Using Customer Insights to Improve Processes

A good place to start is creating lifestyle portraits of 4 to 7 customer types. It can be easily achieved with Microsoft Dynamics 365 Customer Insights. Using a combination of analytics and buyer and store manager knowledge, 4 to 7 customer lifestyles will typically represent 85% or so of the company’s customer base. The remaining 15% will be random shoppers or cherry pickers who do not justify the cost of the effort to treat them in a more specific way.

Analysing customer and basket data will give a first estimate of which lifestyle a customer belongs to. As an example, basket analysis at an individual consumer level will tell you a lot about a customer’s age range, whether they have a family, whether they have pets, whether they are upmarket or middle market, do they do their own home improvements, etc.

Store observations and data capture at the POS device can supplement it. This latter step has to be done unobtrusively by trained store staff at the registers reviewing or entering a lifestyle code for each shopper they serve without making it obvious. This should be an occasional exercise rather than a daily routine.

A good move is to tag each customer in the CRM system with the lifestyle profile that best matches them (including none for the bottom 15%).

Here’s an example of a home improvement retailer. The photograph reminds us of the lifestyle portrait for each customer type. The percentage in brackets is the estimated share of sales each lifestyle accounts for.

Each photograph is supported by a detailed written lifestyle profile of that customer type. In this case, these profiles account for an estimated 83% to 85% of sales. These profiles are specific to the country and the retailer’s product range.

This analysis will not be perfect, but it will be a lot more helpful than thinking about an average customer and accuracy will get better over time as more data and better analytic tools become available. As AI research progresses, AI tools will be able to finesse this type of analysis and lifestyle attribution more accurately.

The chart above shows an example decile analysis of your customer base, in this case for a supermarket. In the decile analysis break each of the top 7 deciles out by what percent is in each lifestyle using the individual tags referred to earlier. The paragraphs below discuss several ways to improve business performance using this data.

Different lifestyles often value different aspects of customer service:

• Some value expert assistance in selecting product choices, and some like to be left alone to make their choices
• Some like self-checkout, where queues may be shorter, and some will always want to be checked out by a person
• Some prefer home delivery, others click and collect

Knowledge of customers' lifestyles and the share of a store’s sales accounted for by each of them will help optimise staffing levels, skill mixes in stores, and the number of self-checkout terminals each store needs, plus a host of other customer service dimensions.

This knowledge should guide even low-level decisions like what background music to play in each store.

Stores that sell everything for the new mother or the new mother to be, are more successful when the sales associates have a good proportion of current mothers among them. For new mothers the cleanliness of bathrooms and the availability of baby changing facilities are important.

Stores that sell home improvement items are another example. In this case, if the customer is a tradesperson like Professional Peter above, they mostly don’t want a store associate telling them how to do a job. A home-owner though like Daiva may be very grateful for the advice. This is why store staff needs to understand the customer lifestyles the store supports, so they can behave accordingly. This is a different and extremely valuable dimension of personalisation.

This type of thought process applies equally online. When a customer logs in to the online store, the pages they see first should be biased to their lifestyle.

Adopting initiatives like these means being able to capture the right data and having the analytics tools and people skills to get the most from them.

Retailers with online stores are now building much bigger databases than they had before online retailing emerged. Couple this development with technical approaches that allow customer data to be aggregated across channels provides significant opportunities to grow sales and profits. We have given a number of real-world examples of how to exploit these possibilities above.

The key steps to be able to do this include:

• The data necessary to support these innovations can and needs to be captured. It must be done at the right levels of granularity and it must have the underlying data warehouse technology to hold it all and support the many kinds of analysis to evaluate and exploit the possibilities.
• The core systems of the business must capture this data and feed the data warehouse.
• A variety of analytics tools are needed to provide analysis and modelling capabilities.
• AI is becoming more widely adopted and its value to retailers will grow substantially over the next 10 years as more and more applications become available to exploit these possibilities.
• Retailers will need their own analytics teams or very good support from industry third parties like HSO in order to invest wisely and achieve the benefits that can be realised.