Tempo Quickly Gets Their Data Infrastructure in Shape With Mozart Data
Results
- 99% decrease in implementation time
- 91% decrease in query runtime
- 76% decrease in time to insights
Company Size
200+
Departments
Operations, Product, Marketing, Customer Support
Data Sources
Tools
Decentralized data and lack of infrastructure prevented Tempo from making use of its data
At Tempo, the award-winning AI-powered home gym, the operations team is responsible for supply chain, member experience, customer success, sales ops, and analytics. While they had lots of member and product data to work with, they didn’t have the data infrastructure that’s needed to easily make sense of their data. “We had data in each of our respective tools that we used, and those obviously have their native reporting. But naturally, it’s decentralized. So it was challenging for anyone to do any sort of analysis,” explained Dan Cho, Senior Director of Operations at Tempo.
They needed to figure out how to centralize their data and get it to a place where they could access and query their data through their business intelligence tool, Mode. Plus, “There was no way I could spend much time transforming data, and it was unclear what tool to use,” said Cho. Without the right tools and resources, the team wasn’t able to make use of their data.
Tempo implemented Mozart Data around the same time they hired their first analytics engineer. This set them up to finally turn their raw data into insightful reports, and it saved their engineer from the dull work of talking to vendors and setting up contracts.
“Mozart simplified a lot of the setup. I didn’t have to create our own version of Snowflake or Fivetran. It all came as a package. All I had to focus on was writing the transforms and connecting the data sources to Mozart,” said Doo Shim, Team Lead, Analytics Engineering and Business Insights. It took less than an hour to set up Mozart Data, while building their own data stack would have taken at least a few weeks.
Moving from manual spreadsheets to automated reporting
With more accessible data, this opened up endless possibilities for reporting. The first initiatives the operations team tackled were executive, finance, and supply chain dashboards. These teams had been working out of Google Sheets, which required people to update their numbers every week.
The operations team replaced the spreadsheets with automated dashboards that also established a baseline structure for company reporting. “The first iterations started with customer support data.
And that led us to being able to show that with this pretty simple, straightforward data stack, you can do things that previously took a lot more effort and money,” said Cho. “We wouldn’t have been able to build these dashboards if not all the data was centralized and organized. That was our first win as a team.”
Since building out their first dashboards, the team expanded on that and found creative ways to help other departments quickly access the data they need. For example, they built a lookup tool that enables the customer support team to quickly pull up a customer’s history and product information when they call in. Prior to this, customer information wasn’t centralized, which made it difficult and time-consuming for support agents to find the information they needed to help customers.
Saving data analysts and engineers time
In order to extract information from their data, Tempo’s analysts had to figure out what data was coming from which table and understand all the raw tables and joins. It could take up to a few days to sift through the data and build the logic for a query. Mozart Data provides them with one place for all their data and automated transforms that run consistently, which has sped up Tempo’s time to insights and simplified their queries.
“If we didn’t have Mozart, we would have to connect the raw tables or the transforms,” said Gautham Sista, Senior Analytics Engineer. “I’m able to combine data from multiple systems and build one core table that we get most of our answers from. What used to be multiple joins in a query is now a select star statement with some filters, so it makes life so much easier.”
Shim estimates “A query that would have taken us 20 minutes to write is now down to five minutes because we have to join fewer tables and we’re able to clean the data in a data warehouse.” Runtime has decreased too — from more than two minutes to as little as 10 seconds.
Capturing a view of their data at a certain point in time, which is known as snapshotting, enables Tempo to track their historical and updated inventory statuses. In other transformation tools, “Snapshots was a wild animal. It was very hard,” described Shim. “Mozart just gave us a toggle option, which was very easy to use. We didn’t have to spend the time writing code for snapshots. We could just toggle it on and off, and it saved us a lot of time.”
Growing into a proactive, data-driven organization
“The executive dashboards really proved the value of data at Tempo,” said Shim. When the company didn’t have a centralized source of truth for the business, they were reactive to issues on a daily basis. The executive reporting gave them visibility into their weekly and monthly performance. They could now see what was happening before an executive asked or a customer contacted them about a problem.
For example, if they see their sales for the past seven days is a lot lower than that of the previous week, they can proactively work with the marketing team to figure out what’s happening. Or if they see cancellations spiking, they work with the member experience team to get to the root cause and come up with ideas on how to reduce cancellations.
“These reports were a great place to proactively start asking questions and come up with solutions. The reporting helped us ask a lot of follow-up questions when the numbers were too low or too high and be less reactive.”
With more data, they’ve also been able to measure the effectiveness of their campaigns. “One of the things we weren’t able to do before was measure what the impact of our lifecycle marketing campaigns was on engagement. Now, we’re able to get insights on was this campaign actually effective and are users actually going back, taking classes, and being more engaged,” said Sista.
Mozart Data has given every department the tools they need to make sure data is accurate, properly collected, transformed, and deduped. This doesn’t just help them get the insights they need to do their jobs, but it also sets up each area of the business for long-term success. “For the long-term success of this company, we have to be collecting data well now so we can make better decisions, we know what’s happening, and we have that visibility,” said Cho.
