Objective:
Improving user workflows by providing a better search experience and results.
My Role:
Leading the end-to-end UX process (Research & Design) and gathering user needs from stakeholder groups.
Working embedded in a cross-functional team with backend developers, Quality Assurance and Product Owner and contributing to the Design System.
Project Goal:
- Empowering other product teams to provide a great search experience to their users by creating guidelines and components adhering to best practices and resolving user pain points.
- Enabling the company goal of technical transformation by providing a search library usable in the new service architecture.
How can we create a search experience for Support Specialists that empowers them to help their customers quickly and effectively?
UX and Search best practices
To understand what constitutes a great search experience and match the mental model of our users, I researched the current UX best practices for search. I looked into the psychological theory of search, how search is presented in a digital context and how search results are presented. From these I derived design considerations and presented these findings to my development team.
Competitor Analysis
I chose the analysis criteria based on the results of our stakeholder interviews and on the UX best practices I derived through my research. These included whether competitors had a specific functionality in place and what design patterns they were using.
Usability test and System Usability Scale
To understand when in our users’ process they use a search, I investigated their behavior and collected pain points when using the current search functionality. As an additional usability indicator, I used the System Usability Scale (SUS). Comparing the SUS of the old search with the new search functionality, will give us an indicator of improvements.
10 internal users were asked to complete 3 tasks of a typical Support Specialist doing task handling. I asked the participants to think out-loud to follow their thought processes and noted down problems as they were working on the tasks.
Main findings
After summarizing the results I found :
- Search behavior can be categorized into lookup and browse (berry-picking)
- Search is always available and can be accessed quickly
- The default scope is set to ‘all’ and the user can choose whether to limit it
- Suggestions are shown while typing and may include filtering and/or autocomplete
- After results are shown, they can be narrowed down further
- Some companies have own query language (Jira, Zendesk)
Stakeholder interviews
Equipped with a wealth of knowledge, we were ready to find out from our stakeholders what they needed from a search solution. I facilitated interview sessions to gather the functional and technical requirements of mostly other product teams. We also needed to understand the users of search in these different domains, so we collected use cases of searching, what information they’re searching for and from where in the software they’re searching from.
Focusing on the main pain points of the current search experience
Federated search
“Right now if you want to search for something you must specify the object type via drop-down, but you don’t always know what type of object you’re searching for.”
Support Specialist
Usability issues
“The cursor does not automatically jump to the search bar. So you have to click manually in the line to search”
Application Manager
Better search engine
“Operators can’t remember entire ticket numbers, so operators need the option to search with wildcards, like *.”
Service Desk Lead
Problem statement and Design Principles
Support Specialists need to be able to both lookup and browse information to solve a ticket in an easy to learn search system.
Learnability
How easy is it for users to accomplish basic tasks the first time they encounter the design?
Trust
Users should only see results that they are allowed to see and results are not hidden from them.
Transparency
Presenting information in a transparent manner, including the origin of search results.
I organized ideation workshops with UX Designers across 3 countries to gather ideas from different product areas and evaluate these based on their requirements.
How might we…
…ensure that Supporters can effectively lookup information?
…support users in picking information from different sources to solve an incident?
…provide Supporters an easy way to learn how to effectively use the search system?
I facilitated 6 ideation sessions (remote and local) with UX Designers and Researchers of different product areas to gather input from their domain when searching. Sketches were organized on Miro and summarized.
Bringing it all together
Iteratively, I worked with stakeholders to find solutions to fit search needs in their domains. Bringing it all together in a first user flow to cover task handling.
After these first results, I worked together with the Design System team to contribute to the further development of the existing search components.
Contributing to the Design System
There were two search components available in the Design System, but came without a suggestions flyout and scope selection. I contributed by analyzing best practices and designing a solution adhering to the design language. Additionally, I also worked on a way for users to scope their search and provided design guidelines to separate filtering and searching.
Process:
- Competitor analysis
- Gathering inspiration
- Ideation and creating 1st iteration
- Design Reviews and iterating
- Preparing design specs for development team
Scoped search component
- Default set to ‘all’ to search through all areas
- Suggestions show top results in selected scope
- Increased efficiency: Selecting suggestion will open item
Guidelines
- Clarifying the difference between filtering and search
- Screen layouts to guide teams while implementing search result pages
First release: A new search experience for the Knowledge Base
Together with the team working on the new Knowledge Management, I provided guidelines on page layouts for initiating search, search result pages and the improved search component.
Dealing with technical dependencies
Due to technical dependencies, the search engine couldn’t be released and the front end was also still not planned on the roadmap. As a team, we decided to build a prototype to gather feedback on both the search library and the new search concept UI.
Currently in progress: Building a front end prototype with the team
Since there would be a delay in the release of the search engine (backend) and subsequently the future of the new search UIs were uncertain, we decided to build a prototype with the backend and new UIs together.
This will enable further testing the quality of search results (e.g., relevance and precision) and finetuning of the search engine. I can also gather user feedback on the new concept, with a system using the newly improved search engine and new UIs.
Final Concept
Before
- Search suggestions limited to keywords
- Users had to select a scope to search in
After
- Search suggestions show top results in the scope selected
- Increased efficiency: Selecting result opens item
Before
- Search results always shown in data table
- Filtering options hidden
- Preview panel located below results, difficult to scan results
After
- Search shown in list view for better readability for less information dense items
- Preview panel next to results
- Easily accessible and transparent filtering options
Before
- Empty state had no communication to user
- There were no next steps for the user
After
- Empty state gives user continuing pathway
- Effectiveness: User can continue their journey
Learnings
Stakeholder Management
In the beginning phase of the project, we were without a fulltime Product Owner, so I took on leading the requirements gathering from our stakeholders. I learned the importance of building strong and transparent connections with individuals invested in our project. By incorporating their feedback and involving them in the ideation phase, I fostered a collaborative atmosphere.
Workshop facilitation
Conducting ideation workshops emerged as a powerful tool for me in this project. I experimented with different ways to gather ideas and summarize results. The experience reinforced the importance of structured facilitation techniques. To further learn about these techniques I joined a workshop about workshop facilitation (no pun intended ;)) at this years Hatch Conference in Berlin. Here I met many Senior and Lead Designers that showed me what potential there was in organizing a great workshop and what I can still improve on.
Dealing with dependencies
We were aware of an upcoming technical dependency but were not able to negotiate a better result for our team. Nevertheless, this provided us with the opportunity to quickly create a prototype to showcase our work and continue developing the search experience. In the future I would welcome a more strategic approach to product management and will look for more opportunities within the product to improve the user’s experience.