Last week Fred wrote a post on marketing that attracted a lot of comments and replies – in part because it had some “bugs” but more because it was provocative. One thing that seemed missing from the discussion (and given that there was so much I may simply have overlooked it) was the relationship between customer acquisition and network effects. I am going to use the word customer acquisition to mean the entire set of tactics (PR, marketing, advertising and even sales) that results in new customers, which for a consumer service might simply be registered users and for a business service might go all the way to paying customers.
The key idea is as follows: if your business has network effects, the all-in cost to acquire an additional customer is likely to decline meaningfully as you grow. Network effects exist when a product or service becomes more useful the more people have it. The canonical example for network effects used to be the fax machine and I still like it (for nostalgic reasons). It is easy to see how selling the first few fax machines is difficult and expensive – in fact you might have to give a few away to get any adoption at all – as the saying goes “the first fax machine was very lonely.”
As some people have fax machines and start to use them successfully with other owners of fax machines they will help lower the cost of acquiring more fax customers. Obviously there is word of mouth: “If you had a fax machine I could send this to you right now!” and what we have come to think of these days as virality (one enduser “infecting” another). The importance of user-to-user marketing has increased tremendously in a world where publishing is free, as a single customer with a large audience (e.g. Twitter following) might be responsible for thousands or more additional customers. Beyond that though there is a cost reduction (per customer) impact on all acquisition tactics.
Take PR. Without network effects PR is largely reduced to talking about stuff that gets boring quickly such as individual customer wins or new features. When you have network effects it’s because you have a network and there tend to be a wealth of stories about things that are happening in the network. Prior to the Internet that was somewhat harder to come by (how do you discover what’s happening in the fax network?) but for most web based networks this has provided a well spring of interesting stories. For example, Twitter has an endless font of PR based on stuff happening on Twitter that the company needs to spend no time or money on promoting.
Even consider sales. Without network effects your 100th sale is no easier than your 10th and may well be harder. Yes, you may have a few more referenceable customers but you are also likely to have competitors who are out there telling a similar story. You are generally reduced to arguing on the basis of price and features – we do more and/or cost less (a combination of arms race and pricing pressure). With network effects sales get easier over time – we have a larger network so you get more benefit!
That is why network effects businesses are so powerful. As you grow your cost of growing goes down. I should point out that this is on a per customer (user) basis – since you may see exponential growth the total cost may be rising dramatically, but what matter from a competitive perspective is marginal cost. I have argued in the past that GroupOn and LivingSocial (and others like them) may not have strong network effects. The fact that GroupOn and LivingSocial are both already spending meaningfully on TV advertising can be seen either as a sign that they are so profitable and have such a high allowable per customer that expensive advertising makes perfect sense or as evidence that their cost of customer acquisition is rising rapidly (admittedly since this is part of their total spend and we can’t really see their marginal spend this might still mean that they have network effects, but it still and indicator to watch closely – Facebook and Twitter did not have to do any TV advertising to grow huge).
