© 2020 by Ben Roach. All images and logos owned by their respective brands.

Case study 01: Sears Parts Direct

The quest for a better home page

Set the stage

It looked like this when I started...

In summer 2014, Sears hired me to redesign the searspartsdirect.com home page.


At the time, Sears Parts Direct was the nation's largest provider of parts and accessories for things like appliances and electronics, carrying over 7 million parts from more than 400 manufacturers. Annual revenue for the business vertical was around $1 billion.

I set out to learn

Was the site experience matching customer expectations?

Did the experience enable customer journeys that resulted in conversion?

What were customers actually doing on the site – what paths were they following?

Which journey were converting visitors into buyers?

My learnings
  1. Users needed help finding and locating their model number so that they could search for parts.

  2. Users were reaching 'zero results found' pages at a high rate due to not understanding search functionality.

  3. Bounce rates were very high once users reached an interior page.

  4. Browsing was encouraged from the home page, but the browse path was not optimal.

  5. The home page had an overwhelming amount of information.

  6. The home page was not surfacing DIY content that was important to the business owners.

  7. Primary reasons for visiting: 48% to purchase parts; 36% to research parts; 8% to download manuals; 2% to check order status; 6% other.

It boiled down to this

The page was not optimized to help users find the item they were looking for. This meant reduced conversion opportunities and a high bounce rate.

I set project objectives
  1. Allow users to quickly and easily find a part via search on any device.

  2. Emphasize ways a user can contact Sears for help via phone, chat or email (as this resulted in a high conversion rate).

  3. Highlight existing videos, guides and manuals to transform site from a purchase destination to a hub for DIY and repair needs.

My overall process
  1. I used simple card sorting exercises with business stakeholders to prioritize page components.

  2. I moved into sketching and whiteboarding to come up with new content layouts and ways that I could design an improved search functionality that was more intuitive for users.

  3. I looked at existing site DIY content and began brainstorming how this would fit on the page within reusable content templates.

  4. I implemented a number of research methods including: videos of users interacting with the site, taxonomy tests, A/B tests, and visual heat maps of user clicks.

  5. Throughout the process, I held multiple weekly meetings with key stakeholders to ensure alignment.

White boarding and sketching.

Component explorations

When redesigning the search functionality, I explored ways to reduce friction. I realized exposing selection options as a toggle had advantages over using a drop down menu.


This type of diagram was useful in helping convey design concepts with stakeholders.

Wireframes & early design iterations
Final design proposal












User validation

I worked with our lead researcher to use visual heat maps and determine if customers were able to execute common user scenarios.

"Recently the ice maker of your Kenmore refrigerator has been malfunctioning and you want to buy a replacement unit. You checked your refrigerator and found out that your model number is 106866830. Please click where you would go to help solve your problem."

In this scenario, almost all of our users were able to take the correct action.







Continuous optimization

I worked with engineering to implement my proposal for an ongoing page optimization plan. I identified a number of areas number for in-production ongoing A/B comparative testing and/or personalization.

  1. Headline language

  2. Value propositions

  3. Brand logos

  4. Flexible advertising space

  5. Featured video content

  6. Featured repair guides

Case study 02: Fiat Chrysler Automobiles

A modular campaign framework

Case study 03: Amazon

Reachable refinements