MEC, a leading global media agency, wanted to provide better insight into online behaviour. The company needed better tools for handling data, so it based its new CookieCutter service—a digital attribution platform—on the Microsoft SQL Server
Parallel Data Warehouse. As a result, MEC is cutting query time from four hours to minutes and gaining better insight into cookie-level data across multiple media channels. As a result, the company is helping marketers unlock the value of digital data. In
addition, MEC is empowering users with access to self-service business intelligence (BI) tools and building a platform for more innovative services.
Headquartered in London and New York, MEC is a leading media agency with 150 offices worldwide, 5,000 employees, and$25 billion in global billings$. In addition to its core planning and buying services, the company’s offerings reflect current technology
and marketing trends, with solutions that include online and social media, mobile platforms, gaming, and TV.
MEC wanted to help its customers get greater return on investment (ROI) from digital information by providing more insight into online activity, but it needed better tools for handling data. The key challenge was digital attribution, which tracks user behavior
online prior to a sale.
Most analysis is based on last-click attribution, which focuses on the final touchpoint before a purchase. Instead, MEC wanted to look at the consumer’s entire online journey, which could include multiple website visits, searches, email, and social media.
MEC was already working with large volumes of complex data pulled from ad servers, and it could take several hours to run queries. To dive more deeply and quickly into digital attribution, the company needed different tools. Christian Bonilla, Director Analytics
and Insight at MEC, says, “We were looking at data that was not cookie-level data, and it was already taking three to four hours. We knew our current solution wasn’t the right fit.”
||SQL Server Parallel Data Warehouse gives us massively parallel advantages. Whereas it would take up to four hours to run queries scaling across multiple nodes, now it takes just minutes.
| Trevor Attridge
Chief Information Officer
In June 2012, the company began implementing the Microsoft SQL Server Parallel Data Warehouse (PDW). In addition to boosting performance and analytics capabilities, MEC knew it would gain faster time-to-market. A prebuilt appliance with software, hardware,
and networking components already installed, SQL Server PDW would simplify deployment as well as provide built-in BI tools.
Working closely with the Microsoft Services Premier Support team, MEC began implementing its new CookieCutter application with SQL Server PDW. The company can use SQL Server Integration Services to access and transform large volumes of data from virtually
any source. Client teams and data technology specialists then work with a variety of analytics and visualization tools. For example, specialists can use SQL Server Analysis Services to build complex queries, and business users can quickly model data with SQL
Server PowerView, an interactive data visualization feature in SQL Server Reporting Services.
SQL Server in-memory technologies such as the xVelocity memory optimized columnstore index also significantly speed query performance.
The solution runs on the HP Enterprise Data Warehouse Appliance. Launched in May 2013 in the United States, CookieCutter is now rolling out worldwide. Multiple clients are already using the service, Xerox and Michelin being the first to onboard onto the
new platform for digital attribution.
By implementing a digital marketing attribution platform based on the Microsoft SQL Server Parallel Data Warehouse, MEC is significantly boosting performance, improving business insight, and enhancing services.
Speeds Access to Information
The new data warehouse is improving data management from acquisition through analysis. The solution not only speeds performance, it also helps the company to work with larger volumes of complex data. “In the past, it could take us about two and a half days
to load a large data set,” says Bonilla. “With SQL Server PDW, we can load the file in about an hour.”
Query performance is significantly faster too. Trevor Attridge, Chief Information Officer at MEC, says, “SQL Server PDW gives us massively parallel advantages. Whereas it would take up to four hours to run queries scaling across multiple nodes, now it takes
Gains Better Business Insight
Analysts now spend less time acquiring data and more time helping customers optimize marketing across multiple digital channels. The solution helps MEC better serve an increasing number of clients that want to quantify the ROI for digital media.
“With SQL Server PDW, we gain a new, scalable, and flexible solution that adapts to the nuances of a marketer’s business,” says Bonilla. “At MEC, we take a consultative approach to supporting our clients’ marketing decisions. CookieCutter with SQL Server
PDW reflects this approach and helps us deliver more streamlined reporting and faster recommendations for more actionable results.”
Improves Access to Data
Attridge believes that by improving access to data through self-service BI tools, the company can stay on top of media trends more easily. “This project is also about empowering users as much as possible,” he says. “As the business evolves, it’s essential
that we get these advanced tools into the hands of people who are experts in their field.”
Supports More Innovative Media Services
With the new CookieCutter platform, MEC can offer clients a customized, highly detailed analysis of data that goes far beyond last-click attribution. “We are actively looking for new data sources because with SQL Server PDW, we know we can handle it,” says
Bonilla. “Right now we’re working with online data, but we’re looking forward to bringing in TV, print, or any other data set. Instead of looking at media channels in silos, we will be able to see the client’s business more holistically.”
This case study is for informational purposes only. MICROSOFT MAKES NO WARRANTIES, EXPRESS OR IMPLIED, IN THIS SUMMARY.