Cannondale Bike Finder

C A N N O N D A L E B I K E F I N D E R

Team Members:

Eve DeAngelis, Danika Landers

tl;dr

At Cannondale we have what we call “Bike Finder”. It’s essentially a product recommendation quiz that promises to help shoppers discover the right bike for them. In the summer of 2020, my colleague and I embarked upon a re-design of this tool. To design Bike Finder 2.0 we spoke with bike riders and bike shop owners, checked out competitive product quizzes, researched e-comm UX best practices, and explored the world of information architecture. Check out Bike Finder 2.0 here, and subsequent user research findings here.

The Problem

Stay-at-home orders, a decline in public-transit ridership, and the need for lockdown-compliant ways to get outside meant tons of people were exploring the world of cycling for the first time in the Spring and Summer of 2020. In fact, leisure bike sales increased by 121% in 2020 compared with 2019. This boom in the cycling industry meant that Cannondale was perfectly positioned to help usher novice riders into the world of cycling. One way of doing that was by revamping their Bike Finder tool (essentially, a product recommendation quiz).

The above screengrab is from Bike Finder1.0. Would a novice user know if a carbon frame or drop handlebars were a must-have? And for that matter, would even the most experienced cyclist know what was meant by a “confidence inspiring” bike?

Going in to this project, we already had a few ideas about what we didn’t like in Bike Finder 1.0

• Questions were a confusing mix of technical terms and marketing jargon

• Results weren’t tailored enough to make decisions at the end of the quiz

• There was no opportunity to change your mind or go back

Research

We knew we wanted to improve the tool beyond its most obvious issues. To that end, we began to wonder how this digital bike finder tool could mimic, and even improve upon the experience our user was having in a brick and mortar bike shop. What was working about that experience? What wasn’t working? We set out to learn more by speaking with bike shop owners and employees, checking in with novice riders, and combing online forums.

Here is what was working about the traditional bike shop experience:

• The bike shop employee can play a massive role in educating the novice buyer

• Bike shops often have multiple salespeople who are experts in unique areas— if the salesperson you’re speaking to doesn’t know the answer to your question, a different employee could likely step in to help!

• It can be difficult to get information about nuanced differences between models through online research alone— speaking with a bike shop employee is helpful in sorting out those differences

Here is what wasn’t working about the traditional bike shop experience:

• People who don’t fit the “typical” fit, male, cyclist image sometimes feel unwelcome in bike shops and will avoid shopping or pursuing cycling because of that

• Not all bike shops are beginner-friendly. Bike salespeople may be experts in bikes, but they’re not necessarily experts in teaching people about bikes

• In a 2019 study, 51 percent of women report that they have felt unwelcome in a shop “often” or “a handful of times,” compared to 34 percent of men

In the final stage of our research, we took a look at the competition. Which other bike brands had similar product recommendation tools? What did we like about their “Bike Finders”? What could be improved upon? We used Trek’s Bike Finder as a case study. At first glance, it seemed fairly simple. But visually mapping the quiz’s structure revealed a complex tree of information behind it. By doing this exercise, we were able to convince the business stakeholders that solution could not be as not as simple as just rewording some questions and swapping out icons.

bike finder 2.0 goals

• Use logic in the tool’s structure that mimics bike shop employee expertise by allowing for multiple, customized user journeys, and that adapts to the user’s inputs

• Use photographs instead of icons that represent a diverse group of bike riders for both greater representation and greater education

• Educate along the way with a friendly but authoritative voice

• Allow the user to go back, start over at any time, or review the answers they previously supplied

• Apply filters to their final results set in order to further tailor their results

information architecture for bike finder 2.0

After establishing some goals for Bike Finder 2.0 , we collaborated with the team to nail down the right questions to ask our users. We then created surveys to test and hone these initial questions with novice users. How we worded our questions was only part of the challenge — we needed to see how this overall architecture would be laid out. We used a tree structure to map out how these questions turned into unique paths:

Bike Results

We used the above map to create our first prototype using Figma. This prototype didn’t yield results, but was used to test the question type and structure with our users. Meanwhile, our engineering team was able to build a working prototype that combined our initial design structure with tagged bikes from our inventory in order to yield bike results. With this tool, we were able to test the logic, questions, and product tagging.

Because accurate data tagging was crucial to delivering accurate, tailored results, we created spreadsheet of editable tags that were tweaked as we further refined questions and answers using the rough working bike finder from our engineering team, and the results from our initial user testing.

UI Design

With two prototypes and a round of testing under our belts, it was time to start UI design. In this phase of the process, we focused a lot of energy on designing a filter bar that would appear alongside the user’s final results and gallery of recommended bikes. This was important to include because it allows the user to further narrow their results by selecting the price range, color, and product family that best suits their needs.

Because filtering capabilities is dependent upon abundant, well organized data, we designed two filter bar states. The future state includes categories recommended by the UX research organization, Baymard Institute. The MVP design is a reflection of what we can achieve given our current data.

Below are some of the semi-final designs for bike finder:

NEXT STEps

Now that Bike Finder 2.0 is live to the public, we will continue to tweak the questions, solicit user feedback, and refine the interface. Below is an example of a document we used to invite internal stakeholders to test Bike Finder 2.0. In addition to internal research, we conducted a user test using TryMyUI, and synthesized our findings here.

appendix

BIKE SHOP EMPLOYEE + OWNER DISCUSSION GUIDE

Internal use: How might we build a tool that helps anyone from a novice to an expert discover the Cannondale bike that’s best for their unique needs? How might we educate when necessary? How might we empower the user to be confident when purchasing their Cannondale? We’re from Cycling Sports Group and we’re trying to understand the bike buying process.

GENERAL INTRODUCTORY QUESTIONS:

○ Tell us about your bike shop.

○ What’s your favorite bike right now?

○ What’s your favorite bike brand?

○ How did you get into the bike industry?

○ Would you describe yourself as a bike expert? If no, why not?

○ How did you learn everything you know about bikes?

○ How long have you owned the shop?

○ Do you offer any demos, classes, lessons, group rides, etc?

MORE SPECIFIC QUESTIONS:

○ Who is your typical clientele?

○ Who is your ideal clientele and why? (Expert? Novice? Open to anything?

Knows exactly what they want? Unlimited budget?)

○ Do you feel you take on an educator role when making a sale?

○ Is it typical that someone will ask you to explain terms to them? Ex: “What’s a hardtail?”

TARGETED QUESTIONS:

○ What initial information do you need before helping someone with their bike buying journey?

○ What are typical questions you ask when helping a bike buyer discover which bike is right for them?

○ In your experience, what characteristics are most important to the expert rider? The novice?

○ In your experience, what’s the biggest mistake one can make when purchasing a new bike?

○ What is the most common price range buyers are interested in?

○ When people come in looking for a bike, do they generally know exactly what they want or are they open to your recommendation?

○ Do you think today’s consumers do the majority of their research and buying decisions online? Or at a bike shop?

○ Is carbon fiber vs. alloy a part of the buyer’s decision making process? Is it something you ask the rider?