The Snowflake IPO

Snowflake has launched its initial public offering (IPO) for a mind-boggling $33 billion, despite being currently loss making. This puts Snowflake into Tesla country as, in my opinion, being way overvalued. But, hey, what do I know about the craziness of markets? From a technology perspective, however, I can comment.

There is no question that when Snowflake first came to market it was disruptive, and it has been riding that wave of disruption ever since. So, what did it do? Firstly, it offered data warehousing as a managed service. Secondly, it is cloud-native, based on the use of microservices (containers). And, thirdly, it offered separation of storage from compute, which allows you to scale these independently.

Now consider the market today. Many of Snowflake’s traditional competitors, as well as those that are cloud-based, now offer managed services, so that is no longer a differentiator, while leverage of microservices is becoming increasingly common. And finally, the separation of storage from compute, is now commonplace and, in any case, is not as significant as it sometimes claimed to be.  This is because it is only useful when the two scale at different rates: if they are more or less in sync, or constant, then you don’t get any benefits. And while you can scale down compute power it is non-trivial to scale down storage. This is why some competitors to Snowflake are offering a choice of separation or not. Microfocus’ Vertica data warehouse even offers on-premises separation if you want that.

The biggest thing that Snowflake has got going for it is momentum. It is has seen very rapid growth over the last couple of years, even bearing in mind that it started from a low base. And there has been a lot of buzz about it. But is it worth $33 billion? I don’t think so.