Current/Upcoming Integrations/Features
Which data source integrations are supported today?
AppD, Datadog, Dynatrace, Google BigQuery, Graphit, Lightstep, New Relic, Prometheus, Splunk, Splunk Observability, Thousand Eyes
Why do we need a solution like Nobl9 when I can build internally?
Fair question, teams we are talking to felt building internally was the way to go, but quickly realize they need to allocate/hire proper resources to build the solution, it is time consuming, there is a substantial financial investment that needs to be made, you need to maintain the solution, and you will need to figure out how to horizontally scale the solution as customer expectations change and business goals are realigned. Nobl9 has done all of the work to build a best in class solution that will meet and exceed what our customers are trying to accomplish with SLOs.
How does Nobl9 compare with Honeycomb?
Honeycomb has a good SLO product, and we have many mutual connections / friends between our two companies. The biggest difference between Nobl9 and Honeycomb is that Honeycomb requires specific instrumentation to be installed across everything you want to monitor, whereas Nobl9 can pull in existing data from your already existing monitoring in place.
We know we want to do SLOs, but we’re too busy and can’t allocate cycles for an SLO migration.
I totally get it, I feel like you may be overwhelmed with the whole idea of SLOs. Many of the teams we’ve worked with felt the same way at the beginning - you are not alone! After working with our team what, what they found was they already had all the right data/metrics needed for SLOs today. Let’s take a step back for a second.. Do you have SLAs? What are those based on? How are you measuring whether you’re hitting your SLA or not?
How much does this cost? I would like to see pricing before moving forward.
Pricing is tied to the # of data sources you are connecting to as well as the # of objectives you are measuring.
As much as I would like to put out a number for you, it’s not that easy: it's specific to your environment and goals.
How many SLOs are you managing today? Although the # of SLOs may be small now, most companies scale quickly once they get the hang of Nobl9. As we continue our discussion, my team can help you define your current state as well as desired future state. After we uncover this my team can create a package that makes sense for you.
Does that make sense to you? Change the subject.
Which Alert Methods are currently supported today?
Discord, Jira, Opsgenie, PagerDuty, Rootly, Servicenow, Slack
We are already doing SLO’s today, why would we need Nobl9?
Companies we talk to are in different stages of their journey with SLO’s, how are you doing SLOs today? Do you use a dashboard? Now that you have implemented SLOs, when I talk to teams, they see a decrease in pager-fatigue because they are alerting off the burn-rate of the error budgets, have you seen that too? Traditionally we see teams who are doing SLO’s, but are not doing them with error budgeting and burn-down rates so they do not see the benefits of reduced alerts, proactively being able to get in front of problems before they occur, creating common language across the board so cross functional teams can see and have discussions around should we focus on reliability or new feature development and that’s where you start seeing the benefits of using Nobl9
We use Datadog as our monitoring tool; why wouldn’t we just use their SLO platform?
#1 -- Vendor lock-in #1 -- What if another group in your org wants to use New Relic? Datadog isn’t really interested in integrating with New Relic, they want all the data in Datadog.
#2 -- Turf battles -- do you really want to have turf battles with a bunch of different groups across your org to get everyone on the same monitoring platform? With Nobl9, people can keep using whatever they’re using; we’re platform-agnostic.
#3 -- Different monitoring tools are good for different things -- frankly, your org probably will want different tools in different places, and most orgs will probably want 3,4, maybe more different monitoring tools. So why pick an SLO product that encourages you to focus on just one tool? Pick a product like Nobl9 that gives your teams the flexibility to choose.
"Moving to SLOs is our goal but too big of a shift in culture."
I guess the question is: if you were able to have a framework that fostered conversations on where to focus when it comes to reliability, would this will help you work towards implementing SLOs? Would it be worth taking a look?
Why do you price on data sources and objectives?
Like any software company, we have to choose some method of pricing that delivers value to our customers at a fair price relative to that value. In our case, we’ve chosen to price on # of data sources and # of objectives because those are two of the factors driving the most complexity in our platform (in terms of compute and storage) -- and that roughly equates to cost. (Example - it makes sense that it’d be cheaper for us to process 500 SLO / objectives than 5000, right? Or to set up integrations for 5 data sources versus 1?)
Which data export integrations are supported?
Amazon S3, Google Cloud Storage, Snowflake
Why do I need SLOs? I already have DataDog, New Relic and Splunk.
Monitoring helps us gather data to get a clearer picture of what’s happening in our software and systems. SLOs filter and translate all that monitoring data and help you identify precisely where to expend your energy to achieve the biggest return for the business.
We are also considering Blameless as a platform, how does Nobl9 differ?
Nobl9 can scale SLOs beyond the monitoring product. As the monitoring datastore grows, calculations can become heavier with slower response times in Blameless. Blameless has error budgets but they are static and require a page refresh and you need incident response since that's their core product.
"We’re getting foundational pieces in place & SLOs is our next phase."
Great to hear you’re starting somewhere on your SLO journey. Would it be a waste of time to connect over a quick call to give you a better sense of Nobl9's technical capabilities? This could give you some perspective on what other SRE teams are doing to balance reliability and feature velocity and also give you some options to consider for your SLO implementation in the future.
Most software companies price on users; why don’t you?
Great question, and you’re right -- it’s very common to price on users (per-seat license). Right now, as you remember from the $200 & the $400 question :-) , we price on # of data sources & # of objectives because that’s where our computational heavy lifting comes from. However, I’m also glad we don’t price on users right now, because one of the key benefits of SRE is that it drives cross-functional collaboration from different teams looking at the same SLO / error budget data -- so really, it’s in your best interest for it to be easy to add new users to Nobl9 without a lot of process / budgetary overhead.
We need Nobl9 to export to Graphana for visualization? Do you have that on the roadmap?
Yes, we do! It is scheduled to come out during Q1 of 2022
What do you do? Why are you calling me?
I work with developers and reliability engineers like yourself, we allow you to measure the reliability of the services you own, ensuring you hit optimized reliability targets for what your customers expect, while also being imperfect enough to allow profitability for the business.
I see that your organization is practising SRE, are you leveraging SLOs/EBs as well?
We’re already building out SLO management using Dynatrace/Keptn.
Great news! Glad to hear you’re already sold on implementing SLOs! We actually have a great partnership with Dynatrace; we support them as a metrics source out of the box, and frequently collaborate with them on joint customers looking to build out APM and SLO capabilities. The thing to be aware of with Keptn, is while it’s a great tool, it does require substantial custom work to get up and running with your stack as out of the box SLOs really isn’t their focus. If you’re interested, I’d be happy to help you benchmark your current progress, and see where we may be able to offload some of your manual work rolling out these capabilities with an out-of-the-box solution.
"We’re building this internally already."
Some of the teams I’m talking to have also felt this was a natural first step but found that allocation of resources and time limited their full vision of implementing SLOs. Have you explored calculating error budgets yet?
Most companies have pricing on their website - where is yours?
That’s a good callout and I’m glad you brought that up. While nothing is finalized we are close to releasing S&MB and mid-market pricing. Keep in mind, Nobl9 is geared for large Enterprise companies that need to scale their SLO function. At the same time, we recognize companies of all sizes need SLOs for optimized reliability for customer satisfaction. In light of that, we are always willing to work with our potential partners on price. When is a good time to have a pricing scoping call to understand a package that would fit your needs best?
We need Nobl9 to integrate with a few different data sources that you don’t support today. When can I expect those integrations?
Nobl9 is an agnostic solution and we are continuing to scale out with new integrations and features. We have a long roadmap of features and integrations we are working on and what you need could be right around the corner. If Nobl9 is something that you want to move forward with and you have the budget for this year, I can work with our product team to see if we can get these integrations marked as high priority. Keep in mind, we need to make investments on our end to accomplish these integrations and it comes down to urgency from our clients. So we are prioritizing the integrations for customers ready to move forward with our solution. We can always take some time to walk through our roadmap with the product team.
What’s stopping me from just setting up these alerts with my monitoring tools?
You’re totally right - monitoring tools often times do have alerting capabilities out of the box. The issue is, these metrics and alerts don’t necessarily tell you whether or not an issue is actually customer impacting. For monitoring alerts to be effective, you need to 1) choose a threshold thats not so rigid that you’re flagging too many issues to reasonably respond to, and 2) not so low that you’re allowing legitimate performance issues to slip through the cracks. SLOs are by definition customer aligned, and contextualize these metrics so you understand which performance issues are actually affecting your customers, and ensure your teams time and resources are being spent efficiently, improving margins and reducing team burnout.
Was this a fantastic exercise that was loads of fun?
Yes
"We don’t have the budget for something like this."
Sure, most of the people I’m talking to right now have had budget changes. I’ve helped teams build an internal use case by exploring how Nobl9 can help you improve the productivity of your reliability and development efforts, provide better customer experience and reduce strain on your team to deliver on these targets. Interested in benchmarking your current strategy?
"I told you how many objectives we may be measuring currently and data sources we are connecting to, why have you still not given me pricing?
Great question. Here are some more things we would need to better understand before having a separate pricing discussion. Although the # of objectives you may be targeting now can most likely be defined, companies usually scale quickly and this number frequently increases. It is important we uncover a future state as well. On our next call we can dive deeper into defining a current and future state with you and then my team can go back and create a package that makes sense.
Here are some things to think about in the meantime.
- How many teams are you thinking of rolling SLOs out to? Dev team size is a good proxy for this.
- How many developers in each org?
- How many services you have now?
- How many users?
- What’s the scale of your services?
- How many data sources?
- Who is responsible for rolling this out? System admin? Project Manager?