Revision of the Marketplace 100 index by a16z

A recent analysis by VC a16z, based on an extremely interesting data set, is somewhat uninsightful.

The article starts by saying: “Today we introduce the Marketplace 100, a ranking of the largest and fastest-growing consumer-facing marketplace startups and private companies.”

Largest and fastest are two very different dimensions. When doing this it’s very important to clearly define them and put a base point for one of the two dimensions. For example: the fastest growing marketplaces with revenue above X, or the largest firms with growth over Y.

The article goes on drawing a conclusion that seems misleading: 

As the subset takes two data points (size and growth), mixing the two can lead to non-insights.

The chart above shows how a few marketplaces account for the majority of GMV. This data point would be interesting if marketplaces were ranked in term of absolute size. In that case companies like Amazon, eBay, Booking.com, Expedia, Thumbstack, Uber, Netflix and many others would figure in this list.

As companies don’t really compete in terms of growth rate (e.g. Uber and AirBnB were founded at the same time and grew quickly but they were never in competition), grouping them like this doesn’t tell us much.

What is probably the more interesting insight is that a few large marketplaces like the ones above are still growing remarkably fast. Understanding exactly why that is would be a golden nugget of information.

The fact that there is no base also slightly weakens the whole data set: growth is only insightful when measured in relation to size. For example if a marketplace grew from 10 orders to 100, that would be a 10X growth, but in reality this data point would not be significant. This however is secondary.

The chart below shows another decontextualised analysis: 

This visualisation is quite misleading as it shows AirBnB as extremely dominant in the travel market sector, but the sample is in no way representative of the whole sector as companies like Booking.com and Expedia are not being considered. 

The chart below is more interesting.

This chart has potential for a more interesting analysis but it's another unfortunate visualisation, though for a different reason. What the chart shows us is that there are four outliers – but including those four in the analysis makes the rest of the visualisation quite obsolete due to the scale: all other marketplaces are so small in comparison that they appear as a mass of dots along the Y axis. This makes it hard to extract information from the visual.

One could also argue that Faire, the fastest grower, being in the wholesale category, is not a B2C marketplace and therefore shouldn’t be included, but that is also secondary.

I think the dataset is extremely interesting and it can lead to more interesting questions like:

-      What are the fastest growing marketplaces in large markets?

-      In which markets do we see a winner take all scenario and why?

-      Why and how are some outliers growing so quickly?

Ultimately I believe that the cut of data chosen has led to a slightly flawed analysis – this goes to show that we must be very careful about the data we gather and the visualisation we chose when looking for insight. 

I did not feel compelled to criticise this piece of content for the sake of it, I have done it because I felt it was genuinely confusing / misleading.

P.S. I would like to add that the content published by a16z on this topic is usually of the highest quality and I thank them for that.