Marketing with Big Data: Customer Spotlight Merkle

CTO, Digital Media and General Business, Netezza, an IBM Company

Architecting Big Data Solutions Series

KP12.8.png"Our real secret sauce is to take our analytical skills and capabilities and blend that with our offerings around Big Data technologies to execute more effective marketing for our clients", says Russ Pearlman, CTO of Merkle Inc.

We are seeing a growing trend, especially in the Digital Media space, where the most successful and innovative organizations claim that their key to success lies in their unique ability to marry complex analytics with Big Data know-how. For example, Zynga, the hottest social gaming company in today's market, claim that their "secret sauce is that they are really an analytics company masquerading as a games company".

Russ was speaking at a recent Direct Marketing Association (DMA) webinar on how Merkle pioneers Marketing with Big Data. Merkle is one of the largest Customer Relationship Marketing (CRM) agencies in the world and at the core of Merkle's Big Data platform is a "Knowledge Warehouse" that aggregates customer information across multiple sources that include data from all customer touch points such as call centers, point-of-sale systems, web sites, social media, email campaigns, contact preferences, offline channels and third party consumer data. The ability to aggregate these disparate sources and analyze data quickly to enable more targeted marketing campaigns is critical to their business. Within the Knowledge Warehouse, they maintain customer reference bases, support campaign analytics, power business intelligence data marts and enable real-time analysis.

Russ provided great insights into some of the challenges and imperatives in architecting their Knowledge Warehouse data platform:

  1. Deal with Big Data

Merkle manages over 125 CRM databases for some of the best brands in the world. Their data includes customer information provided by these brands and includes online behavioral activity and offline transactions. This is in the order of Petabytes of data with hundreds of millions of new transactions coming in daily that need to be managed and analyzed. Also, they have to deal with a wide variety of data that they receive from their customers. Their data files may be in different formats and may include complex data such as digital activity (click-stream) on a website. They also may have to act upon customer information in real-time and manage offers in real-time. Managing high data volumes across a wide variety of types at an extreme velocity creates unique challenges that most traditional data systems are unable to handle.

  1. Reduce Data Latency

There is an increasing business demand from Merkle’s customers to reduce campaign execution time. This means that the amount of time it takes from acquiring customer data to providing effective analysis needs to be shortened. Data latency also impacts human productivity, especially analysts, who have to wait for data to land into their Knowledge Warehouse before it can be acted upon.

  1. Manage Complexity and Cost

Hardware, software, data center and personnel costs for managing Big Data across multi-clustered systems can be high. Also, it can take a lot of resources to maintain the underlying infrastructure. Keeping complexity and cost under control and passing those benefits back to Merkle’s clients is a key consideration for their system architecture.

  1. Improve Analytical Results

Delivering accurate marketing results is critical to Merkle’s business success. In order to achieve that they need to run a large number of complex analytical models, decrease execution times for those models and enable analysts to work on their client’s entire data set and not just a sample.

Merkle standardizes on the IBM Netezza Datawarehouse appliance as the key component of their Big Data platform as it helps them overcome the above-described challenges. In the webinar, Russ outlined their technology evaluation process and rationale for standardizing on IBM Netezza. Their investment in IBM Netezza technology has enabled them to decrease data latency and improve revenue lift with a 66% decrease in the cost of managing client environments. Ron Par, Vice President Quantitative Solutions at Merkle, summarized the business benefits achieved via adoption of IBM Netezza as thus:

  1. Response and revenue lifts of 20-35% in marketing campaigns enabled by leveraging a significantly more robust source of data.
  2. Improved ability to control campaign cadence to create better control groups allowed for identification of incremental marketing opportunities yielding up to $10MM in operating income.
  3. Regularly received a 70 percent reduction in processing time for complex marketing campaigns decreasing time from hours to minutes.
  4. 50 percent decrease in end-to-end run time for marketing campaign execution—from sample to test to final version.

Architecting Big Data solutions can be a complex endeavor. It is always fascinating to learn from practitioners about their process and successes. We have also published a more detailed case study with Merkle. Hopefully it helps you with your journey.