It’s a confusing process.
As someone who manages eCommerce experiences, you have 1,000 competing priorities, and it’s nigh impossible to rank them all.
In the midst of dealing with fulfillment, checkout, paid advertising and many other priorities, how can you find time to make sure you do everything you can to get users to the products they’ll buy?
- Should you invest in machine learning and personalization? If so, how do you do it?
- Which product discovery improvements should you prioritize? Search? Browse? Recommendations?
- When do users struggle to get to the product they want? Which queries are most frustrating to your users? Where are they getting stuck? If you solved their frustrations, what would the outcome be?
- Do you even have the time, patience, and budget to test complex improvements on top of everything else?
With so many potential business objectives on your hands, you need a way to effectively discover, evaluate and rank new opportunities in product discovery without expending all of your team’s resources.
We’ve partnered with dozens of enterprise retailers over the last 5 years. Time and time again, we’ve found the best way to inform priorities is to dive into the user clickstream data — but that’s neither quick, nor easy.
That’s why we’ve developed a process that provides retailers tangible insights into their discovery gaps and opportunities — without requiring any commitment to a contract or that they devote any team resources.
How to Simplify Retail Decision Making (And Achieve Unforgettable Product Discovery Experiences)
Before we dive into the process of clickstream analysis, let’s first answer the question:
What is clickstream?
We define clickstream as data on the actions each user takes on your website — what action each user takes — whether it’s searching, browsing a category page, clicking, adding to cart and purchasing.
Clickstream analysis is the process of gathering and analyzing the clickstream data from your users to help you understand and plan critical business objectives. In other words, clickstream analysis allows you to find and plan the tasks that will result in the best outcomes.
How Clickstream Analysis Can Help You Make Decisions
At its simplest, clickstream data can provide myriad insights into your search and discovery experiences:
- Where are users getting stuck on your site? What browse pages, search result pages, etc. do users not convert on? How can these pages improve and what would be the outcome of those improvements?
- What pathways do users take to find specific products? How can those pathways be optimized to ensure users find what they’re looking for in the quickest time possible?
- Where are users seeing zero-result pages? When they do hit zero-result pages, how do they refine their searches?
- How do a user’s geographic location, time of day, etc. affect their purchasing decisions? How do these factors play into the difficulties they may experience in the above examples?
…but that’s only the starting point. Machine learning is now expanding the possibilities of clickstream analysis, allowing many companies to answer complicated questions:
- What are the best positions for any given item within any given query to maximize business KPIs that we care about?
- What are the best positions for any given item in a result set for any given user? What is the outcome of introducing personalization?
- What would be the outcome of tailoring each user’s experience across the entire product discovery experience (search, autosuggest, browse, recommendations)?
Use Case #1: Grocery Retailer
We recently worked with a popular grocery retailer who struggled to achieve their eCommerce goals, particularly in the realm of search and discovery.
After helping the grocer set up a simple system for capturing and analyzing their clickstream data (we’ll discuss this system shortly), one glaring issue was presented to us immediately:
While zero-result rates were low, frustrated search counts (i.e. searches where users bounced after searching any specific query, or reformulated their search) were high.
This was a clear sign of sub-par search results. The upside to resolving the issue was clear, so we made solving it a priority.
After analyzing the data, we were able to implement systems which used the reformulated search data (and more) to automatically correct the search results across hundreds of queries.
Use Case #2: Apparel Retailer
Bonobos, a popular apparel retailer, also recently came to us with issues surrounding many of the same concerns we mentioned previously.
After analyzing Bonobos’ clickstream data, we found that:
- Long-tail queries received lower clickthrough rates and purchase rates
- Recommendations and browse results were receiving comparatively low click through rates and average order values
- Some popular products like Weekday Warrior dress pants weren’t being discovered due to user confusion
Without conducting the proper clickstream analysis, finding and fixing these issues would most likely have never happened and Bonobos would have never realized the benefits of resolving them.
The challenge of collecting and analyzing clickstream data
Many companies don’t know exactly what data they have on their users — and even if they do, they struggle to make the sort of data-driven prioritization decisions to optimize search and browse.
Some log and store user data into database cold storage in batches (not in real-time) for economic reasons, which means their data is out of touch with what users are actually doing on the site. They aren’t in a place where they can effectively use their data to improve search and discovery, and this causes them to incorrectly evaluate the outcomes of search improvements (or just ignore the idea of improving search altogether).
On top of the difficulties of collecting and using clickstream data, there are many questions left unanswered:
- “How highly should I prioritize search and discovery improvements?”
- “Are my search & browse pages doing all they can to drive conversions?”
- “Is machine learning applicable with my catalog size and monthly visitors?”
- “Is now really the best time to experiment with more advanced search and discovery systems?”
- “Do I even have the time, patience, and budget to test complex search improvements on top of everything else?”
That’s why we developed the Constructor Beacon.
Constructor Beacon
The beacon is a two-line snippet of JavaScript that takes your clickstream data, feeds it into Constructor’s machine learning models, and presents you with a panoramic overview of improvements you’d see by implementing machine learning in your search — as well as the low-hanging value you can capture today.
The Beacon also gives you insights into your user behavior patterns:
- Where are users getting stuck on your site?
- What pathways do users take to find specific products?
- What are patterns of product attribute affinities across search results, browse pages, and in terms of geography and interest graph?
- Across all touchpoints, where is product ranking upside-down? Meaning, where do users interest and your ranking diverge?
- Where are users seeing zero-result or frustrating result pages — and when users hit those pages, how do they refine their searches successfully or unsuccessfully?
The process for installing the Beacon is simple, and can be completed in the span of 2 to 5 minutes as shown in the video below:
There are no front-end or back-end changes that take place after the Beacon is installed — our systems will take care of the rest. Whether you’re making changes to your back-end, re-platforming, or anything else, Beacon will stay up-and-running without interfering with other important tasks.
……
Now more than ever, we believe in providing true value and insights with the Beacon. Improving your search and discovery experiences shouldn’t cause confusion — we all have enough of that on our hands already.
Interested in installing the Beacon on your retail site? Shoot us an email today.
