Without knowing your customer, you won’t successfully emerge from the recession. Here are some simple and effective ways to optimize your customer analysis.

In June, Starbucks announced a big change to their business model in response to COVID-19 that included shutting down hundreds of traditional café locations in order to open (or repurpose) pickup-only stores in their place. According to Starbucks, even prior to COVID-19, data from its mobile app and store sales showed that 80% of their transactions were for “on-the-go” orders. This insight put into motion a new strategy that better caters to the majority of their customer base who has a preference for convenience and to-go orders. Then came COVID-19 which prompted Starbucks to accelerate the rollout of Starbucks Pickup, knowing their market would be even more receptive to it in the middle of a pandemic.

There are countless stories like the one of Starbucks where companies are using their data to uncover what’s in front of them—customers’ preferences and changing buying behaviors—to skillfully adjust their strategies in the midst of COVID-19.

This ability to use data to respond to customer desires will be the defining characteristic that separates companies who successfully emerge from this crisis and those who don’t.

How to Use Your Data to Understand and Respond to Your Customers

Companies should be able to answer basic questions such as:

  • How have customers buying habits and behaviors changed?
  • What products and services do they want or need?
  • How do they want these products and services delivered?
  • How are my competitors servicing customers? How can we do it better?

If you can’t answer these questions about your customers — and do so quickly — you need to fix it now. Here are some recommendations on how to get started with analyzing customer behavior.

Source and Gather Your Customer Data

Fortunately, most companies already have a wealth of customer data needed to perform meaningful customer analytics. Customer data can be found in a variety of different source systems throughout your business. Data from CRM systems, ERP systems, Google Analytics, social media, surveys, customer service tools, and loyalty program data (just to name a few) all provide useful information about your customers—who they are, what and when they buy, and how they interact with your business.  In order to glean these insights, you must integrate your data — a task that is often easier said than done.  How you integrate your data depends on your level of analytics maturity and the BI tools you utilize. While Excel can be a great tool (and better than nothing), we always recommend a central data repository and a modern analytics tool.

Look to External Sources to Complete the Picture

In addition to internally-generated and captured customer data, data sources from outside your organization can help build a more comprehensive picture of your customers. Demographic data, market data, macroeconomic data, healthcare data, and location data (such as foot traffic data, driving traffic data, public transportation data, and airline usage data) are examples of sources that can be used to augment your internal customer data.

Keep in mind that while third party data may be free or inexpensive to acquire, integration with your existing datasets can be a challenge due to format, poor documentation, and naming conventions. The good news is that most of the major data platforms (like AWS, Azure, GCP, and Snowflake) now offer “data marketplaces” where you can easily access and procure well-formed datasets curated by third parties.

Use Machine Learning to Perform Customer Segmentation

Customer Segmentation is the process of dividing customers into similar groups or segments.  Customers within a segment share similar characteristics, and thus their buying/affinity/retention behaviors and pain points are likely to be similar.  We’ve known for a long while that taking a “one size fits all” approach to marketing and selling to customers doesn’t work across the entirety of your target market, but often one size is appropriate within a customer segment.

Companies have hundreds of data points about each and every customer, so manually trying to classify customers into segments can be difficult-to-impossible.  This is why machine learning can be a very effective – and sometimes the only viable – approach to customer segmentation.  Unlike traditional, manual methods of grouping customers based on a handful of characteristics, machine learning looks at the entire data set to identify characteristics and patterns that are not visible to the naked eye.

There are a wide variety of machine learning techniques and clustering algorithms available — K-means, Hierarchical Clustering, Affinity Propagation, Agglomerative Clustering, and other density-based algorithms (e.g. DBSCAN or HDBSCAN), to name a few.  Each approach has its own strengths and weaknesses that should be evaluated to determine which method is right to explore and segment your [prospective] customer base.  A data science expert can help you determine which machine learning approach is best for your use case and will bring you value quickly.

Read our blog: Predict and Control Customer Churn with Machine Learning

Revisit Your Data Strategy

The most important aspect of your recovery efforts is your Data Strategy. Customer behaviors have never been static, but in the face of COVID-19, customer behaviors are more fluid than ever.  Agile organizations that can quickly respond to customers’ changing needs and behaviors will not only survive, but also have what they need to thrive. A Data Strategy addresses more than the data; it accounts for the people, processes, and technology so that everyone across your organization is empowered to use high quality data for decision making. Answering the immediate, mission critical questions stated above about your customers is necessary, but establishing a data culture and infrastructure that takes a long-term approach is critical to your future success.

Customer success story: Captive Confidently Faces Major Growth with Data Strategy

While we cannot predict the future, if your company is agile and has the data and analytics practices in place to predict and respond to shifting customer behaviors, it’s possible to emerge from this crisis stronger.

Josh Goldner Josh is Analytic8’s Google Practice Director and is also a certified LookML Developer. Josh implements modern analytics solutions to help his clients get more value from their data. Josh is an avid outdoorsman and balances his professional work with hunting and teaching his coworkers how to fish.
Nate Sather Nate is a technology professional with data analytics experience in retail, manufacturing, health care, and life sciences. He is a creative problem solver focused on implementing technical solutions that drive real business value in changing markets. In his free time, Nate likes to ride his bicycle around the Twin Cities and local mountain bike trails.
Subscribe to

The Insider

Sign up to receive our monthly newsletter, and get the latest insights, tips, and advice.

Thank You!