Network effects from existing and new incumbent users

Draft essay on how different roles within a community may impact how the network evolves.

This essay looks at the indirect relationships between segments of users — and the positive and negative effects as a platform evolves. The first case, from LinkedIn, shows a direct impact between different generations of users. The second case refers to Mariam’s company,; which shows positive and negative outcomes measured as a new group of users were introduced to their platform.

As indicated by David Evans and Richard Schmalensee, the concept of Network Effects were initially influenced by the assumption that users in a network were part of the same group. Such original notion end up being shaped as the Direct Network Effect:

Direct network effect: The impact of the addition of another participant to a network on the other participants in the same group. A positive network effect occurs when an additional participant makes other participants of the same sort better off because they can reach and interact with more participants. (Matchmakers, 2016)

During the .com boom days, this (direct) network effect idea strongly influenced entrepreneurs, investors, and many companies in the belief that the first mover would be able to acquire the whole network and unleash its full potential. As also pointed by David and Richard, such view became obsolete with the increasingly recognition of different roles of users and how they collaborate within a network.

As such users have different needs, and in many cases showing interconnected dependencies, a platform may well consider going beyond being the first-mover for one segment of them. Instead, to be positioned as an ideal mover for the multiple kinds of users depending on each other — the recognition of indirect network effects.

While direct and indirect network effects are indeed key considerations for a network to sustain, the following examples focus in the network effects among existing and new incumbents users in a growing network.

In the case of LinkedIn, Allen Blue reveals how a segment of users impacted further generations of users:

We basically discovered that there was a set of people that really, really, really, really, really cared about networking. They loved to meet new people, they loved to add connections and so forth, and they were the primary drivers of our early growth. They were a place where we had product market fit literally from day one. So, sometimes, you have to figure out a way to leverage that particular group of audiences. (Blitscaling 05, Oct 13, 2015, 1h6m34s)

LinkedIn cofounders also learned much more amidst their growth. “[T]he people that were most fanatical about our product wasn’t necessarily the people we knew” says LinkedIn cofounder Reid Hoffman. Reid adds that when they did analysis, one year later, it turns out that 75% of the network was sequenced after the generation 1000.

Another example is the case of Minted(.com). This case shows a different kind of relationship between segments, one with positive and negative effects generated through a quite indirect interaction. Minted started to offer products such as wedding invitations, featuring designs by top brands:

“We’ll be launching soon to bring you wedding invitations and custom stationery from the best independent brands as well as selections from up-and-coming members of the design community.” (, 2008, from Web Archive)

The above message, from 2008, introduced their customers to two potential sources of users (suppliers) — the top brand designers and the design community. As Minted continued to learn about their network, they recognized, first, that customers enjoyed new designs from the community instead from the top brands.

Such learning led them to drop the relationship with their “best independent brands”. But another finding, that came as a surprise for them, was that their new community designers were entering the platform exactly because the top brands existed in first place:

So we had signed up all these brands for exclusive distribution, and then had to have the courage to go face all of them and say “oops, sorry, we made a mistake. We can’t sell any of your products, it turns out, at all, and instead we’re going to go completely compete with you and source all these crowdsourced designs. And not only that, but the reason why all these people are entering in the competitions to begin with is because you are here with us right now. “ (Mariam Naficy, 2015, 27:56)

Mariam classified the positive effects as aspirational. At the same time, she recognized the negative effects that they learned along the way, which led to the termination of the relationship with top brands. Perhaps we could look at such case as a creative destruction phenomena arising within the network:

  • Her investors were more inclined, initially, to support the relationship with the said best brands;
  • They recognized that the two communities of suppliers were different; later that the aspirational component were a growing force for the new incumbents which wanted to compete with top brands;
  • As they let the crowdsourced community to raise, such incumbents designers took over the supplier side; which led into a sustainable business model that saved the company.

With the above examples, my hope was to focus in the network growth potential. Beyond the consideration of the direct network effects, I wanted to highlight examples of relationships that can influence new incumbent users.

Matchmakers. David S. Evans (Autor), Richard Schmalensee. Matchmakers: The New Economics of Multisided Platforms. (23 May 2016).

Blitzscaling 05, Greylock Partners, Published on Oct 13, 2015; John Lilly on Leveraging Community to Scale Mozilla,

Mariam Naficy. Greylock partners. (Published Oct 26, 2015). Blitzscaling 07: Lessons From The Dot Com Days and Knowing When To Blitzscale [video]. Retrieved from Web (Captured in 2008). Retrieved from