If you aren't incurring losses, it means you are throwing away potential income by growing too slowly, and inviting competitors to take up the slack.

In case of a market or tech hiccup, you might find yourself without "sufficient cash flow", and collapse, but that would be the investors' problem. The investors are assumed to be able to absorb the loss and disappointment without undue discomfort, and so they are. Risk to your dreams takes last place.

I understand a lot of silicon valley is funded in this manner right now but please re-read this command and realize how silly it is. It's proper to say that taking a loss for long term gain is sometimes the right measured decision but saying that failing to lose money constantly is a failure is just ridiculous.

Well in their case if they had to they could axe most of their sales and marketing spending, and become cash flow neutral and survive a hiccup.

No, this is a healthy high growth SaaS company.

Elastic is not predominantly a SaaS company. It's part of the play, but not the main source of their revenue.

I think they are transitioning from license/support to SaaS. Also looking at their statement SaaS is taking over in revenue. License 25 Million, subscriptions 123 Million in FY 18.

Subscriptions do not imply SaaS, though Wall Street often equates the two. It’s just a different licensing model that’s replacig the up-front perpetual license + 25% annual maintenance that is the old school enterprise software model.

The majority of that subscription revenue (they make this clear in other statements in the S-1) is enterprise deals running Elasticsearch & Friends on enterprise datacentres or clouds where Elastic isn’t providing a managed service - they’re providing traditional enterprise software support.

This isn’t unique to Elastic. Splunk, Pivotal, Mongo, MuleSoft, all are similar. Wall Street mostly cares about the revenue model, but there’s a nuance to it.

Subscription was the bulk of their revenue, over licensing and services...

Subscription includes on-premise installations; it’s just another licensing model. It is common to move away from perpetual license up-front followed by maintenance models to a more spread-out revenue model and better predictability for Wall Street (plus potentially better quarterly deal volitility insulation because of more deferred revenue, especially if they’re doing up-front multi-year subscriptions).

They also say that the majority of their revenue is from enterprises running their own instances.

Anyway, the point is mostly moot for an investor, but it is interesting as a technical person to observe that many SaaS plays aren’t entirely or even mostly *aaS, they’re open source software sold as enteprise software subscriptions (priced at enterprise or SaaS levels too, not a bargain basement OSS support contract)