David Norton’s 4 Secrets to Understanding Customers Through Analytics

 

If you've ever vacationed at a Caesars Entertainment property and used a Total Rewards card, you've probably received an offer from David Norton. Every move made by customers at Caesars' properties is tracked, the data feeding into Caesars' central customer database–data that is later mined to improve the casino's direct marketing efforts.

Norton spent 13 years in the gaming industry, first working for Harrah's Entertainment, which eventually purchased Caesars in 2004 and renamed as Caesars Entertainment in 2010. In charge of customer insights, loyalty and multichannel marketing, Norton rose to become the SVP and CMO, where he won CMO Magazine's 2010 Chief Marketer of the Year award.

Norton has since moved on to a new position at MDC Partners, where he acts as the EVP of Integrated Customer Engagements and Analytics. Norton is now responsible for building a unified customer centric and analytics practice throughout the 50-plus advertising agencies owned by MDC Partners.

I had the chance to speak with Norton recently to discuss how he develops excellent customer analytics practices that uncover insights other companies usually miss. Some of the key advice that Norton shared:

  • Create simple customer segments first by mining data;
  • Prioritize data collection projects to build momentum;
  • Abandon intuition. Test offers to optimize marketing campaigns; and
  • Ensure marketing analytics is embedded with decision makers.

In this article, I'll share details on Norton's insights.

Create Segments Around What Customers Do, Not What They Say

First things first: to improve marketing initiatives, organizations need to understand who their customers are and what they want. While focus groups and market research can be a good start, Norton says the best source of insight doesn't come straight from the customer's mouth–it comes from looking at existing customer transaction data.

“You can do other research, but to some extent, what people do and what they say are different,” he explains. “You need to mine the transactional data and look to see what your groups do differently.”

So trust the data, not the customer. But it's important to keep segmentation and analysis organized–driven by simple hypotheses and straightforward data–before exploring new customer segments hidden within Big Data. Otherwise, companies can lose focus.

“The mistake I see is when companies leap straight to Big Data, but they don't have a basic view of their customer,” says Norton. “They should start at a more basic level by understanding performance by key customer segments before getting more complicated with a broader data set.”

Prioritize Data by Expected Value and Complexity

Customer analytics are only as good as the data the company collects and the resources it has to learn from that data.  Many organizations still struggle with understanding their customer data, due to the data being disorganized, outdated or impossible to access and integrate with other data sources. Norton explains that, while it can sometimes take years to create the perfect data management system, using an iterative process that gets smaller portions of data to marketing sooner rather than later is a company's best bet.

“To me, it's about finding the right data and getting it in an environment that marketing and operations can use it now,” says Norton. “If it's a two, three-year project, you'll lose momentum, people will change jobs. You'll miss the opportunity.”

To determine priorities, Norton uses a low/medium/high bucketing system that assesses the expected value and relative complexity of getting data to a useable state. By starting with the high-value, low-complexity initiatives, the company has a greater chance of impacting the business quickly. A visualization of this hierarchy is displayed below:

David Norton's Priority Heatmap

This process builds momentum for the initiative. And importantly, it helps organizations paint a clearer picture of how customers act today–so the company can influence this behavior to their advantage in the future.

Abandon Intuition and Learn to Love the Test

Norton is a big advocate of creating customer segments, scientifically testing offers and rewards for these segments, and measuring results to find which initiatives drive the greatest satisfaction, loyalty, and eventually, revenue. By learning from and refining these tests, marketers are able to determine which incentives, messages and interactions are best for different segments of customers. And segment marketing, rather than personalized marketing, is often the most effective method to improve marketing initiatives.

“One-to-one marketing is unrealistic in many instances because of the executional burden to customize and refine, but creating a segmentation scheme that is actionable can significantly increase marketing efficiency,” says Norton.

Testing different offers and measuring their impact on customer behavior is essential to refining segments, further improving offers and increasing customer loyalty. I asked Norton how he incites his teams to embrace the scientific model, and he shared two key tenets:

  1. You have to build a true “analytical culture.” Executives must habitually answer questions scientifically and ask their employees to do the same. Norton advises executives to ask questions that inquire analysis, which forces the team to adopt a decision-making approach based around analytics.
  2. You have to settle professional disagreements through analysis. Gut intuition doesn't have a place in customer analytics–unless you have the data to back it up. In our conversation, Norton referenced a direct-response television campaign Caesars tested after the economic downturn. The campaign hoped TV ads would fill hotel rooms in its top-of-the-line properties. There was internal debate whether the campaign would work, and unfortunately, it did turn out to be unsuccessful: while the ads did generate a response, most of the respondents were only looking for rooms with low rates at non-premium properties.

Position Marketing Analysts at Ground-Level

At Caesars, Norton and his team spearheaded a major technology project to organize all of its data across all properties into a single database, creating a central view of every customer throughout over 50 casino properties. But while Caesars centralized its customer data, it distributed its marketing analytics function geographically so they could be near those in charge of marketing decisions throughout all of the company's properties.

The marketing analytics function at Caesars was built around a team of MAMs, or marketing analysis managers. To Norton, embedding these analysts within operations was crucial, as it meant these employees were already familiar with key market and business issues before diving into the data. “Being close to those they serve helps facilitate turning insights into action quickly,” says Norton.

When hiring these MAMs, Norton's ideal candidate looked like this:

  • Analytics-focused over management-focused. The ideal candidate was more likely a graduate of a research institution such as Carnegie Mellon or MIT than a great business program at Stanford or Harvard.
  • Driven to communicate and partner. The candidate must prove capable of not only analyzing data, but articulating the best recommendations for decision makers
  • Data-expertise over domain-expertise. Analytical skills trumped experience within the organization or within the gaming industry, in general. Candidates often came from consulting firms, where professionals obtained expertise analyzing data and advising clients in a wide range of industries.

“I think we created a world-class marketing analysis function at Caesars, which I am looking to replicate at MDC Partners,” says Norton. Only in his position for a few months, Norton has already begun to replicate the success of his MAM program, placing analysts in regions close to the agencies MDC Partners manages.

“So technically while they may report to me, they sit in those agencies, they build the relationships and become a true partner. I think there's going to be a ton of value, and I've already received positive feedback from the agencies that we've done this with so far,” says Norton.

Share this post:  
Michael Koploy

About the Author

Since joining Software Advice in 2011, Michael Koploy's work has been cited in a variety of online publications, including ReadWrite, O'Reilly, The New York Times and SYS-CON. Michael manages content related to the Business Intelligence (BI) market for Software Advice, writing on BI tools and applications, (big) data-related news and industry trends.

Connect with Michael Koploy via: 
Email  | Google+