Reverse Network Effects

Network effects exist when the value per user rises as the number of users for a particular product or service increase.  Classic examples of firms with network effects include eBay, Uber, Google, and Netflix.  Sangeet Paul Choudary has written extensively about network effects in the past few years.  His recent post examines the concept of reverse network effects.  Is there some point where adding more users decreases value per user?     He explains in this excerpt:  

Reverse Network Effects may sometimes set in with scale i.e. online networks may become less useful as they scale. I do not imply that all online platforms lose value as they grow. However, in the absence of robust curation, online platforms may lose value as they grow.  Under what conditions do online platforms lose value as they scale?  Since the participants on an online platform create value, an online platform loses value with scale when the participants it allows in OR the information/value that they create are not curated appropriately. Poor curation leads to greater noise which makes the platform less useful.

He goes on to give some examples of the ways in which increased noise can become a problem.  For example, he points out that less sophisticated participants may began to use the platform, and that might reduce value per user.   He cites Quora as one platform in which this danger may arise.  Quora worked very well in the beginning as experts answered interesting and challenging questions.   As people with less expertise use the system, however, may create a problem.  As their answers offer less value, some experts may leave the platform.  As experts leave, that can create a downward spiral of value loss and further expert attrition.   His work in this area is interesting, because it stresses the fact that more is not always better in the realm of network effects.  A reverse mechanism can begin to take hold on certain platforms.  

Why Lock-In and Switching Costs are More Substantial at LinkedIn than Facebook

Network effects exist when the value per user rises as the number of users increases.   In a situation where strong network effects exist, we often see high switching costs and a lock-in effect.  Take eBay, for example.  Why has eBay remained so dominant for so long in the online auction market?   Well, compare the value per user at eBay to the potential value per user at a startup that might challenge them.   Users place a high value on eBay because the existence of many fellow buyers and sellers makes it likely that they can find someone with whom they can engage in a mutually beneficial transaction.   They would have a much harder time making that match on a new auction site with  a limited number of users.  For that reason, the network effects create switching costs for users, and a lock-in effect emerges.   That's all good news for an incumbent player with high market share, such as eBay.

Ok, so clearly LinkedIn and Facebook benefit from strong network effects.  Does that mean that they will remain as dominant for as long as eBay has in the auction market?  I would argue that Facebook and LinkedIn face quite different levels of switching costs and lock-in effects.   Consider the cost of switching from Facebook to Instagram or Twitter.  What's keeping you on Facebook?  Well, it's probably that circle of fairly close friends and relatives with whom you keep up to date and exchange photos and messages via the social network.  Suppose that a critical mass of that small circle of friends defects from Facebook to another social network such as Snapchat or Twitter.  Would it be costly for you to switch as well?   Well, as long as a critical mass of friends has left, then your switching costs are rather low.  Yes, you have other acquaintances who might still be on Facebook, but are they really going to keep you there?  Perhaps not. 

Now compare Facebook to LinkedIn.  What's keeping you from potentially switching from LinkedIn to another prospective professional social network?   You should recognize rather quickly that your small circle of close friends is not what is keeping you on LinkedIn.  Instead, you derive value from a much wider range of users on LinkedIn.  Many of those folks are not people with whom you communicate often.  They might be former classmates, co-workers, and the like.  They are people with whom you share weaker ties, but maintaining a connection to them has potential value to you professionally.  Now suppose that a few close friends defect from LinkedIn. Would you consider defecting?  Well, the problem, of course, is that all those acquaintances and former associates remain on LinkedIn.    In short, the switching costs are higher on LinkedIn.  The lock-in effect is stronger.   You would be much more reluctant to abandon that expansive "rolodex" that LinkedIn has become for you professionally.  

Why the discussion today about LinkedIn vs. Facebook?   Consider the latest news from an iStrategyLabs report.  It found that the number of Facebook users ages 13-17 have declined by 25% in the past three years.  The number of users ages 18-24 has declined by 7.5%.   These young user defections show that Facebook's lock-in effect may not be as strong as people once thought.  We've seen the story before of a social network losing traction; remember MySpace?  I'm not saying that Facebook will face that fate, but it's worth noting that network effects don't guarantee that an incumbent will remain on top forever. 

Network Effects in Reverse: Facebook's Problem with Teens

Facebook spooked investors yesterday when it announced earnings.  Actually, profit levels themselves were not the problem.  Facebook reported robust financial results.  However, management informed investors that the number of teens visiting the site declined in this recent quarter.  What's happening?   I believe that teens are migrating to other social media sites, one of which Facebook now owns (Instagram).   I hear many parents of middle school and high school students noting that their kids are not asking first to join Facebook.  Instead, they are first asking to join Instagram or some other social media sites.   

Here's the potential bigger problem coming down the road for Facebook.  When you are growing as a social media site, you have network effects in your favor.  The more users on the site, the more value per each user - it's a virtuous cycle.    However, when the number of users begins to decline, you have the potential for a downward spiral to begin.  As someone's friends leave Facebook (or don't join in the first place), he or she derives less value from the platform.  Then that person defects perhaps.  Their friends now derive even less value from the site, and they may defect.  We've seen this story before, of course... just ask the folks at MySpace.  Now, I'm not predicting such a dramatic fall for Facebook, but I am suggesting that investors have a right to be worried about the drop in teen usage.  It will be an interesting metric to watch going forward.

Are Network Effects Over-rated?

Nir Eyal and Sangeet Paul Choudary have written an absolutely terrific column for TechCrunch about network effects.  The essay is titled, "The Network Effect Isn't Good Enough."  Network effects, of course, exist when the value per user rises as the number of users of a particular good or service rises.  When network effects are strong, one firm can become the standard in a market - think eBay in auctions, for instance.  In these situations, firms tend to try to "get big fast" so as to move up the curve and achieve a high value per user.  Then they hope that switching costs will lock users into their good or service.    Eyal and Choudary point out that many startups in Silicon Valley bank heavily on the notion of network effects.  They willingly run up huge losses in the early years, hoping to cover those losses with venture money.  Then they hope to reap the rewards when they have become the dominant player in their particular market. 

The authors point out, however, that network effects may not be the elixir that many startups believe they will be.  Why?  The authors make several key points.  Here is an excerpt from their argument:

For one, in the old days, consumers paid to access the network through their upfront investment in hardware. These upfront costs locked users into the network and once they were in, they were in for good, thus erecting barriers to entry for would-be competitors. However, the cost of providing access to the network has fallen precipitously. The days of customers buying expensive hardware to use a network are gone as is the correlating lock-in effect. In addition to access costs falling to zero, another key component of what once kept users locked into a network has vanished. Once, porting contacts onto a new network, like switching instant messaging services from Yahoo! to AIM, was a non-trivial task. Today however, customers use their Facebook, Twitter or Google profiles to join a new service in seconds. A burgeoning network, take Instagram or Pinterest, can leverage the single sign-on enabled by the social graph to reach critical mass faster than ever before.

In other words, switching costs may not be nearly as high as many startups believe them to be.  Thus, the lock-in effect doesn't materialize as hoped.  Consumers flock to a platform, but then they may just as quickly move on to a different product or service.   Think MySpace relative to Facebook.   This limitation on switching costs may explain why many startups have a hard time "monetizing" their eyeballs.  If they try to extract more value from their consumer, they face the startling possibility that people will simply switch to another product or service. 

The Challenge at Etsy

Etsy is the fast-growing online marketplace where artists and craftspeople can sell their products.   Wharton Professor Barbara Khan calls it a "very well run, high-quality street fair."   Etsy has 39 million unique visitors per month.  More than $525.6 million of items were sold on the site last year.  Etsy generates revenue through a small listing fee as well as a transaction fee equal to 3.5% of an item's price once it is sold.  

As this article on Knowledge at Wharton notes, the site faces an interesting challenge as it grows.   As it becomes very popular and goes more mainstream, it risks alienating some of its hard-core original customers.  Will the artists and craftspeople see it as less quirky and unique?   Will some people accuse of it selling out when they learn that some vendors, who are larger than perhaps it may seem at first, appear on the site?   This challenge isn't new, of course.  However, it creates a dilemma for a company such as Etsy.  After all, the company benefits from network effects, i.e. the value to each buyer and seller rises as the number of buyers and sellers increases.  People want to be where many other buyers and sellers are.  That network effect drives Etsy to get bigger at first.  Yet, at some point, that urge to get big fast runs up against this potential backlash... are you becoming "too mainstream" and "losing your soul"?  

The challenge for Etsy moving forward is to maintain that quirky appeal, to stay appealing to the unique artists and craftspeople who made the site popular.   They can continue to grow, while staying true to those values and positioning. However, it will take hard work.    To achieve this objective, the firm ought to be very clear now about what they absolutely won't do moving forward.  What is absolutely out of bounds?  Moreover, they need to be very clear about their core values.   What do they stand for as an organization?  What values are non-negotiable?  By being more explicit about these issues, Etsy and other firms like it will be more likely to avoid alienating their original, hard-core customers.