BUSI 331 Week 7 Web Analytics Research Paper Example

Web Analytics for a Subscription Retailer: How Chewy Could Use Its Data to Guide Decisions

Defining Web Analytics

Web analytics is the collection, measurement and analysis of data about how people reach and use a website or app, with the purpose of improving decisions. Kaushik (2010) argues that useful analytics combines what happened, from clickstream data, with why it happened, from sources such as surveys, testing and competitive data. Modern tools fall into several categories: tag-based platforms that record visits, pages and events; product analytics that follow individual users across sessions and devices; experimentation platforms for controlled tests; customer data platforms that join website behavior with purchase and service records; and competitive intelligence services that estimate competitors' traffic. Malhotra (2015) places such data within the broader research process, reminding marketers that measures are only useful when tied to a defined problem.

What Research Shows About Clickstream Data

Research on clickstream data shows that browsing behavior predicts purchases. Moe and Fader (2004) modeled shoppers' visits to an online retailer and found that a visitor's history of visits helped predict the chance that a given visit would end in a purchase, so conversion could be understood as a process that develops across visits rather than a single event. Bucklin and Sismeiro (2009) reviewed clickstream research and described how it can model which sites people visit, how they move from page to page within a site and how that movement leads to purchase. The useful question is not how many people visited, but which visits moved a shopper closer to buying.

The Company: Chewy

Chewy is an online retailer of pet food, supplies and pharmacy products. In its annual report for the fiscal year ended February 2, 2025, the company reported more than 20 million active customers, net sales of about $11.9 billion and sales from Autoship, its recurring delivery program, making up roughly 80 percent of net sales (Chewy, Inc., 2025). Because most revenue comes from customers who reorder automatically, Chewy's most important analytics question is not how to win a single sale but how to start and keep subscriptions.

Evaluating Chewy's Web Analytics

A subscription retailer needs measures across the customer's full life, not only at checkout.

Two of these measures deserve special attention. First, the Autoship start rate on first orders connects website design directly to long-term revenue; a change to the checkout page that raises it even slightly is worth more than a change that only lifts one-time sales. Second, cohort retention, the share of each month's new subscribers still active after three, six and twelve months, shows whether improvements are lasting. Following the insight from Moe and Fader (2004) that conversion develops across visits, Chewy could also track how many visits new customers make before their first Autoship order and target reminders to visitors who have browsed several times without subscribing.

Data Sources Beyond the Website

A subscription retailer also needs to connect website data with what happens after the order. Customer service contacts, delivery delays and product reviews often explain why a subscriber cancels better than any click does. Kaushik (2010) describes this as combining the what of clickstream data with the why from customer voice. For Chewy, linking cancellation reasons recorded by service agents to the browsing and ordering history of the same customers would show whether cancellations follow a late delivery, an out-of-stock item or a price increase. Each of those causes calls for a different response: logistics, inventory or pricing.

Mobile app behavior deserves separate attention. Many repeat customers manage subscriptions in an app rather than on the website, and app sessions tend to be shorter and more task-focused. Measuring how easily subscribers can change a delivery date or swap a product in the app matters, because a subscriber who cannot easily adjust an order may cancel it instead. A simple measure, the share of subscription changes completed without contacting customer service, would show whether the app is helping or hurting retention.

Turning Measures Into Decisions

Analytics only creates value when someone acts on it. Bucklin and Sismeiro (2009) note that clickstream models are most useful when they are tied to specific marketing actions, such as which visitors to target or how to arrange a site. For Chewy, that means assigning each measure in the table to an owner. Marketing would own acquisition cost by channel and decide where to shift spending each quarter. The site team would own funnel conversion and the Autoship start rate and run tests to improve them. The customer care team would own cancellation reasons and work with logistics on their causes. A monthly review that looks at these measures together, rather than in separate reports, would let managers see when a gain in one area causes a loss in another, for example when a discount raises Autoship starts but also raises early cancellations.

Testing Before Deciding

Analytics shows patterns, but controlled tests show causes. Kaushik (2010) recommends experimentation as the most reliable way to learn what improves results. For example, Chewy could randomly show half of first-time buyers a checkout page that preselects Autoship with a clear discount and the other half the current page, then compare not only subscription starts but also cancellations over the following months, to make sure the change produces lasting subscribers rather than surprised ones.

Limits and Conclusion

Web analytics has limits. Privacy rules and browser changes reduce what can be tracked across sites, and measures can mislead when they are not tied to decisions. For a subscription retailer like Chewy, the most valuable analytics follow customers from first visit to long-term relationship, focusing on the Autoship start rate, cohort retention and customer value. Used with careful testing, these measures can guide decisions about where to spend, what to change on the site and how to keep the customers who generate most of the company's sales. Above all, the company should resist the temptation to track everything; a short list of measures, each tied to a decision and an owner, is more useful than a dashboard no one acts on. For a student studying any online retailer, whether it sells pet food, clothing or electronics, the same discipline applies, and it is the habit this paper has tried to model: start with the decisions the business makes, and choose the few measures that would change them.

References

Bucklin, R. E., & Sismeiro, C. (2009). Click here for Internet insight: Advances in clickstream data analysis in marketing. Journal of Interactive Marketing, 23(1), 35-48. https://doi.org/10.1016/j.intmar.2008.10.004

Chewy, Inc. (2025). Annual report on Form 10-K for the fiscal year ended February 2, 2025.

Kaushik, A. (2010). Web analytics 2.0: The art of online accountability and science of customer centricity. Wiley.

Malhotra, N. K. (2015). Essentials of marketing research: A hands-on orientation. Pearson.

Moe, W. W., & Fader, P. S. (2004). Dynamic conversion behavior at e-commerce sites. Management Science, 50(3), 326-335. https://doi.org/10.1287/mnsc.1040.0153