Loom improves support response speed by activating real-time customer data

Loom is a video messaging software that gives users a more effective way of communicating with coworkers, employees, and even students and teachers. Since 2015, they’ve been providing a platform to make simple, clean videos in the hopes of removing long back-and-forths over email and in-person meetings.

When Senior Data Scientist Buddy Marshburn was hired at Loom, he realized there was a data access issue; Segment, their existing solution, was only able to pass through event streams to the tools the teams used. With only raw event data, it was difficult to power use cases that depended on aggregated metrics or modeled insights. For example, triggering targeted Intercom messages based on usage patterns or lead scores simply wasn’t possible without a way to sync modeled data from Loom’s Snowflake warehouse. 

Marshburn knew from experience that managing this type of data flow from Snowflake to Salesforce, Zendesk, or Intercom isn’t simple — it requires several people to build custom data pipelines, which then need to be maintained and monitored independently. As a data team of one, Marshburn lacked the time or capacity to take this on.

A large part of Loom’s feature release process depended on data collected through Intercom forms, but Marshburn noted that they couldn’t access the information they needed to confidently move those releases forward. This limitation created a bottleneck the team needed to overcome before shipping new functionality.

That’s when Marshburn began searching for a tool that could help him bring transformed data from Snowflake to Intercom.

Discovering Census (a Fivetran company)

When Marshburn started looking, he wasn’t sure anything on the market could solve the problem, so he turned to the dbt community for recommendations. That’s where he was first pointed toward Census (a Fivetran company), a tool he hadn’t encountered before. “I looked into it and immediately thought, ‘This is kind of like Fivetran flipped around to the other side of the data warehouse. It's exactly what I'm looking for!’” Marshburn says.

He could see the impact that Census could offer right away. He says, “My whole career as a data scientist and in data engineering roles has been about setting up foundational architecture components, so I very quickly realized the value prop Census offered. I feel that anyone looking to send data to any of the [destinations] Census provides will quickly see the value.”

Census: The solution to tackle a sea of support tickets

Initially, Marshburn began using Census to help Loom connect data to Zendesk, which allowed them to prioritize the increased volume of tickets that were coming in. This capability became especially critical during the pandemic, when Loom opened the product to students and teachers for free. The move drove a massive increase in users and, in turn, a dramatic spike in support volume that Census made possible to manage effectively.

Loom wanted to prioritize tickets by a user’s plan type, but this quickly became complicated because a single user could belong to multiple plans. “You can't really use Segment for this,” Marshburn says. “Ultimately it requires SQL to create a list that says, ‘you're a member of all these plans.’ Now let's take this list and map it into one single value.”

Instead, Census allowed Marshburn to take this transformed data, relevant for ticket prioritization, and easily populate the necessary fields in Zendesk.

“Census made something that was impossible before very highly functional. Before, we were diving through a sea of tens of thousands of tickets. Census solved the foundational problems that we needed to have fixed.”
— Buddy Marshburn, Senior Data Scientist at Loom

The impact: Activating data for every team

Census organized all of Loom’s data activation needs in one place, saving them time and eliminating the need to hire additional engineers. With dbt already transforming thousands of raw tables into a clean, curated layer, Census allowed Loom to fully leverage that work. This ensured that Salesforce, Zendesk, and Intercom were all powered by the same accurate, up-to-date customer data, effectively creating a more consistent customer 360 across the business. No additional engineering, pipeline building, or maintenance was required — a process that previously would have demanded substantial development effort and ongoing monitoring.

Marshburn also points out that as data teams grow, so does the amount of effort needed to get data out of the warehouse, which won’t be the case now that they have Census. “That's where Census is the missing puzzle piece,” says Marshburn, “because you're easily able to take all the work that a few people have done and put it in the hands of your stakeholders.”

As adoption grows, Loom is focused on scaling collaboration for enterprise customers, with data playing a key role in that strategy. Marsh sees Census expanding far beyond its initial support and customer success use cases, powering data-driven decision-making across teams at Loom.

[CTA_MODULE]