4 ms·

Ultimate includes a bunch of bells and whistles - full k8s integration, static/dynamic security testing, deep monitoring integration, etc. I have considered it

Ultimate includes a bunch of bells and whistles - full k8s integration, static/dynamic security testing, deep monitoring integration, etc. I have considered it but ultimately at any given time, there are one or two features in the top tier that sound good, but usually you can get 80% of the way there with a FOSS alternative. For example wiring up ArgoCD to run your k8s deploys, or writing your SAST rules yourself.

acdha 6y ago
Also it's all-or-nothing: the security tools are a good example where it could easily be worth paying a reasonable amount for either an optional feature or upgrades for specific users but when that applies a big up-front multiplier to your entire organization it becomes a much harder sale.
- theptip 6y ago
  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.
  - acdha 6y ago
    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.
harrisonjackson 6y ago
They push features down to the lower tiers pretty frequently. Easier to launch them on ultimate and then move them down than the other way. Silver got a lot of great project management features around roadmapping and epics last year that allowed our team to cancel jira sub, so that was sweet. edit: wouldn't count on the fancier devops tools getting moved through :/
- addicted 6y ago
  That's not reassuring in a comment on an update where they are completely eliminating a tier.
  - harrisonjackson 6y ago
    My response was to someone considering the ultimate tier. I'd recommend they keep an eye on the premium (silver) tier for the functionality they want which may at some point become available. Premium was an easy sell for us as we were able to eliminate other subscriptions. Ultimate does not have features we'd be willing to pay an extra $80 per user per month for.
dogma1138 6y ago
If you want your SAST tool to be worth a damn you need to write your own rules, at a minimum configure sources and sinks manually especially in an app that has a lot of dependencies. Last time I checked the security scanning tools that GitLab provides don’t really allow you to do that, SAST is expensive and also require manpower to manage well I guess if you only want to tick a box it might be worth it but I wouldn’t call the security features of GitLab worth the money you’re paying for especially with a large number of users.

acdha 6y ago

Also it's all-or-nothing: the security tools are a good example where it could easily be worth paying a reasonable amount for either an optional feature or upgrades for specific users but when that applies a big up-front multiplier to your entire organization it becomes a much harder sale.

theptip 6y ago
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.
- acdha 6y ago
  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.

harrisonjackson 6y ago

They push features down to the lower tiers pretty frequently. Easier to launch them on ultimate and then move them down than the other way. Silver got a lot of great project management features around roadmapping and epics last year that allowed our team to cancel jira sub, so that was sweet. edit: wouldn't count on the fancier devops tools getting moved through :/

addicted 6y ago
That's not reassuring in a comment on an update where they are completely eliminating a tier.
- harrisonjackson 6y ago
  My response was to someone considering the ultimate tier. I'd recommend they keep an eye on the premium (silver) tier for the functionality they want which may at some point become available. Premium was an easy sell for us as we were able to eliminate other subscriptions. Ultimate does not have features we'd be willing to pay an extra $80 per user per month for.

dogma1138 6y ago

If you want your SAST tool to be worth a damn you need to write your own rules, at a minimum configure sources and sinks manually especially in an app that has a lot of dependencies. Last time I checked the security scanning tools that GitLab provides don’t really allow you to do that, SAST is expensive and also require manpower to manage well I guess if you only want to tick a box it might be worth it but I wouldn’t call the security features of GitLab worth the money you’re paying for especially with a large number of users.