Strategies For Building Network Effects

The arrival of web 2.0 has revived the network effects business model. Every where I look, web 2.0 companies are extending the existing model of 2 sided networks to re-integrating multiple part of the value chain. Take Zazzle for example, it’s a marketplace that extends the existing model of buyer-sellers by chunking out a previously vertically integrated seller responsibility of manufacturing into a separate role. I would not be surprised that in the future, once Zazzle is able to codify its manufacturing processes, that it enables third party or franchised “sub contractors” to be the manufacturing & logistic provider for its sellers/artists. (thus increasing its ROE and margins)

One of the major problems many “Network Effects” driven businesses face was the “empty chat room” issue. Like a chemical reaction, a certain amount of activation energy needs to exist in the value network in order for the virtuous adoption cycle to take place. How do you achieve that critical mass of users needed in order to attract more users and eventually create barriers to entry to dominate your market? There are several strategies employed by companies.

1 First to market by atleast 2 years. The classic example is of course eBay. What many people forget was that eBay was formed in 1995, at the time, the mere presence of a platform or venue was enough of a draw to end users given that there weren’t any alternatives. Word of mouth spread fast enough and conversion rate from visitor to user(buyer or seller) was high enough that the “empty chat room” problem occurred only for a small set of very early adopter. Driven by “newness & coolness” factors (it was actually one of the first commercial websites along with Yahoo!), traffic grew exponentially and trial sellers enabled Pierre to build an early lead that it never relinquished.

2. Luck. Everyone understand the concept of random walk and noise. So think of it this way. On the X-axis, time. On the Y-axis, # of SIMULTANEOUS users. With Z# of simultaneous users equaling to the critical mass required to build network effects. The company/site has some constant AVERAGE X# of simultaneous users, however, given noise and random chance, the actual # of simultaneous users fluctuates based on a normal distribution. Most of the time, the company is barely hanging on with that weekly average holding steady. BUT given time, the normal distribution WILL prove itself out (if the standard deviation is high enough to reach Z) by having X>Z simply by “luck.” At which point the business takes off and VC’s and founders become billionaires. Highly risky strategy and dependent on having a wide normal distribution of simultaneous user which is rarely the case anymore for todays’ websites. I sometimes call this strategy “build it and they will come.”

3. Blow it out. Raise a ton of money and spend it on acquiring a huge # of customers in a short amount of time – hoping to juice the marketplace through marketing and getting to Z# of users immediately. This was the strategies of many failed dot-coms and b2b marketplaces in the early 2000’s. Unfortunately when there is so much noise in the advertising channels it was really hard to make the message stick and for the ROI to justify both the cost of the campaign and the users acquired. Some worked (think Paypal’s pyramid scheme in its early days), most didn’t. VC’s don’t invest in startup that do this anymore, in most cases you better be at critical mass before raising series A.

4.Build “localized” network value. A variation of strategy 3, but focused on creating critical mass in one segment of the customer base via marketing and sales. Some examples include social networks, local trading/classifies, and superniche marketplaces. The whole point is that users place different value on having different type of users in the same marketplace based on their own preferences. Find a segment of the target market that values each other disproportionately higher than any other segments (you value your friends much more than random people) and concentrate on building critical mass in that segment before moving on to the next. Really an extension of Crossing the Chasm for network effects businesses.

5. Seed the marketplace. Find all your friends, recruit “mavens” and “influentials” (read the Tipping Point) and just beg them to sign up and use the product. Sometime companies will hire part-timers to act as initial users so that potential customer will perceive value when they first visit and convert to regular users. Its the equivalent of the age old strategy of opening a restaurant and hiring people to hang out/eat at there to give an impression of popularity.

6. Create “Point” product value. This last strategy is probably the least understood and could potentially be the most powerful and cheapest. Christina Jones of (a blast from the past!) explained it to me once when I was doing technically called consulting (but more like minimum wage internship) work for the company on their rebate strategy.

In essence the strategy is to design products that has 2 components to its value proposition – point and network value. The point value is value proposition for the product independent of the # of people in the network. As a result, users are compelled to adopt the solution regardless of general adoption. Personal productivity software such as word, excel etc derive most of its value through its ability to allow its users to be more efficient, more accurate, and more productive. The hard part is to create coherent synergies between the two components that encourage the usage of the network component once the user begin using the point product. Intuit is doing this via Quicken (point value) + MyQuicken (network value) which includes Investing Center, Banking Center, Loans, Insurance etc. As another example, Trumba’s calendaring service could create more point value for its software if it is able to synchronize the various personal work/home/online calendars of a single person before relying on the ability of the service to publish to multiple users.

Of course, none of the strategies presented here are mutually exclusive, try it all if you have the time & money. But given limited resources some evaluation will need to be made on which combination will be the most effective. Also, I’m sure I’m missing some as well as this is based on my personal experiences and observations, so feel free to add more.