Agreed, this is the problem with Gitlab's bloat; the more features they add, the harder it is to justify a cheap price-point like the old $5/month -- and everybody's idea of what should go into the $20/month tier is going to be different. You can't disaggregate the offerings to pay for just the pieces you are interested in.
If they are focusing more on the Enterprise side of things, I could see it being a good idea for them to split out the licensing and billing for different offerings; see Datadog for an example of the opposite extreme, where every service is optional, and you pay (handsomely) for whatever you use.
If Gitlab split their offerings into a few packages, say $10/mo for k8s integration, $10/mo for security, $10/mo for the Jira replacement that I've never touched, and so on, they could potentially get more stickiness - users would be using the top-tier functionality for the modules that they care about, instead of missing a few killer features that are stashed away in the gold tier. I'm not paying $100/mo for a Gitlab seat, but I would have considered $20-30/mo extra for the k8s/prometheus CD integration without having to pay for all the rest of the top-tier features I'm not interested in.
This might help Gitlab to focus a bit more, too, as they would be able to see exactly what product lines are bringing in the most revenue, and focus on making those better.
It's especially tricky with some competitors' choices: for example, Github's free dependency update service is a HUGE chunk of the day-to-day security value for most projects and there's not much room to say you can justify a double-digit dollars-per-user-per-month to get the other tools.
If they are focusing more on the Enterprise side of things, I could see it being a good idea for them to split out the licensing and billing for different offerings; see Datadog for an example of the opposite extreme, where every service is optional, and you pay (handsomely) for whatever you use.
If Gitlab split their offerings into a few packages, say $10/mo for k8s integration, $10/mo for security, $10/mo for the Jira replacement that I've never touched, and so on, they could potentially get more stickiness - users would be using the top-tier functionality for the modules that they care about, instead of missing a few killer features that are stashed away in the gold tier. I'm not paying $100/mo for a Gitlab seat, but I would have considered $20-30/mo extra for the k8s/prometheus CD integration without having to pay for all the rest of the top-tier features I'm not interested in.
This might help Gitlab to focus a bit more, too, as they would be able to see exactly what product lines are bringing in the most revenue, and focus on making those better.