Is Internet retailing a sick industry?
The conventional wisdom today thinks that Internet retailing is overvalued and overrated; that the phenomenon of consumers directly buying over the Internet is not particularly profitable for Internet-centric retailing firms, or particularly earth-shaking for the retailing sector or the economy at large.
There is little doubt that valuation notions based on anything other than profit projections are highly suspect. I have often argued that many of the attempts at setting up .com firms without thinking through a business model and a revenue stream are quite doomed. However, in this article I will argue there are aspects in "the B2C backwaters of the high-tech world" today which are exciting, which offer important profit opportunities for some firms, and constitute an important reshaping of the modern economy.
First, let us look at the traditional argument. When Internet retailing first appeared, some aspects of it were obviously appealing:
- Internet retailers would avoid expensive store--fronts and sales persons, and hold smaller inventories.
- Internet retailers could have IT integration from their store-front to order processing all the way back to their suppliers. This has been achieved by a handful of traditional retailers (e.g. Walmart) but it could be done by every Internet retailer.
- Building a national or international presence is much easier on the Internet than is the case with physical stores.
From an economy perspective, there are three important implications of Internet retailing:
- Individuals have easier access to information when shopping on the Net. The lowered cost of information gathering is likely to go along with greater information gathering by individuals. This is good for economic efficiency.
- The capital and labour devoted to store-fronts, salespersons, inventories, etc. is freed up for other productive applications in the economy, which helps GDP growth.
- Consumers are much less likely to buy a product on the net at a price much higher than that available elsewhere. A world with prolific Internet retailing betrays fewer arbitrage opportunities in consumer product markets.
My sense is that these promises have been substantially fulfilled. Consumers buying over the Net is clearly a reality today, with very high growth rates. There is a slow process through which an individual gets access to a computer, to bandwidth, and becomes comfortable with doing purchases on the net. But once all these preconditions are met, there is little doubt that buying on the net takes place on a large scale.
Where things have gone sour is on the question of profit. Internet retailers have had to put up steep discounts in order to tempt customers into embarking on Internet shopping. Amazon offers 50% off the list price on the books which are its 100 best selling titles. This implies that Amazon suffers a loss on each of these books sold. These books (the 100 bestselling titles) may well account for the lion's share of Amazon's revenues. It is clearly hard for Amazon to turn a profit under such conditions.
The key question, then, concerns entry barriers. From a technological standpoint, is it quite easy for competing retailers to establish competing operations. If Amazon raises prices to a point where it becomes highly profitable, would customers switch to other low--profit resellers? If this happened, we would be living in an economist's ideal zero--profit world. The prospect of such a world has spooked the shareholders of Internet retailing stocks.
How does Amazon obtain loyalty? It uses its one great strength: a large number of customers. It tracks every purchase made by me. It uses this information to find other "people like me", who have purchases which are highly correlated with mine. It uses their purchases to show me things that I might like. This works uncannily well. It works fine even if you buy nothing: you just have to login.
Amazon carries this through to an uncompromising extent. Every screen of the store is altered using this information. When I walk into the "History" section of Amazon, the books which are prominently shown to me are those which "people like me" have bought. This makes it much more interesting for me to walk into the "History" section of Amazon as compared with any physical bookstore.
I can indicate my preferences for things that I have not purchased. I have every incentive to devote effort to this, since the system is then more tightly tuned to my preferences. Hence, I have taken the trouble of telling Amazon that I really liked Sunil Khilnani's book "The Idea of India" and the movie "The Matrix", and that I really disliked the book "Programming as if people mattered". The more I teach the site about me, the more effective it becomes in hunting down intellectual and cultural materials that I'm likely to take interest in. When Amazon hears me say that I really like books by J. R. R. Tolkien, it's able to say that people like me are hooked on Harry Potter.
I am now highly loyal to Amazon, since I've voluntarily fed it with knowledge about me for several weeks, knowledge which goes well beyond my purchases. I now find that wandering in Amazon is an uncannily exciting experience, especially outside my field (financial economics). I'm discovering a host of cultural materials that I had never noticed before. Now that I have done this with Amazon, walking into any other store on the web has a "dead" feeling for me. Competing stores know nothing about me, they are wasting my time by treating me like an anonymous customer. I'm loyal enough to Amazon to pay them a few percent of a retailing margin.
Looking two moves ahead, Amazon will build up a huge database about consumer profiles, which could then be used in product design. If a product is tested on a focus group, the extrapolation from the focus group to the population could be done by correlating the characteristics of focus group members against the full database.
This is explicitly the revenue model of another remarkable B2C project, which is working wonders on information gathering by consumers. This site interacts with the individual in detail and helps the individual in choosing products that he will like. The interaction is anonymous; only demographic information about the user is taken by the site.
I have used it in one setting, and was amazed to find that it led me to an answer within fifteen minutes, which I had taken weeks to find out by myself. As with Amazon, I have every incentive to take trouble in interacting with the site in detail, since the more I tell the site about my tastes and budget constraints, the more effectively it is able to help me. Its user database is the ideal foundation for scientific product design. Compared with using this database, any traditional notions of market research are sheer guesswork.
What does all this mean for us in India? The complexity of such systems implies a booming export market for Indian software services. However, firms like Amazon can easily come into India and compete against the existing, low-quality Internet retailing efforts. Indian Internet retailing firms will either need to rapidly jump to this level of quality, or they will die.
Back up to Ajay Shah's BS column