EyeBall Division is a data science-driven performance marketing agency based in Park City, Utah. Their team of expert media buyers uses machine learning and data science to acquire target customers for their clients. According to EyeBall Division CEO and Founder Andrew James, “What we ultimately are striving to be is a platform and a place that you can go to — a mix of machine and human — and get whatever idea you want in marketing turned into data and revenue.”
Since becoming a TUNE customer in 2010, EyeBall Division has built and operated their advertising platform entirely on top of the HasOffers API. As a data-driven performance agency, the team relies on up-to-the-minute information to power their evolving set of algorithms and tools. But with every new client, EyeBall Division CTO Chase Cottle was having more trouble consolidating multiple data sources and validating data integrity, due to being limited to pulling data from the API in batches every 30 minutes.
“We were pulling data as fast as we could, but as our business grew we were approaching API rate limits, and we didn’t have all the data we needed fast enough,” Cottle said. He had to wait until after midnight each night to update the data warehouse and pull data from the previous day into client dashboards for the next morning. He also spent up to 5 hours each day validating data, because correcting just one error (such as an incomplete tracking parameter) required him to reload data from an entire hour or day. “We were quickly approaching a serious issue that would have caused major interruption to our service,” Cottle said.
When EyeBall Division heard that HasOffers provides real-time data streaming through Firehose, they knew it was the solution they needed. By working closely with HasOffers solutions engineers, Cottle’s technical team was able to implement Firehose on their custom HasOffers platform. Now, the data transfer process happens seamlessly with less than 100 lines of code running through AWS. “It’s made a huge impact on our time to focus on our next priorities, and we are using the data to draw better conclusions and deliver to our clients faster algorithmic analysis,” Cottle said.
Since implementing Firehose by HasOffers, Eyeball Division has moved from batched data updates with 30-minute latency to real-time data streaming. This has allowed the company to save 42% on monthly data extraction and transfer fees while dramatically increasing data integrity and availability. With no need to spend up to 5 hours a day validating data, Cottle now uses his time to build new systems and products, improve algorithms, and work more closely with clients. In addition, CEO Andrew James has been able to spend more time focused on improving his team and finding new opportunities to move the needle: “The things I worry about have changed. And I think that’s actually the biggest benefit.”