed preference method is really an analog of the method that is used on radio and TV. The users subscribe to certain published content on the mobile web. The mobile web is an analog of the more familiar www with the important distinction that the mobile phone is the access device not the PC. So if publishers publish interesting content, and users choose which content streams to subscribe to, then once again advertiser and users can meet around those content watering holes.
That is what the whole (watering hole?) idea behind what Rajesh Jain calls the “publish subscribe mobile web.”
To end this rather long piece, allow me to make a prediction. In the not too distant future, we will have the freedom to choose whether we wish to pay for our mobile calls or have it free. Soon enough we will be given a free mobile phone and be allowed to make and receive calls for free provided we are willing to pay with our attention to short advertising messages while using these phone. For instance, when we answer or make a call, before we get connected, we will hear the advertising message. This message will more than likely be of some interest to us because the advertiser would have information about who we are based on what content we subscribe to from the ocean of published content.
As for me, I will probably just choose to pay for my mobile calls and avoid the advertisements. I just don’t need any more information. My brain is bursting with too much information already. 🙂