In the space Atlassian is in, choosing to be unprofitable while quickly growing could make sense.
It is not necessarily true for other sectors (e-scooters, different delivery services, maybe even Uber), as there is no stickiness in those areas. Whenever a cheaper Uber would come to my city, I'd drop them within a second. They are usually cheap while they can burn VC money.
If Atlassian can keep their system for a couple of years in a big, slow legacy company, the company will end up with tens of thousands of tickets and pages.
Then, no matter how everyone thinks that all the products of Atlassian is terrible (there is a thread about it every second month here), the company will never leave Atlassian because nobody wants to spend the time on migrating all that stuff and everybody is afraid to say that "we will probably not going to need poorly written user stories from three years ago", and they don't want to risk that the new system doesn't cover everything that e.g Jira does.
Once a company is locked in, it will keep paying because paying any amount is easier for them then migrating, training the employees to use the new system, broken links, missing features, etc.
You have it nailed: they offer low-churn products.
Atlassian is well-managed. Their revenue is fine. Their cash flow management is good. Their compensation plan is effective at retaining employees--the stock employees receive is likely to be more valuable at the end dof the vesting cycle than it is today.
Atlassian has got a collection of offerings that are industry standard. When used at large companies, they become quickly entrenched, with very low churn.
It is also a collection of offerings that is chosen by startups. Last quarter through a year from now, I expect that stream to be dry. But they are on 3-year contracts at an awful lot of got-my-Series A startups that will have to tighten their belts in other ways before trying to rotate off of Atlassian.
If you are actually producing e-scooters it could make sense as well when you have plans for reducing your costs long term through economy of scale so you reach the minimum efficient scale.
It is not necessarily true for other sectors (e-scooters, different delivery services, maybe even Uber), as there is no stickiness in those areas. Whenever a cheaper Uber would come to my city, I'd drop them within a second. They are usually cheap while they can burn VC money.
If Atlassian can keep their system for a couple of years in a big, slow legacy company, the company will end up with tens of thousands of tickets and pages.
Then, no matter how everyone thinks that all the products of Atlassian is terrible (there is a thread about it every second month here), the company will never leave Atlassian because nobody wants to spend the time on migrating all that stuff and everybody is afraid to say that "we will probably not going to need poorly written user stories from three years ago", and they don't want to risk that the new system doesn't cover everything that e.g Jira does.
Once a company is locked in, it will keep paying because paying any amount is easier for them then migrating, training the employees to use the new system, broken links, missing features, etc.