it seems like companies like DDOG’s product sells itself lot easier unlike ESTC where its capabilities need to be explained which perhaps contributes to the lower margins.
I think this is a classic example of the fundamental difference between a product play vs. an infrastructure play. And I’m now kicking myself for not recognizing it sooner!
People on this board have endlessly compared the numbers between ESTC and MDB. And no one has really pinpointed why there is such a discrepancy between the two. Your comparison of DDOG however, hit me like a brick.
Infrastructure plays always sound like fantastic ideas, but seem to never really take off as one expects. This is a lesson I thought I learned a long time ago, and yet, got sucked into ANET which was the same thing. And now, once again, I find myself holding ESTC as well.
So, what is an “infrastructure” play, vs. a “product” play? Simply put, the infrastructure is the thing required to build a product. Thinking back to the “Internet Boom” of the late '90s/early 2000s, there were all sorts of companies promising to light this world up by laying fiber optics. They all imploded. They laid so much capacity that we still, 20 years later* haven’t lit up all the dark fiber they laid down! They were “infrastructure”.
ANET - infrastructure. Just as the fiber was required to build high-speed internet capabilities, ANET’s infrastructure is required to build cloud capabilities.
ESTC - infrastructure. They’re picks and shovels are required to build monitoring and logging applications. They are a means to many different ends. Just like fiber, just like ANET’s switches and routers.
Compare those things to a “product” play:
NFLX - Pure product with every human being as a potential customer. They rely very heavily on that laid fiber. And they rely on AWS (which uses ANET’s switches/routers). And they rely on ESTC for a lot of their logging/monitoring, etc.
Amazon - (somewhat of a hybrid infrastructure/product) - Amazon.com relies on AWS to provide it’s online shopping services just as it relies on UPS for delivery. (UPS is an infrastructure play).
DDOG - Product - Relies on high-speed internet provided by fiber, the cloud, and things like ESTC among others.
TTD - Product - Relies on all the same infrastructure as everyone else to deliver their services.
MDB - hybrid, but mostly product - Provides infrastructure to some extent, but really they are providing the service of that infrastructure in the form of things like Atlas. But even with out Atlas, they rely on a tremendous amount of infrastructure to pre-exist before their database can be useful.
OKTA - product - relies entirely on infrastructure to exist and sell it’s services.
ZS - I’m thinking infrastructure (which means I might need to get out now!)
TWLO - I’m beginning to think infrastructure as well…
The last two are interesting. If ZS weren’t cloud-based, and sold an actual widget, it would be unquestionably infrastructure. But they’re leveraging the cloud, so maybe it’s a hybrid…
TWLO too, they provide the things that others use to create great products and services.
If it’s not clear, my point is, infrastructure never wins big, but is required for those products and services which everyone benefits from. I don’t benefit from lots of fiber any more than I did from a PC in the pre-internet generation. What I benefited from was the software (Microsoft, Adobe, etc.) that ran on the infrastructure created by the PC manufacturers. I don’t benefit from fiber optics. I benefit from Netflix for streaming and shopping at Amazon, which are only possible because of the infrastructure provided by the fiber and the ISPs who provide access to it.
I don’t benefit from ESTC’s ELK stack. In fact, I suffer because I need it. I need to understand it all, build it up, configure it, manage it, etc.
But I DO benefit from DDOG’s service, which is built in part on ELK. THEY manage it, configure it, deploy it, worry about it, etc. I merely pay them for an amazingly easy to use product that does what I need!
I think TWLO is a lot more like ESTC now that I think about it. I’ll have to do more research on ZS, but my gut is telling me that they’re more like ANET and therefore infrastructure, than they are like NFLX or DDOG…
MDB, like I said, kind of a hybrid, but ATLAS moves them more toward product/service…
Anyway, my $.02. I think I need to get out of ESTC, TWLO, and ZS. I don’t like infrastructure as an investment, they tend to be races to the bottom (e.g. every fiber optic company, every PC manufacture, etc.).