OMNI Search

How might we improve the Omni platform's search and navigation to boost efficiency, reduce frustration, and avoid overwhelming users?

UX Research

UX Strategy

Data Analysis

Memo: OMNI is a tool used by many throughout the MillerKnoll ecosystem and requires a login. Out of respect and confidentiality for the organization’s internal tools, I will be omitting screenshots and specific findings from this report, and only sharing generic findings. Any visuals/illustrations on this page will reflect the process, NOT the outcome, and may have been altered for confidentiality purposes.

OVERVIEW

Omni, a tool frequented by employees of dealers and internal MillerKnoll employees, houses many resources one may need while working with MillerKnoll products. Past research has identified finding information, and OMNI search specifically, as one of the top unmet needs of dealers.

As the team searches to update the technology contributing to these issues, the UX team is looking to inform their work with user research to identify pain points regarding the search function.

OBJECTIVE

How might we improve search functionality within the Omni platform to enhance user satisfaction and efficiency, reduce frustration with irrelevant results, and ensure quick access to updated and accurate content through effective navigation?

🔑 KEY TAKEAWAYS

Rigidity and search engines do not mix well

Users need segmentation and top-down hierarchy

For complicated tools, simple UI is best

The balance between ‘too much’ and ‘too little’ information is fragile

GOAL SETTING / TOOLS

🎯 PROJECT GOALS

The project aimed to uncover usability issues, including search expectations and pain points of current users.

Ultimately, our goal was to curate recommendations for enhancing the search experience and understand the sentiment surrounding the navigation menu.

We knew we would utilize multiple methods of research and end up with a lot of information thorughout the process. We decided to utilize Miro, Google Analytics, Excel, Baymard, and Dovetail to transform information —> insight.

MODERATED USABILITY / DATA ANALYSIS

🔬 UX RESEARCH

Moderated Usability Interviews

Talking to current users of the tool, we sought to gain an understanding of how it was being used. We presented them with a series of questions to deepen our understanding of the tool’s uses and uncover pain points. The term most frequently used to describe the Omni experience was overwhelming.

One of the first things we realized was that users had established a few routine interactions, essentially building solutions around existing problems. Identifying commonalities between these paths was crucial to understanding the users' true needs.

Users also expressed significant frustration with the search results page. We found that the issues arising from their experience were a mix of poorly designed UI and a search engine that did not perform as expected.

Benchmarking and Secondary Research

Our issues seemed to fall into one of two categories: 1) the search engine algorithm itself, and 2) how the information was being presented to the user. Baymard had much input on questions I asked, such as: Should we be presenting the total number of results on a search results page? What about time it took to search? When are filters effective, and what kind of filters are commonly used? What icons/next steps are usually available for search result tiles? How many results should a user be given? What parts of the page should be prioritized?

Data Analysis

While the tool itself wasn’t tagged for deeper analytics, I was able to use Google Analytics to identify a URL parameter that changed as users applied filters. Analyzing 3 months of usage, I was able to make recommendations to simplify and reorder the filters, creating a less overwhelming experience.

SYNTHESIS / FINDINGS

💭 INSIGHTS AND RECOMMENDATIONS

Rigidity and search engines do not mix well. Users were unable to replicate past sessions without recalling the exact phrase (including spacing and capitalization).

💡 Recommendation: Provide more flexibility in querying. Basic needs include allowing extra spaces, punctuation, and spelling errors. Additionally, creating a search engine that better understands the MillerKnoll catalog and ecosystem would be highly beneficial.

Users need segmentation and a top-down hierarchy. Having broad categories presented at the top of a search results page was more satisfying to users than only trying to provide very specific results.

💡 Recommendation: Showcase a product’s category (e.g., collection pages) as one of the top results returned, in addition to more specific results.

For complicated tools, simple UI is best. Due to its multi-use nature, the tool already has a high learning curve, and further complicating the page with non-essential embellishments added to the poor user experience.

💡 Recommendation: Remove anything that does not directly and quickly benefit a user’s search experience. This includes several of the brand tiles showcased. Additionally, have filter options closed by default until a user opts to open the drop-down.

The balance between ‘too much’ and ‘too little’ is fragile. Users appreciate exposure to information that can confirm their actions, such as the category of a result or the number of results returned. However, presenting too many details was detrimental to their experience.

💡 Recommendation: Simplify numbers on the page (e.g., 3,295 to 3.3k) and remove commonly un-referenced fields such as file size.

Visual cues and their purpose. Often, users rely on visual cues to quickly evaluate whether they are on the correct path.

💡 Recommendation: Create more descriptive thumbnails, update icons, and re-evaluate the placement and contrast of the eyebrow (which consists of the category) and the date.

Filters should be used sparingly. The original logic behind the current design was for the search engine to return all relevant results, which users would then narrow down using the filters. We found that users hardly used the filters, and even when they did, only a few were used.

💡 Recommendation: Using Google Analytics, I identified which filters users used. Removing the less frequently utilized filters and reordering them by common usage would create a simpler experience.

CONCLUSION / REFLECTION

🔮 FUTURE WORK

This research was conducted as the first step of updating the Omni tool. There are many components to the tool, however, our scope was only for the navigation menu and search results sections of the tool. From our research, we were able to give pointers for updating the search engine itself, as well as providing a review of the site’s UI. However, this update is yet to come and continues to be a work in progress.

👩🏻‍💻

Irene J. Kim

WORK

Geo-Redirect

Omni Search

Oura Mobile App Case Study

NAVIGATE