Case Study
Jumpstart Automotive Group
07/21/2016
Using Big Data to Plan for the Future
A few Silicon Valley tech giants are famous for having started in a garage. Jumpstart Automotive Group (JAG) is a San Francisco-based company that made the actual garage it’s business. This vertical ad network manages websites for both publishers and advertisers and has since grown to include offices in Los Angeles, New York, Detroit, and Chicago.
JAG monetizes automotive websites for publishers like Car and Driver, JD Power Cars, and US News Best Cars. This focus on premium sites puts them in a position to serve ads to customers who are ready to buy. With people buying cars year-round, there is a seemingly endless supply of customers and this rapid influx of data must be sorted and properly maintained to maximize opportunities.
“For our advertising and publishing partners, the level of granularity we offer into their business and our business is something we put forward all the time,” said Libby Murad-Patel, head of the strategic insights and analytics group at JAG.
Their detailed approach meant that as the business matured, JAG needed a data solution that could handle large amounts of information.
That’s where Zencos came in.
“We knew to try to consolidate company performance metrics and getting a good understanding of our business on a regular basis was very complex. We were doing a lot of manual work to pull in data sources,” said Murad-Patel. “We were spending a lot of time trying to come up with Band-Aid solutions.”
In 2010, Zencos helped JAG build a system that consolidated data. Consultants from Zencos sat down with the JAG team early on and got the stakeholders together, looked at the needs, and customized a solution to fit the business. Before Zencos, one weekly report took 12-18 hours for a person to compile. Today, the same report is pulled daily in under an hour.
“The Zencos team is like an extension of our team,” said Murad-Patel. “Ben, Tricia, and Kim have been involved since day one. So when we need to troubleshoot they get our business.”
It turned out, not only did JAG need to control the flow and output of data, but it also needed to clean incoming sources of data and produce packets of information for easy analysis. Zencos implemented a structure that allowed JAG to review information in one-hour packets. This bundling of data helped JAG more accurately forecast revenue, isolate trends, and plan for the future.
“Now we can spend more time on analysis and ad-hoc reporting,” said Murad-Patel. “We can go straight to a deep dive without having to manipulate and cleanse our data. We can get into the data and start looking for trends.”
JAG looks at the consumption of ads and site traffic and finds patterns in the data, filtering through about 900 million impressions a month. To handle this demand, Zencos built a web portal, a custom interface with dashboard and alerts, a data integration platform, and designed the architecture to scale with a growing business.
A long way from Excel, JAG now uses pure log data in a raw format and fewer outside vendors for data sources. Reducing the number of data sources helped JAG control what was coming in and assign the same variables to each piece of data, allowing for easier manipulation.
“We plan 18 months in advance,” Murad-Patel said. “By having the ability to forecast and look at trends of previous years and also the way our inventory is pacing regularly, it gives us a much better lever to look at the way we monetized the website and if we need to shift the way we capped a particular site.”
Zencos developed an easy-to-use web-based system for analyzing ad campaigns. The system produces
a number of financial measures by integrating and analyzing huge volumes of clickstream data, and other online sources to provide the optimal ad placements for advertisers across their publisher’s networks.
The system architecture leverages both physical on-premises infrastructure and a cloud component to manage the cluster required for analyzing vast amounts of data. This provides a system to track under-performing and over-performing campaigns, total ad impressions, revenues, and other financial data. The system combines data from several online sources and supports dashboards for decision-makers.
Prior to partnering with Zencos, JAG had a KPI forecasting accuracy of about 7 to 10%. Today, they can achieve accuracy within 1 to 3%. This has increased their ability to assess profitability. Often, they will pull a KPI report halfway through each month and measure projections. If they are within 5% of their target, they can move forward confident that they will hit their goal.
When a Company Becomes Big Data
“As a stakeholder in the project, I’ve been involved from the beginning. Our goal was to get more accurate with our data. This helps us when we go to our clients and when we communicate internally,” said Murad- Patel. “We have a complex business model, and as we refine our data sources we will continue to make our projections richer.”
More confidence in her data freed Murad-Patel and her team to focus on resolving issues and making decisions. Each phase of development improved accuracy and allowed for more confidence at multiple segmentations. The vast amount of data JAG brings in is very detailed and granular all the way down to the head level, making manipulation and calculations very complex. The system Zencos designed performs “flight period” calculations that account for the time period and ad grouping, and it incorporates billing rules, allowing them to forecast revenue.
But there is room to grow. Moving forward, there are existing opportunities to bring in new data sources and unite stakeholders at the company around data-driven decisions. As JAG continues to collect more data and works to maximize efficiencies in their business, the next step is finding ways to apply visualizations to data and facilitate further communications internally. There is also increased pressure in an ever-growing digital landscape to more accurately forecast revenue and target resources. Zencos is currently working with team members to develop a strategy for the future and continue to shape solutions that have the ability to scale alongside organizational growth.
For Jumpstart Automotive Group, the key is clean, accessible, and accurate data. Zencos is willing and able to assist. After all, there’s Big Data in what consumers park in the garage.