How Zeplin Uses the Modern Data Stack to Answer Growth Marketing Questions
Growth Marketing, Operations, Product, Customer Support
Zeplin needed a reliable source of truth
Founded in 2014, Zeplin is a design delivery platform. Zeplin helps technology teams — designers, developers, PMs, and more — collaborate on, organize, and deliver beautiful products together on schedule.
By 2020, the growth marketing team at Zeplin had reached a data management roadblock — namely, that by default they were using Salesforce as their data source of truth. Using Salesforce as an impromptu data warehouse meant constantly seeking workarounds for problems in the form of strung together integrations and countless Excel spreadsheets. The process was unwieldy and prone to breaking, resulting in unsynced data, CSV errors, and limited reporting.
Jason Feng, Zeplin’s Director of Growth Marketing, knew that data options had drastically improved since 2014. “When I first heard about a data warehouse, it was described to me as a database of your CSVs that you can query across. We were used to having everything in spreadsheets. That’s what resonated with me,” said Feng. He set out to find a modern data stack that could help them make the most of their high volume of customer and prospect data.
They considered building their own data stack, but chose Mozart's fast, reliable out-of-the-box option
Zeplin considered a number of approaches to the modern data stack. One option was to build a data warehouse themselves, but they lacked the expertise, knew it would take resources away from other departments, and were concerned about how long it would take.
Another option was to hire a consultant to build a data stack for them, but the estimated timeline was too long for their near-term goals, and they didn’t have someone in-house with the experience to manage the consultant.
A third option was to piece together a collection of tools themselves. They considered using Segment (a customer data platform) alongside Salesforce and Mixpanel. But Segment wasn’t designed to be a data warehouse and would have limited organizational and analytical capabilities compared to a purpose-built data warehouse.
They ultimately chose Mozart Data because Mozart offered them all the capabilities they needed in a modern data stack, along with the expertise Zeplin lacked in-house.
How the team answered questions about user limits, revenue, and lead scoring with the modern data stack
Once their data stack was set up, Zeplin could tackle critical projects, starting with getting a better understanding of their customer. “With all the data in one place, we’re able to look at the data together, in a combined way, to give us a much better view of the customer,” said Feng.
Some questions Zeplin was able to answer:
What’s the right user limit on our free plan to better monetize users?
Zeplin wasn’t sure how to set user limits on the free plan. They wanted to create an experience where new users could use Zeplin without many restrictions, and power users would upgrade to a paid plan. So it became important to determine the right user limits.
They took product usage data in Mongo and examined how many users were on each of their free plans, breaking them into groups based on what percentile of active collaborators they fell into. By looking at how many plans there were at each level, and combining that with payments data from Stripe to compare the number of users on free plans to paid plans, they could identify where reasonable limits were for their users.
How do we split out self-serve and sales revenue and use that information to assist growth marketing initiatives?
Zeplin didn’t want to just identify customers that came in through the self-serve channel versus sales channel. They also wanted to understand their customer profiles and the different actions those types of customers take in order to drive further growth.
This required combining data from Mongo, Stripe, Salesforce, and Intercom in one place, so they could examine the entire journey of the customer before they converted to sale — including the channel they came from, their product usage history, and any engagement with a Zeplin representative after joining the platform. They also examined traffic profile, the size of the customer, and the customer’s own business growth to accurately categorize them. As a result, they accurately determined whether or not a customer was a true self-serve user or a sales-assisted customer, and used this information to strengthen their customer profiles for Marketing and Sales.
What’s the criteria for a product-qualified lead?
The work Feng’s team did to assess revenue directly led to a project to better qualify leads.
By combining product usage data and Stripe transaction data, Zeplin began to identify the actions that indicate a happy customer who is likely to renew or expand their account.
When users with free accounts start taking similar actions as these customers, Zeplin classifies them as “product qualified leads” because the way they’re using and getting value from the product indicates a strong opportunity to upgrade them to a paid plan. They also use targeted marketing campaigns and sales efforts to increase the likelihood that these users upgrade.
To aid customer satisfaction and retention goals, Zeplin sought a way to communicate more consistently and productively with their customers. They automated a monthly report showing users facts about their usage — like the amount of work they have published and the number of people they collaborated with — and unified it with their lifecycle marketing tool to share it, all without a time-intensive repetitive process.
Growth Marketing leads the way in data fluency and democratization
It’s unusual for a marketing team to implement and drive adoption of a data stack. But as marketing moves into owning the full sales cycle, Zeplin’s Growth Marketing team is making this the norm and leading the way. “If you believe Marketing is responsible for full-cycle sales, you need to have a ton of fluency in your data, to be able to know who are the people signing up, what are they doing in the application, what package did they buy, when are they going to renew, and so on,” said Feng.
Mozart Data’s modern data stack enabled Zeplin to put down their collection of spreadsheets and short-term Salesforce integration fixes. Zeplin can now trust their data and ask more complex questions because they have the tools to easily find answers in their data.
Although Marketing was the first Mozart Data user, other teams have also benefited from having a modern data stack. The customer support team has been able to dig further into customer adoption of their product to provide better support. Engineering has used product usage data to better understand customer needs as they refine product roadmaps and organizational features within the platform.
“You’re going to realize these sorts of data problems are in every single department. If you can uncover these problems, you’re going to have a business case where you can say that 50% of the company will benefit from building a data warehouse. And at that point, the ROI becomes very evident,” explained Feng.