Executive summary
Dentsu, one of the top five global advertising agencies, leverages its deep analytics expertise to deliver data-informed services to its more than 11,000 clients worldwide. The amount of data Dentsu accumulates on a specific campaign is huge: 10 terabytes, 100 terabytes, or even a petabyte for a single campaign.
Challenges
Dentsu first built their data management platform in-house taking more than a year, hard work, and substantial expense to complete. Each client needed a highly customized version of Dentsu’s platform, blending contextual and psychographic data with the client’s own success metrics to create the best algorithms. This meant that each new account required months of custom integration work to collect data from the customer’s POS, CRM, production databases, and more. And each new instance required engineering resources to maintain it.
Solution
Using Treasure Data Customer Data Platform (CDP), Dentsu’s data scientists could blend data from multiple sources and create advanced analytical workflows on the fly without any resources from IT. Instead of building custom connectors for an ever-increasing variety of data sources, Dentsu used Treasure Data’s plug-and-play integrations to set up new data feeds in minutes.
- In under a month, they centralized hundreds of customer datasets in Treasure Data’s flexible, scalable cloud
- The implementation time for each new customization dropped from months to weeks
- In addition, the Treasure Data CDP scales well and provides better performance than the original in-house system
With the addition of Treasure Data, Dentsu is now a one-stop customer insight portal where all of its clients can discover consumer trends, launch highly targeted campaigns and intelligently nurture leads. By removing the obstacles between data and results, Treasure Data has exponentially increased Dentsu’s ability to deliver cutting-edge technology to its global brands.
Results
With Treasure Data, Dentsu was able to:
- In less than 1 month, centralize hundreds of customer data sets
- Customize client instances from months to weeks
