SmartBear offers a full range of products to help companies with the quality of their software infrastructure.
Needed an easier way to access cost per customer metrics to gauge usage and maximize value. Also, needed to overcome tagging issues for when new products joined the company.
Implemented CloudZero and developed an increasingly granular understanding of cost incurred by customers.
- Altered strategies according to customer types, thereby raising revenues without expending additional engineering resources
- Enhanced their understanding of the marketplace, improving go-to-market strategies
- Created better alignment between engineering and finance orgs
- Cleared away infrastructural excess
SmartBear is a software company offering more than 20 tools for developers, testers, and business teams. While they cover just about every aspect of the development lifecycle, every decision they make centers around the one thing that never changes for software: quality.
With the tagline “Quality isn’t just a goal. It’s the whole point,” to help guide them, SmartBear has some high-profile, innovative customers. But no matter who they work with, their aim is to help customers make software as fast, accurate, and error-free as possible.
SmartBear has gradually grown through both evolving its tools and acquiring new ones, and unifying them all under the SmartBear umbrella. What results is a bevy of quality assurance tools ¬– not to mention some prestigious recognition.
For SmartBear, the theme of quality runs deeper than customer only – they use it as an internal compass as well. One challenge came in the form of cost analysis.
An Atlassian “Platinum Top Vendor,” SmartBear offers numerous apps through the Atlassian marketplace – effectively an App Store for businesses. By selling through the Atlassian Marketplace, SmartBear can make greater investments in product quality. But, until recently, they lacked the kinds of granular insights that would allow them to maximize their margins.
“As a company driven by quality, we wanted a tighter concept of what our unit cost per customer was,” said Laurent Py, SmartBear Senior Vice President of Product Management. “We could see overall costs on our AWS bill, but wanted it to be simpler to itemize and sort through. The goal was to make it easier to attribute specific costs to individual customers.”
As a result, SmartBear felt that its pricing models were based on incomplete information. A multi-tenant SaaS environment, SmartBear wanted to quickly know how much it cost to house each individual tenant, not just the overall cost of running the environment.
“Our customers span a huge spectrum, from tiny customers on the free tier of Jira, to enterprise-cloud editions of 100,000 or more customers,” Py said. “Not having immediate access to cost per customer was a gap we wanted to fill in our business.”
In the past, SmartBear had used an array of cost-analysis methods, including manual AWS services tracking, the AWS Enterprise Discount Program (EDP), and CloudCheckr.
“The value from our existing tools left more to be desired,” said Martin Loewinger, Vice President of Cloud Engineering. “We had the same issue with the EDP, so we ultimately went back to tracking AWS services manually. Our costs were stable for about a year, but when they started to go up, we knew that manual tracking wasn’t going to cut it.”
SmartBear knew that accurate cost per customer metrics would keep them ahead of the curve, allowing them to optimize their pricing and maximize their margins. Those metrics would also help them achieve manageable, predictable growth. Prior tools hadn’t done the trick – and they needed to try something new.
Py directed SmartBear to CloudZero, a cloud cost-intelligence platform that delivers useful data about products and features to the engineering teams responsible for building them.
When asked to describe the onboarding process, Loewinger described it as “Quick, good, straightforward,” recalling that CloudZero was configured into all 10 of their cloud accounts in less than an hour. “By the next day, we had the data we needed. We had known the overall cost, but now it was broken up by services, groups, customers, etc.”
In particular, CloudZero helped SmartBear overcome the tagging challenges that accompanied its growth-by-acquisition strategy.
“Everyone’s got their own tagging strategy,” Loewinger said. “So when another group joins the fold, it’s very hard to say, ‘Conform to this, this is how we track cost.’ Most teams have a hard time because they’d have to redo all their automation. But our bottom line is to focus on the customer, not get caught up in things like tagging.”
CloudZero was built partly around the premise that while organizations leveraging AWS benefit from perfect tagging practices, very few can maintain perfection, especially as they scale. Resisting the tedium of tagging can lead to data-analysis issues down the road, particularly when it comes to shared and/or untaggable resources.
CloudZero gave SmartBear the ability to gain consistent visibility across their many products without undergoing lengthy tagging exercises.
In its early stages, SmartBear and CloudZero had weekly meetings to ensure that SmartBear was accessing the full value of their new cloud cost-management product. CloudZero initially provided high-level cost per customer metrics, which SmartBear made increasingly granular through extended collaboration.
“We initially had information from Atlassian about billings and revenue,” Py said. “Now, we monitor customer-specific API calls in proportion to our overall AWS spend to figure out who’s using our products, how much, and at what cost.”
This provides them with essential visibility into key financial metrics.
“Having CloudZero really helps us understand cost per customer,” Py said. “In turn, that lets us understand multiple segments of customers, including their size and the markets they’re in. When we have information around our customer base, it lets us understand pricing, improve our margins, manage EBITDA, and optimize our business as a whole.”
Aligned costs and go-to-market models
The visibility CloudZero provides into cost per customer and product has enabled SmartBear to reassess optimize our strategies to ensure our costs and our go-to-market models are well aligned.
“That information allowed us to work with our product marketing team on how to continue to package our products in a way that so our customers always receive maximum value, while also supporting our business model. It has also helped us make earlier-stage development decisions to support strong cloud economics.”
Reduced operational excess
CloudZero allows many customers to see their data landscape at a level of detail not previously possible. This often turns up some very apparent excess. Because CloudZero puts this information in the hands of the engineers best equipped to handle it, the most obvious excess is easy to clear away quickly.
SmartBear was no exception. “When we saw anomalies, our head development architect very easily determined what we did and didn’t need, and disabled certain infrastructure quickly,” Loewinger said. “In one account, disabling an AWS config drove down cost by $1,000 per month.”
Additionally, a more refined portrait of what business needs cost produced greater alignment between the SmartBear engineering and finance orgs. “We’re doing a better job of dividing up financial information by individual products. We build cost centers in CloudZero, which makes it simpler to tell finance, ‘Here you go, now you have a better picture of where and how.’”
Improved engineering use and go-to-market strategies
“CloudZero has helped us raise revenues without using additional engineering resources,” Py said. Efficiently gauging usage allowed SmartBear to maximize their products’ value, and to invest more resources in building additional functionalities within each product. “We’re building overall value,” Py added.
Plus, a more thorough comprehension of their customers’ usage habits helped SmartBear better understand the market(s) in which they operate as a whole, thereby refining their marketing decisions without incurring additional R&D costs. “It helped us make really good decisions on go-to-market strategies,” Py said.