How to Conduct a Customer Behavior Analysis

This article is an excerpt from the Shortform book guide to "Hacking Growth" by Sean Ellis and Morgan Brown. Shortform has the world's best summaries and analyses of books you should be reading.

Like this article? Sign up for a free trial here.

How do you conduct a customer behavior analysis? What kind of data should you collect on your customers?

Analyzing your customers’ behavior allows you to find patterns. When you find a pattern in user behavior, you’ve found an opportunity to optimize your product. For instance, you might compare users who leave after a month to find a pattern in behaviors that occur just before a user leaves. You can then use this information to tweak your product and reduce churn.

Here’s how to track and analyze your customers’ behavior.

Analyzing Your Customers’ Behavior

The more data you have on your customers’ behavior, the better insights you can glean from it. Start tracking customer behaviors at initial contact with your marketing and through to your sales funnel (the process you use to market your product to customers). Additionally, track customers’ behaviors throughout the product: How they use it, their favorite features, how often they use it, as well as your churn rate (how many users you lose).

(Shortform note: Data gathering has been the subject of ongoing conversation around online privacy and whether companies should be able to track and monitor users. In 2018, the European Union passed the General Data Protection Regulation (GDPR), a privacy law that requires companies to inform users of how they’re collecting data. This is part of why we see so many cookie notices—“cookies” are small files that websites and advertisers attach to your online signature in order to track and measure your activity. Each time you consent to cookies, you’re on the user side of the authors’ above strategy.)

The authors recommend that you continually gather two types of data:

Quantitative: How your users are using your product. Get this from in-product tracking.
Qualitative: How your users feel about your product. Get this by talking with your users.

(Shortform note: One way to interpret your customer behavior analysis is to ask what story it tells. Qualitative data, gained through customer interviews and surveys, tell you how individual users experience your product—what emotions they feel, what interactions they have, and so on. Then, you can compare across your whole set of qualitative data to find the larger story (for instance, what emotions and interactions are most common across your user base). Quantitative data tells a different sort of story: It gives you the granular detail of how your users actually use your product or service.)

When you can see how your users use your product as well as how they feel about it, you can make educated guesses about how to improve it and increase growth. Thus, the authors explain, each growth hack combines data deep dives and customer outreach. For instance, you might analyze your data and find that 85% of users ignore a certain feature. Use this information to then reach out: Using surveys or email, ask your users why they don’t use this feature. You might learn that the feature is unintuitive or hard to access. Now, you can test improvements to your product onboarding to better teach its purpose and value, and this could improve user engagement and drive growth.

Run Effective Surveys

To run an effective survey, clarify your goals, select the right population to survey, and develop effective questions. Your goals should aim to clarify information you need about your product—for instance, whether users really have the problem you think they do, or whether they’re interested in a paid service that solves it.

Next, your population (or sample size) should be large enough to give an accurate picture of your target market, and your questions should be clear and straightforward. The following two questions illustrate a good and bad survey question:

Good: “Indicate how you feel about this statement: ‘data gathering is a reasonable practice that helps businesses develop better products and services.’ Strongly disagree, disagree, neither agree nor disagree, agree, strongly agree.”

Bad: “Do you think data gathering is unethical, or is it necessary for modern business success?”

The latter question is worse because it sets up a black-and-white situation where users can only pick one of two strong positions. Additionally, it’s really two questions packed into one: “Do you think data gathering is unethical?” and “Is data gathering necessary for modern business success?” These sorts of questions can be difficult to understand and often annoy users. In contrast, the former question gives users several options to indicate how they feel about a clear, straightforward statement.

How to Conduct a Customer Behavior Analysis