Customer value and the network effect

What’s the value of a customer who doesn’t pay you anything? If you’re running a hot dog stand, the answer is probably “zero.” But if you’re running a two-sided market – a market, like eBay or Monster.com or AdWords or YouTube or Digg or even Second Life, that needs to attract both buyers and sellers (or content generators and content consumers) – the answer may be “a lot.” EBay, for instance, earns most of its money from its sellers, who pay the company a fee whenever they sell something through the auction site. The buyers don’t have to pay when they make their purchases. But while eBay receives no direct revenue from the buyers, the buyers nevertheless represent a crucial set of customers for the company – without buyers, there’d be no sellers and hence no business.

Clearly, each buyer in such a networked business has value to the company – but how much value, exactly? That’s where things get tricky. Because traditional approaches to determining the economic value of a customer don’t work when the customer isn’t generating any direct revenue, there hasn’t been any good way to estimate customer value in these sorts of two-sided markets. That means companies have to fly blind when determining how much they should be spending on marketing to attract the non-paying customers. And that, in turn, likely means they’re spending either too much or too little.

But a recent paper by three business-school professors – Sunil Gupta, Carl Mela, and Jose Vidal-Sanz – offers a new approach for estimating the value of nonpaying, or, as the professors term them, “free,” customers. The authors created a mathematical model of a hypothetical firm, with a business similar to Monster’s, and used it to calculate how much every new buyer joining the company site is worth and how that value changes over time.

Some of the results are fascinating. The professors found, for instance, that the value of each nonpaying customer (buyer) was actually slightly higher than the value of each paying customer (seller) – even though there were far more buyers than sellers in the company’s marketplace. (To put it another way, the network effect of a buyer on a seller was far stronger than the network effect of a seller on a buyer.) The research also demonstrates that it’s possible to estimate optimal marketing expenditures as a two-sided business grows. While heavy markering spending is required in the early days to attract a critical mass of buyers, the network effect itself becomes a larger attractant than marketing as the business grows, allowing a company to cut back its marketing budget over time. Knowing the optimum spending amount with some precision at different points in time would help businesses maximize their profits. The information would also, the authors argue, allow a company’s founders, managers, and investors to gain a more accurate understanding of the firm’s overall value.

In an interview about the study, Gupta notes that the model applies only to fairly simple two-sided markets. But on the Net today, of course, nonpaying, “free” customers are also critical to other businesses with more complex network structures, from YouTube to MySpace to Skype to an open-source software company like Red Hat or MySQL. If you have a “community,” you likely have “free customers.” Gupta says that he’s currently

working on understanding and modeling complex network structures such as those of MySpace. Here the issue that we are grappling with is the tangible and intangible value of customers. In other words, customers provide tangible value to a firm through direct purchases but they also provide intangible value through network effects or word of mouth. It is quite possible that some customers have low tangible but high intangible value. Traditional models would label such customers as low value and would miss a huge opportunity for a firm.

This promises to be a particularly fruitful, and practical, area of study in the years ahead.

7 thoughts on “Customer value and the network effect

  1. babken

    Thats an interesting angle on the business plans for e-businesses such as google and ebay.

    i feel however that there is a fundamental discrepency. the customers for ebay aren’t the buyers, they are the sellers. the buyers are the customers for the sellers. ebay as a business doesn’t on its own attract the buyers. the buyers are attracted to the sellers who in this case are on ebay. sure, it’s in ebays interest to bring in more buyers, but it seems to me that their value is greater only because there are more buyers then sellers, not because of any genuine value indicators. essentially, one “buyer” is not more valuable then one “seller”, because at the end of the day its the sellers that will create revenue for the business, not the buyers. the value of the buyer is only that of bringing in more sellers. now, im sure those harvard folks have some insane formula that will calculate that relationship between the number of buyers and an increase in sellers. but to call the buyers “customers” seems innacurate, they are mearly developing and sustaining the interest in building the buyer-seller network.

  2. Nick Carr

    babken, I think you’re actually expressing the problem – the difficulty of determining in concrete terms the economic value of nonpaying customers to a firm – that the study seeks to solve.

  3. Joel P

    And in eBay’s case, there is definitely an even larger economic value to the losing bidders. If the winning bidders are eBay’s “free” customers, the losing bidders are their “negative” customers. They don’t even pay the sellers! But they drive up the winning bids and thus the final value fee eBay collects, and they are the absolute core of the network effects that make eBay work. I just don’t know how you quantify the economic value of a customer two levels removed from ever paying you a cent.

    Now if someone could model the reverse network effect of how each time someone joins my gym, it reduces the value of my membership by making the place more crowded.

  4. enigma_foundry

    “Some of the results are fascinating. The professors found, for instance, that the value of each nonpaying customer (buyer) was actually slightly higher than the value of each paying customer (seller) – even though there were far more buyers than sellers in the company’s marketplace.”

    In transportation infrastructure planning it has long been understood that the value of a system as a whole increases exponentially with linear increases in the number of nodes. Reason is that the network becomes more useful–you can go more places. This is known, but very frequently not taken into account, for example in the case of the Saint Louis metrolink, more people ended up using it then was (too conservatively, but it is Saint Louis I am talking about) planned for.

    “But on the Net today, of course, nonpaying, “free” customers are also critical to other businesses with more complex network structures, from YouTube to MySpace to Skype to an open-source software company like Red Hat or MySQL.”

    Open Source software companies are essentially membranes that mediate between their developers and their customers. The value they add is their skill in deploying, and their integrity, which they obtain by skillfully deploying software, not writing it.

  5. Patrizia

    You can call it also Page Rank, or Search engine placement or in any other way.

    At the end your website or blog is worth for the number of readers or users.

    Just like TV.

    Audience makes the price.

    And I say the price, because what’s worth today is JUST the price, not the VALUE.

  6. Bertil

    Christ ! and I just wrote (on next post comments) that you were probably not found of formal microeconmics: how come I missed this paper so far? I’ve been craving for such insights for month, and I’d love to contribute. Simply reading the abstract made me more excited then ever.

    Academia is good for you.

  7. Ed Kohler

    I think Ebay can behave more like a retailer today since they have so many sellers. Measuring average order size, thus commissions, on traffic from various referring sources would make sense.

    Their affiliate program largely rewards new sign-ups over other measurable transactions. They must see the value of increasing their network as a primary long-term benefit even though it may have less value compared to other factors short term.

Comments are closed.