Case Study:

How Affinio is transforming the way MECCA thinks about its customers

About MECCA

Twenty years ago, a young entrepreneur launched a sole MECCA store in Australia with a specific goal in mind: curate the best-emerging makeup and skincare brands from around the world, and offer them to Australian women in a unique retail environment. The store team knew exactly what they wanted from the products they stocked, and the store philosophy spoke to customers who had similar needs. That original team took pride in knowing who their customers were and having a deep understanding of their individual needs. 

Success came quickly, and soon MECCA began opening new locations across Australia and New Zealand. Today, MECCA has over 100 locations, a large eCommerce operation, and a much-loved loyalty program to reward customers. They are the exclusive retailer in the region for many of the world’s finest beauty brands and have a reputation for helping women navigate a rich and constantly changing landscape of beauty brands.

The Challenge

As the retail footprint grew, so did the customer base and the ability to understand the full range of customer needs wasn’t as straight forward as it was when the business was young.

A key concern was that different parts of the business were unknowingly making different assumptions about the customer.  There was a potential “telephone effect” where different understandings could emerge about customers even from the same information and discussions.

MECCA’s founders wanted an evidence-based understanding of their customers. Specifically, they wanted to know:

  • Which types of customers do they need to work harder to better serve? Are some customers basic needs not being met?
  • Is the loyalty program properly rewarding different types of customers? Or, based on individual needs, could it be optimized?
  • Which types of customers is each new brand launch finding success with? How can they meet more customers’ needs through adapting new brand launches or brand partnerships?

The Solution: Clustering Customers

MECCA engaged David Boyle of Audience Strategies, a consulting company that helps companies develop strategies to grow their audiences, to build capabilities to answer these questions. David is a strong advocate for clustering, which is a form of audience segmentation that surfaces otherwise inaccessible insights from hundreds or even thousands of variables. It is uniquely powerful on data sets like transaction data, where other methods of analysis struggle to come up with actionable insights. 

Selecting the variables to cluster on is a bit of an art, though. Should they cluster by brand affinity? Frequency of store visits? Product category? Overall engagement? In the end, they opted to start by clustering on all of these variables.

MECCA created a de-identified dataset of 50,000 customers’ transactions, and fed them into Affinio’s Augmented Analytics platform, which can analyze a dataset based on thousands of attributes simultaneously, identify statistically relevant commonalities, and surface otherwise hidden distinct clusters or communities within an audience.

Within hours, Affinio identified six preliminary audience clusters with distinct and insightful characteristics.

The clusters were more than just profiles of distinct customer types, they provided a framework for further customer research and the foundations for a detailed roadmap for growth. For instance, MECCA was able to start developing plans to improve its in-store experiences to better meet each cluster’s needs.

Results

Although early in the process, the clusters created by Affinio have already accomplished a key goal for MECCA: they showed that even transaction data, without the rich attitudinal and needs-based data that comes with survey data, can create a schema for describing customers that is insightful, helps better meet the needs of different types of customers, and can be understood by everyone in the company. In other words, it could eliminate the telephone effect. It would make it easy for all employees to keep these customer clusters top-of-mind. 

More than that, these clusters will allow MECCA to transform the way it thinks about its customers. For example, rather than grouping all customers together, the brand now has the basis to understand the needs of each segment, how to talk to them, and how to deepen the customer relationship. MECCA can now see distinct groups emerging, each requiring different products and services to meet their needs. 

Further, this offers the promise of a much better understanding of the products and stores themselves. It offers the ability to understand which types of customers a new product launch or new store launch engages. It will allow MECCA to watch the evolution of a new product or store as it engages different types of customers over its lifecycle, and to adapt marketing or services to ensure they’re doing everything they can for a new product or a new store to better meet its precise customer needs.

Finally, “It’s like having an X-Ray machine for the first time,” says David. “We can now easily look inside a customer base and see all of the different groups of customers that each have different needs. We can better treat individual groups of customers with the care and attention they deserve.”

Going Forward

MECCA is experimenting to find the best solution to roll out across the entire customer data set as the initial trial was only on a subset of data. They’re exploring which additional data can provide further insight on the clusters, trying out time-series data and product-specific data in the next iteration. They aim to come up with one master segmentation and then make use of it across almost every area of their business. The insight derived from the final set of clusters will have the potential to help optimize MECCA’s loyalty program. 

Download the Case Study