Making every marketing dollar count is every company’s dream. One way to achieve this goal is by attracting customers precisely. To get started, companies need to ask themselves four questions:
- Who are the most valuable customers?
- Where do they come from?
- What do they look like?
- What and how do they like to buy?
In Double Down on the Most Valuable Customers, I discussed how to identify the most valuable customers. In Three Approaches to Uncover User Journeys, I explored ways to analyze where customers come from. By answering questions 1 and 2, companies can quickly seize the low hanging fruit by investing more in the best-performing campaigns. However, achieving a high conversion rate is still challenging, which means each campaign is attracting less than ideal users.
Customer segmentation is a useful tool to help companies overcome the conversion rate obstacle. By looking into the profitable customers vs. the unprofitable ones, companies can find out what is attractive to each customer cohort and fine-tune their customer acquisition strategy.
What does each customer cohort look like?
To develop an acquisition plan for each customer cohort, companies must take a deep dive and analyze the cohort from as many dimensions as possible. The dimensions to look at depend on industry, business model, and product. Take consumer products, for example; here are the most common dimensions.
Demographics
Demographics, like a customer’s location, age, gender, and income, are some of the most used attributes in customer segmentation. These are often the only attributes analyzed when companies consider rudimentary customer segmentation for product packaging and pricing. However, demographic attributes only provide rough ideas about customers, which are usually insufficient for business decisions.
Touchpoint interactions
A user’s interactions with marketing campaigns and the home page provide rich information about who they are and what they like. Interactions with touchpoints include view-through, click-through, the time between each action, and more, and it takes sophisticated data infrastructure to track these events. For example, if a user sees an ad on Instagram and quickly swipes over it, that probably indicates they are not very interested in it.
Visiting behaviors
Visiting behaviors refer to what a user does on a company’s website. Compared to touchpoint interactions which are often impacted by advertisers’ programmatic push notifications, visiting behaviors are more proactive. Therefore, visiting behaviors provide enlightening ideas about what a visitor is truly looking for and how they like to shop.
Purchasing behaviors
Purchasing behaviors summarize what a user buys, their purchasing frequency and timing, purchasing price and discounts, etc. As you can imagine, purchasing behaviors indicate a user’s financial situation (whether they are a budget buyer) and more in-depth information like purchasing purposes.
Multi-layer customer segmentation
The backbone of customer segmentation is clustering algorithms. Successful customer segmentation requires data from many dimensions. However, due to technical constraints, the more dimensions there are, the more difficult it is to generate actionable segments. To address this shortfall, companies need to set a clear goal for customer segmentation before they get started with data science work so that they only bring necessary attributes into segmentation algorithms.
Companies can use a simple two-layer approach to understand what the most valuable customers look like.
- Calculate customer value using the formula discussed in Double Down on the Most Valuable Customer.
- A customer’s long-term value = Revenue from the customer’s repeated purchase — its acquisition cost
- Run the first layer segmentation with customer value
- Choose the most valuable customer segment from step 3.
- Run a second layer of customer segmentation with demographics, touchpoint interactions, visiting behaviors, and purchasing behaviors attributes.
The beauty of segmentation algorithms is that they not only generate segments automatically but also reduce the dimensions of customer attributes and synthesize information in a digestible way.
Making every dollar count with personalization
As mentioned, companies often only consider demographics when they segment customers on a high level. When these companies analyze their customers with the above approach for the first time, it’s not uncommon for them to be surprised and discover that each valuable customer segment contains subsets of multiple demographic cohorts.
With richer information, companies can now identify the most attractive campaigns, shopping experiences, and products for each customer cohort and make data-driven business decisions. Therefore, the company can achieve a higher return on ad spend with a smaller marketing budget, yielding a higher profit margin and more room for market expansion.
I discuss how to use data science to level up your business and optimize your marketing in my articles. If you want to chat about them, feel free to contact me on Linkedin.
