Yonder is a fast-growing authentic internet company on a mission to give the online world the same amount of cultural context as the offline world. “We’re the anthropologists of the internet,” says Taylor McCaslin—Senior Product Manager at Yonder—describing how the company “uses artificial intelligence and machine learning to identify the groups and narratives that drive online conversations. This helps organizations determine what matters most and least and creates the confidence to act.” Yonder helped identify Russia’s campaign to influence the 2016 presidential election, uncovered inauthentic behavior on user-generated content sites like Rotten Tomatoes, and today helps Fortune 500 brands contextualize and understand factions and narratives on public social media platforms like Twitter and Facebook. 

A vision for democratized data

Yonder aggregates around 50,000,000 pieces of social content per day. As the team grew, it neared a “breaking point” as it tried to keep track of the derivative reports, data sources, and what people are doing every day with the data. “Our tribal knowledge,” recalls Taylor, “just wasn’t scaling.” At the same time, the company maintains a proudly transparent internal culture that values broad access to data to allow its employees to make informed business decisions every day. 

Once a week at least, we’ve got somebody who says, ‘Hey, look at this cool thing I did’. And we discuss data lineage, the actual query they used to get the data and from what source the data came from’  While we like everyone doing their own analysis, we want to make sure that the analysis is true, is right, is correct.

That’s a big problem to solve, and Yonder knew they needed help. “We’re a modern data science company, and we need modern data solutions. So we reached out to data.world,” says Taylor. 

The following is an example of one of the many areas Yonder is using data.world to scale. 

One of Taylor’s weekly responsibilities was to hand-update a report that powered several core business metrics, including eight public-facing lead generation pages. These pages “are some of the most impactful we’ve created,” says Taylor. After filling out a contact form, a Yonder sales prospect can see some of the online factions related to a specific industry—a small, enticing window into what data products the company provides for customers at scale daily.

Faction Trackers, as the pages are called, are derived from an expensive, complicated base query. Taylor was in the unenviable role of being a human bottleneck—the only person who knew what the queries were doing and could manually run the report and update the page. 

Before data.world, I downloaded the data into an Excel file and merged that file into a Google Sheet. Then I’d create a Google Chart and embed that chart into our sales landing page. Every single week. Monday morning, I had a calendar invite for an hour to go and update these eight pages. As you can imagine, I did not do that very frequently. I either forgot it or had other things that were more important. And so inevitably, this data sort of got stale, which hurt the effectiveness of the pages. We just weren’t getting the outcomes we wanted from them.

Taylor and team looked at a variety of tools to automate this process. Some didn’t fit the use case. Others just didn’t work. Then he tried data.world. 

We’re reading that data through a project that has parameterized queries  in it. I then generate dynamic links for each one of those variables, and ingest that through an import data function call in Google Sheets. And Google Sheets pulls that URL every 30 minutes and updates my graphs. I can then embed those graphs and they just stay updated with live data queried from Snowflake directly through data.world.

Now Taylor has an hour back every Monday. He’s no longer a bottleneck, and people don’t ping him asking why it’s not updated. “I’ve removed myself from that whole process and can do more important things,” he says. Meanwhile, the data is always up-to-date, without anyone having to do anything.

As the relationship between Yonder and data.world progresses, both teams see Taylor’s early win to be a sign of what’s to come.

We’re excited to expand usage of the data.world platform with more of our employees, roll out a data catalog to further democratize data access across the company, and to put structure around the data products and insights employees use every day to provide value to our customers.

Want to dig deeper into Yonder’s vision for democratized data and how Taylor and team use data.world? Check out this webinar.