Buying a diamond can often be a lifetime investment, symbolizing some of the most significant moments in your life. Blue Nile allows customers to search for diamonds and settings, down to the detail but can lead to confusion along the way. The purposes of this research was to not only understand user perceptions of what it means to purchase customized jewelry, but what type of sentiment is held throughout the build process. The end-goal, to enhance the builder experience for Blue Nile.
Assessing the competition is a great starting point, but its imperative to dig deeper. This research aimed to better understand the mental model of Blue Nile's users; from the "just looking crowd" to the "motivated to buy" customer. Creating prototypes for comparative testing helped to better understand what works best for this demographic, all while extracting the most important site features to help them complete a purchase.
The "Build-Your-Own" model applies to earrings, rings and pendants alike, with the most simple journey being earrings and pendants. Using that as a jumping off point, the prototypes were refined and used to further understand the customer's behavior, pain points in shopping the site, and what components would make the experienced more streamlined.
I often reference the automotive world (online vehicle builders) where a user can customize their desired vehicle and assess the characteristics and price. With costs so high, a customer should feel good about what they chose and how much they paid for the product. The model here for Blue Nile, instead of using cumbersome "cookie logic," the prototypes made use of a "builder drawer" and a "builder bar." Tested against each other, the results were clear--users don't want to read, they don't want to think too hard either. Building a ring, earrings, or a pendant should be an easy task, and theoretically linear, however they choose to start (starting with a diamond vs. setting). After five stages of research, results ultimately pointed to the builder bar, and again, forming that linear journey of creating a piece of jewelry that a customer would be proud to own.
Phase I, II, and III – Prototypes Build Your Own Earrings
Prototype – Drawer (Desktop and Mobile)
Prototype – Builder Bar (Desktop and Mobile)
Diamond and Setting Search – List view of all the diamonds was perceived as more difficult to choose diamonds upon initial use. Whereas the visual nature of the setting search was helpful to nearly all users.
Retrieval of Build – Nearly all users added their rings to the basket, and would expect them to remain there as they navigate the site, and come back to them for later comparison in the same session.
Compare Earrings – Users mentioned a preference for having a space/tool to build/save/compare builds.
Product Review Page – Most users noted this page should be distinct amongst the process (basket review page did not offer the detail they were seeking) and images should be larger, in general.
Landing Page – Fairly consistent preference for the landing page design in the drawer prototype vs. the builder bar landing page; noting the explicit reference to “three easy steps,” use of 1-2-3, and the illustrative quality of the page visuals for those steps.
Save for later – It was not clear to some users how they would save the build for later as it stands in the prototypes. Nearly all users were opposed to anything being automatically saved in the basket (this would need to be an opted-in experience), and some users opposed to it being automatically saved in favorites/wishlist.
Revisiting a build in-progress – It had been expressed that our users will want to be able to “play with the tool” to see how much things cost. Overall, users expected the build to disappear unless they proactively save it.
Favorites/wish list – Users liked the idea of adding things to their own curated list—to sort, build, and compare other items with them.
Prototype – Builder Bar Concept (Desktop)
Step-Oriented Way-Finding – Most users gravitated to the simplified guidance of “1-2-3”. Prominence of this placed at the top of the page was reassuring to users.
Prototype – Drawer Feature (Mobile)
Most users either completely ignored the drawer on the footer of the page during the build process or discovered it AFTER building a pair of earrings.
Users did not actively use it to navigate the build itself.
Like desktop, most users either completely ignored the drawer on the footer of the page during the build process or discovered it AFTER building a pair of earrings.
User patterns suggest and may indicate the footer of the page may not be the best placement for drawer functionality, but rather sticky placement at the top.
When the drawer is open on page-load users considered it annoying or “in the way.”
User Testing – Phase IV and V – Build Your Own Ring
Competitor Analysis Desktop / Mobile – Brilliant Earth vs James Allen
Brilliant Earth – Perceived simplicity and reasonable process/number of steps. The wayfinding on this site was perceived as more basic and easier to navigate.
James Allen – The way-finding UX to get to the builder was slightly more confusing for users, general perception that the site was less easy to use and had too many ways of explaining how to build or buy a ring.
The “Ring Studio” seemed like a good experience to some users, though the level of detail seemed to overwhelm some that preferred a more straightforward approach to building a ring.
User Starting Point – All users were sent to the respective home pages for each site, no further guidance was given for this test. Users were either starting their ring builds from the top navigation or via visual starting points from the home page. Balanced interest in starting from diamond or setting. Slightly stronger start from home page visual on the mobile experience.
Build Process – Simplicity and guidance was preferred, specifically the use of “three easy steps’ on both desktop and mobile experiences.
Overall, users preferred the way finding experience of Brilliant Earth over James Allen, though both sites had elements and/or features reviewed as positive by users.
General Knowledge and Education – Users in general, did not feel they needed to know a substantial amount to build a ring at the basic level; however, users could find in moments of confusion once advanced filters were introduced.
Comparing Two Rings – Nearly all users who created rings added them in their cart to compare builds. There was a recurring observation made, users would appreciate a comparing tool or option aside from adding the items to their cart and/or being able to save builds for later consideration.
Browsing for Diamonds – Users found list views to be more confusing / more effort overall as opposed to the thumbnail views.
Some users felt that the list view would take longer to review the PDP of the diamonds than it would seeing them on a PLP page (e.g., having to navigate back and forth). They wanted to be able to get a visual impression quickly and see details efficiently. This was especially important for mobile users with a reduced screen viewport size.
Diamond + Setting Compatibility – There were few errors in diamond and setting searches (both sites filtered to what IS compatible), there was an exception where one user did not get served any results until he made a realistic price adjustment. Users did, however, understand that starting over truly meant starting over, then began the process from scratch after creating the first ring.
Evaluation Tools – Nearly all users utilized tools to examine the diamond and/or build (e.g., 360 view, video, etc.), virtual try on, adjustment of skin tone, changing metals and evaluating the look based on personal preference or imagined preference of significant other.
Mobile vs. Desktop for Final Purchase – Two users in the study contemplated whether they would use the mobile site to purchase a ring, instead they would treat mobile as a search and consideration tool, then finalize their purchase when they have access to a desktop computer.
Competitor Analysis Desktop / Mobile – Brilliant Earth vs Blue Nile
Brilliant Earth – List view was the least appealing aspect to the site’s design, noted by roughly half of all users.
In general, Brilliant Earth was perceived to be more user friendly and visually appealing overall (partly aesthetics, partly UX).
Presented in roughly 70% of users, the visual appeal of the rings and diamonds, clarity of images, and the design variety was rated slightly higher as compared to Blue Nile.
Blue Nile – Similar to Brilliant Earth, clicking from list view was perceived as time consuming when exploring diamond details. Users expressed they would like more picture-based view of filtered results.
Top-level menu was well-rated or easily utilized by all users.
Filter explanations were extremely helpful for nearly all users, especially the the non-power user. Users were quickly able to learn what the filters do after exploring the information provided with each filter.
Maintain Top Level Navigation Design – Research suggests that way-finding on BN is currently effective.
Diamond Search Listing – May be helpful to bring forward UI components for alternate PLP view (list view vs. thumbnail).
Information / Helps – Bring forward “( i )” and/or enhance informational prompts.
Blue Nile – Similar to BE, clicking from list view was perceived as time consuming when exploring diamond details. Users expressed they would like more picture-based view of filtered results.
Top-level menu was well-rated or easily utilized by all users.
Filter explanations were extremely helpful for nearly all users, especially the the non-power user. Users were quickly able to learn what the filters do after exploring the information provided with each filter.
Suggested Refinements for Blue Nile
Build Your Own Ring – Make the 3-step process introduced early in the ring building experience (landing page), and use of illustrations for this process is desired over photographs.
Builder Bar (Desktop) – Highly recommended to implement a step-by-step assist at the top of the page, outlining the three steps to BYO.
Builder Bar (Mobile) – Highly recommended to implement a builder bar, andallow this bar to be defaulted to open on-load. It may be useful to have a sticky version of this feature, with a collapsible option.
Builder Bar (Mobile) – Highly recommended to implement a builder bar, andallow this bar to be defaulted to open on-load. It may be useful to have a sticky version of this feature, with a collapsible option.
Error Matching – Highly recommended that logic of the builder prioritizes compatible options to reduce error frequency, however, allow the user to navigate to options that require the builder to start over (e.g., chose a larger incompatible diamond, then start over with settings that are compatible with that larger selection.).
UX Handling of Errors and Alerts – Make errors, prompts and corrective actions prominent with clear explanation and instructions how the user can correct (e.g., may refer to how live sales reps guide and correct in-store customers).
Demographic Information
Trial Testing Phase I and II (10 Users) - Desktop Prototype (BYOe)
Men/Women, Age: 25-65yo, Income: $80k-150k+, Location: US / UK, Some College or Higher
User Testing Phase III (6 Users) – Mobile Prototype (BYOe)
Men/Women, Age: 26-57, Income 100-150k+, Location: US, Some College or Higher
User Testing Phase IV (20 Users) – Competitor Analysis Desktop/Mobile (BYOr)
Men/Women, Age: 25-45, Income $10—150k, Location: US, Some College or Higher
User Testing Phase V – (20 Users) – Brilliant Earth vs. Blue Nile Desktop/Mobile (BYOr)
Men/Women, Age: 25-45, Income $10—150k, Location: US, Some College or Higher