What do you like best about CloudZero?
* Setup and connection to your AWS accounts is really smooth and easy and utilizes least-privilege IAM permissions. The team has done a masterful job on this.
* The ability to slice and dice your cost data by so many different dimensions is a killer feature, especially the tag support. It's pretty easy to start to see immediate insights even without a reasonable tagging strategy or naming conventions. With a solid organizational strategy is where you find the biggest benefits - rolling up costs for teams/services/features or attributing cost per customer. CloudZero gives you insight into your cloud spend, but it also indirectly guides you towards building better products where such analysis is possible.
* The ability to see untagged resources in context via extrapolated metadata and related resources is a big help in cleaning up messes like old unmanaged AWS accounts and abandoned experiments.
* The performance and UX of the platform is quite good given the volume of billing data, screens are uncluttered and the information is easy to consume.
* The "resource diff" feature, while rudimentary at present, is a useful data point that can help you figure out why a cost anomaly might have occurred.
* The ability to change "cost views", e.g. real cost, billed cost, is supremely useful for communicating with different audiences. The unmodified AWS bill is terribly difficult to interpret if you're a developer trying to figure out why your service is suddenly costing more.
* Cost anomaly detection has saved my bacon on a number of occasions, especially when doing serverless work. Even with EC2-based systems it's useful - helping me find suddenly-overprovisioned clusters or instances left running.
* Seeing relationships for a resource you're analyzing is very useful when trying to solve a billing mystery. Often, a resource might have an incomprehensible name, but when seen in context with other related resources it becomes far more obvious what it is.
* The monthly trends, slack alerts, and document downloads are all great features. Each one helps you see your spend from a different angle.
* I like very much that CloudZero is branching out from EC2 and serverless into other areas like kubernetes and snowflake. Cost analysis is useful everywhere.
* The focus on unit economics applied to your cloud spend is highly welcome and a trend I'd like to see continue. It's a mark of organizational maturity to perform this analysis; I'd like it to be on the minds of every executive I work with. The fact that CloudZero says "yes, we value this philosophy and will help you with answering these questions" is a big reason why I chose CloudZero, and why I will continue to do so. Review collected by and hosted on G2.com.
What do you dislike about CloudZero?
* The ML features around untagged/unknown resources, while extremely useful, do not seem intuitive. I feel like there needs to be some sort of guide or wizard that can help you go from "I just connected these accounts" to "hey this thing is finding structure I didn't know about" to "hey I've mastered my tagging strategy and know exactly what's going on". Definitions and examples on the 3 divisions (cost group, feature, product) would also be helpful.
* There are no in-app definitions and examples for the different cost options (real cost, billed cost). I know them from talking with their team a lot and the views work fine, but this was not immediately clear. How credits, discounts, and support factors in is also unclear (though I understand this is being addressed)
* I use KMS keys a lot, but given their cryptic names I feel I need more context. Unlike many other resources, keys do not seem to have the context like relationships that I would need to identify what they are supporting. There are occasionally resources like this that are hard to track down.
* AWS is notorious for doing weird things with resource names, like putting instance name in a tag and key aliases in a separate resource. CloudZero could probably benefit from navigating some of these rules and showing the friendliest name possible for a resource. This would be especially helpful for things like large fleets of instances. Right now, the cost explorer contains a mix of easy-to-identify resources with incomprehensible ones and that makes discovery somewhat difficult. However it's still a lot better than using other tools!
* Relationship links are not bidirectional when you are drilled-down into a resource. This leads to potential blind spots where you can only see what's connected to what if you pick the right resource to look at.
* There's only a limited ability to drill into the cost explorer and uncover what's behind a "long tail". In a lot of cases I will see the top 5 taking up 20% of the total cost, with 80% being "other". Really unpacking what "other" contains is a bit more work than it needs to be IMO. There are no "negation" filters, where I could say "show me everything except these things I already know about"
* CloudZero has limited options for integration with your auth provider. Would be nice if it supported GSuite. There are also occasional glitches with signin - nothing a page refresh can't fix, but somewhat annoying. Review collected by and hosted on G2.com.
