Posted in: Uncategorized

Commerce Marketplaces, Network Effects, and Winner-Take-All Outcomes

In the past few months I’ve looked at quite a few commerce marketplaces and social networks. One thing I’ve been trying to understand more deeply is the set of circumstances when commerce networks tip toward winner-take-all outcomes, with one dominant network accounting for the bulk of transactional activity. This blog post is very much a work in progress, but it’s a topic I want to understand better as I’m interested in continuing to invest in marketplaces going forward.

Winner-Take-All in Social Networks

In social networks, we’ve seen strong winner takes all dynamics in some geographic markets and around certain horizontal use cases. Facebook and Tencent are global general purpose social networks that offer their respective user bases a wide variety of social sharing and communications features. But even with their horizontal dominance, there are other strong networks that have emerged to compete with them. In some cases, the vector of competition is geographic focus (Kakao Talk, LINE) and in other cases its around a laser focus on a particular feature (Instagram before Facebook bought them, WhatsApp prior to the Facebook acquisition, Pinterest around visual sharing and communication, and many other examples too numerous to list). The one other thing we’ve seen in social networks is the ability to defeat powerful incumbents by focusing attention and product development on interaction models that are orthogonal to the core value proposition of the leaders. I would put Snapchat and Twitter in this boat – in each case the interaction model (asymmetric follow for Twitter, temporal photos for Snapchat) is very different from what Facebook proposes and hence it has been hard for that company to compete with those services effectively.

In most social networks, it seems that they tip to winner-take-all in broad horizontal categories because most people want to participate, share, or interact where others are. Not surprisingly, social networks are largely governed by the power of network effects when it comes to sharing and communicating.

When Can Commerce Tip to Winner-Take-All?

The dynamics of commerce marketplaces is slightly different. Most social networks seem to grow or die based on ongoing usage and participation by one’s social network – that engagement reinforces the value of being a part of the network and encourages you to participate more fully. For commerce networks, the incentive is somewhat different. For commerce networks, buyers want to be where sellers are and sellers want to be where buyers are as that formula tends to drive maximum liquidity for all.

The more commerce marketplaces I see, the more I tend to believe the natural state of affairs for the markets in which they operate is to have one dominant marketplace provider that wins. Late last year, I started asking myself the same question every time I get excited about a particular new commerce marketplace:

Is there a really compelling reason that all buyers and sellers shouldn’t be on one single platform? If that outcome doesn’t maximize transaction liquidity, why?

Framing the question this way has helped me think through the opportunity to invest in marketplaces. In trying to get to the answer to the question above, I try to think about the 3 somewhat durable ways I’ve seen commerce marketplaces differentiate themselves:

  • Geographic segmentation – For some marketplaces, there are reasons that you cannot have a global or national marketplace in a given product or service. That could be due to the nature of the product, cultural differences, fulfillment issues, or some other reason that allows several regional players to flourish over time (as opposed to exploit a temporary advantage). Increasingly, I find geographic segmentation to be more of a temporal, as opposed to durable, advantage for marketplace companies.
  • Supplier or Buyer Self-Selection – One other reason that marketplaces can avoid tipping to winner-take-all is that buyers or sellers self-select into or out of a given marketplace. One way in which this happens is by the chosen business model for the marketplace. For example, if the marketplace sets prices, wages, or rates, service providers who want to charge more might opt out of the marketplace. By the same token, if the quality of service providers is concentrated in either high-skill / high cost or low-skill / low cost, there can be buyer populations whose needs are not served in that marketplace. This can create multiple marketplaces that self-sort based on the needs of buyers and suppliers.
  • Commodity Products vs Experience Goods – Last, I think the nature of the product or service being sold really does matter. As I’ve written about in a previous post, the nature of the product or service being sold matters. Is the product or service a standardized service or product that is easy to evaluate prior to usage or is it something where consumer preferences and tastes matter and the only way to evaluate the service is by becoming a consumer? For commodity or standardized products, I think the power of concentration on one platform is high – if I want to buy an unboxed iPhone, I’ll probably want to buy it at the place that has the most unboxed iPhones for sale as they’re all the same. Service marketplaces tend to be different – in many cases, the only way to figure out if the marketplace meets your needs is to test the service providers offered.

Of the 3 bullet points above, I spend the most time thinking about the implication of the last bullet points. Understanding the nature of the product and service being sold really drives a lot of my thinking. Take ridesharing for example. If you think that a ride from point A to point B is a commodity good where the primary driver in consumer decision-making is price, then you’d suspect that the one provider who can consistently provide the lowest cost and highest liquidity will win the entire market. However, if you think that ridesharing is an experience good where consumers do not think of a ride from point A to point B as a commodity but as an experience where they will trade price for convenience, style, or some other feature, there is a strong argument for why multiple services can thrive.

In addition to ridesharing, the other areas where I think this could play out in an interesting way are the home and apartment rental sharing space (airbnb and onefinestay), peer-to-peer second-hand commerce, and those who are trying to chip away at Craigslist on specific vertical use cases.

As always, comments are open below. You can also share your thoughts with me on Twitter @chudson.

Comments (3) on "Commerce Marketplaces, Network Effects, and Winner-Take-All Outcomes"

  1. Great post Charles!

    What do you think about services like Amazon Prime (or if uber was giving discounts based on the number of rides), that could increase engagement and usage from users of the marketplace?

    I think this is like your example of Snapchat and Twitter, in the sense that the interaction model changes versus Facebook’s. But instead of a difference in feature, it would be a difference on business model / pricing.


  2. I do think that Amazon Prime is an example of Amazon trying to make buying commodity products on their platform a more appealing option. I don’t think Uber really needs a frequent rider program – they are winning already and have lots of engagement. Maybe some of those who are not in a leadership position in the market should consider it, though.

  3. This is a really insightful breakdown that gave me a few things to ponder. I’m working on a marketplace myself, and you raised a question in my mind: given the components of what a good marketplace does and the risks involved with experience & commodity marketplaces, what makes a minimum viable product for a marketplace? And are all components required for initial traction and growth?

Leave a Reply

Your email address will not be published. Required fields are marked *

Back to Top