Snowflake's Competitive Position and Pricing Model

Thanks to @LisaOnCloud9 and others for their valuable analysis of Snowflake. Having run a business for 20 years that enabled customers to pull back on spend or increase it as needed, I have to say that I see Snowflake’s pricing model as a significant competitive advantage over alternative pricing models. It is likely to help them attract and retain customers.

Yes, it makes valuing the business harder: latent future revenue, embodied in those customer relationships where spending has decreased, doesn’t show up under even the most aggressively “non-GAAP” analysis a company might put forward. Nonetheless it is a significant business advantage over the ever-present competitor known as DIY.

I know others have made similar points on this board in the past, but with all the discussion of how overpriced SNOW is rn, largely due to the falloff in revenue, I thought it was worth revisiting.

Looking more broadly at competition, AWS and the other hyperscalers charge according to a similar usage-based model, so no advantage for SNOW there. But we’re talking about a new, fast-evolving market where the game is won or lost on the basis of features and support. I prefer to bet on the focused specialized player against the “everything under the sun” player in that kind of market. Point solutions tend to win until features and support become commoditized; that’s when the “suite solution” wins. We’re not there yet in the data lake market. Meanwhile, the pan-hyperscaler nature of a Snowflake solution trumps what any individual hyperscaler can offer – and the large flow of cash from Snowflake to the hyperscalers should continue to make them less than eager to try to out-compete Snowflake.

What about smaller competitors? I don’t worry about them much either, at least until after they IPO. Until then, any F2000 CFO will choose the financial stability of a SNOW over a competitor who is merely “just as good”.

Consequently I see Snowflake as being in the winning position in an emerging, very large market where today’s revenues will be dwarfed by revenues down the road. That doesn’t make me price-insensitive, just a whole lot less sensitive than many posters on this board. I’ve bought 54% of my position so far this year, and at prices fairly close to the current level. My average purchase price for the whole position is $176 (vs. 170 or so today). I may buy more in the future.

I fully expect to hold for at least 5-10 years, because it will take that long for the “gaping open mouth” of a new market that I see to become mostly realized. In that timeframe, I expect the current price volatility will become insignificant.

Short-term, I want to continue seeing

  • the use of “data edges” climb (now at 25% of customers);
  • the number of customers spending $1m+ continue to grow strongly (up 80% y/y); and
  • the NPS stay high (last reported at 72).

Long-term, I want to have the courage to stay fully invested as the absolute number of dollars involved (hopefully) goes up. Intellectually knowing it’s “the right investment decision” only goes so far in confronting fear. That’s not a SNOW-specific comment; that’s a general comment on my personal challenges as a longterm investor.

For me, this is a high conviction position. It’s the largest position in my “high risk” portfolio, at 11.2% (I am not a follower of Saul’s concentrated portfolio approach, but I appreciate the focus and experience of the posters on this board).

To summarize, for me, SNOW’s usage-based pricing is a “feature”, where for many of you, it’s a “bug”.