Helping Lawyers Make Smarter Decisions

Case Summary

Ropes & Gray is one of the world’s oldest and most prestigious big law firms. The firm remains competitive due to its investment in technology internally developed software solutions. I led our Applications Development team in redesigning one of the firm’s most heavily used applications, Client/Matter Lookup (CML). CML was an aging platform that was limited to desktop use, and unintuitive. The redesign had two major goals: create a mobile version of the application and make the application easier to use. I accomplished these top goals by developing long-term relationships with users, using an iterative approach to releases and collaborating with cross-functional teams.

As the Senior UX Designer, I collaborated with our Lead Software Engineer, Front-End Developer and UI Designer to iterate on and test possible improvements to reduce call center volume. We each had specific responsibilities. My role was to lead the team, organize and design usability testing sessions with users to figure out why they were abandoning the app, and create a solution.

The project began as a quick update to produce a mobile version of the site for lawyers. However, after researching, I discovered an opportunity to vastly improve the product experience and make a bigger impact. Data shows that I drove the team to achieve the following results: 1) a significant decrease in new help desk tickets; 2) lawyers were able to more easily access client and matter records; and 3) power users on the business support side were able to increase efficiency and perform faster searches, due to streamlined search flows and consolidating platforms.

Problem

Needle(s) in a Haystack

Our law firm serves thousands of clients. Each client is assigned a segmented record, which we call “matters.” These matters allow law firms to cluster the work done by any lawyer on a particular issue for that specific client. It is possible for a single client to have hundreds of thousands of matters hidden on our databases, which feels like a twisting labyrinth to many lawyers in the firm.

Even simple searches were akin to looking for several needles in a haystack. Searching for a keyword in one vertical would lead to duplicate results, dead ends or related but irrelevant results in other verticals. Even if you knew for certain that the content existed in the database, it was not easy to find.

Product History

CML was an old standby product that hadn’t been updated in nearly a decade, it was desktop only, slow to load and not friendly or intuitive to new users. The search results page was a tabbed layout. Keywords would return thousands of related records Lawyers did not have the time or patience to jump from tab to tab and they were overwhelmed by search results, making most of their searches fruitless. I endeavored to create an uncluttered UI that would lead to quick decision making.

Primary Users

Design Defect

Lawyers were the focal point in the design update. Before the update, the only way to gain access to CML was to either physically be in an office or, on a firm laptop remotely connected to the database. As the major income generators for the business, they needed a mobile solution to quickly reference firm activity while in court, conducting external client meetings, etc. This redesign allowed them to be more effective at the business of lawyering, which is a huge benefit for the firm.

My Process

I strategized an action-based plan to move quickly and generate data points as soon as possible. I led my team to deliver a live data prototype for the lawyers to use after a short two-week sprint. We created the responsive prototype after a whiteboard session that I conducted with the development team, stakeholders, and project managers. Then, I gathered a focus group of influential lawyers within the firm to test the prototype for a few days.

Our Prototype

The lawyers loved the prototype. They found the simplified search to be nimble and they appreciated the minimal layout of the content. They wanted to know how soon we could roll out the update.

CML was not just used by lawyers, it was used by the business-side of the firm for many different reasons. As a result, in addition to designing CML to be mobile, we also had to figure out how CML could work for a wide range of users, with a wide range of functionality.

Identifying Other User Groups

I wanted to study who used CML the most over a 30-day period. I reached out to Desktop Engineering for a usage report, then I matched employee names to departments, grouped names by department, then searched for clues. I discovered a pattern when I noted three top groups who accessed CML on a daily basis. They used CML far more than lawyers.

Top User Groups:

1. Practice Development

2. Client Intake Client Control (CICC)

3. Legal Executive Assistants

Top Secondary User Groups:

1. Document Processing

2. Finance

3. Client Development

Getting to Know Our New Users

We could not risk only serving the needs of lawyers. As a result, I set out to gather the feedback and experiences of users among the other groups. I expanded the pool of prototype testers to include Practice Development, Client Intake/Client Control, and Legal Executive Assistants.

During these focus groups, I facilitated conversations to better understand the roles and responsibilities of the teams and how CML helps them achieve their goals. Then, I incorporated their feedback and experience into the prototype.

Eliminating Redundancy

I had an early concern about the product strategy. There were also two other internally developed platforms which contained the same content as CML, but they were designed with different end goals in mind. In order to ensure that this new product truly addressed a different use case, I spent time meticulously understanding how each product is used and how the relevant stakeholders perceive each product.

The Squeaky Wheel(s) Get the Grease

I reviewed 8 years of logged help desk tickets in order to understand the points of frustration repeatedly mentioned. Reading through the backlog validated established product requirements, but I was also looking for clues. My goal was to identify long-standing issues that had not previously been addressed.

Key Insights

CML is the Beginning, not the end of a task

Research and discovery revealed several key insights. One of the more surprising takeaways from the user interview sessions is that CML is merely a tool that powers other tools, it is not the destination. Once users find the information that they are searching for, they will begin working in another application.

Power Users Need Powerful Tools

During shadowing sessions with business support staff, I noticed a very interesting set of habits.

Lawyers would often enter a single matter number or client, and immediately find what they needed. Support staff would use multiple fields to perform a version of searching I began to call “hyper-targeting.” They needed to be as specific as possible because their searches would yield hundreds of results.

Less is Not More, More is More

When we tested our proof of concept with CICC and Practice Development, they were not happy about the changes because they use the application in a way that is different from all other groups. When search results were returned, they would scan through hundreds of records to compare several items at once. Most users were happy to find a single entry at a time, CICC and Practice Development approached results as entire sets of relevant information.

Search Only Works Well When You Know Exactly What You’re Looking For

Early in the project, I made an assumption that an iterative search model would work well for all users. I simplified the UI to a single input omni-search. This worked well for lawyers who were closer to knowing exactly which clients or matters were current for them. However, support staff are concerned with many clients and many matters at once. They may not know the specific search terms related to their work, which makes features like autocomplete nearly useless for them.

All Information is Connected, Searches are Not Always Direct

Searching for one information type will always lead to related information. Users tended to search for one known parameter, like matter number while really having an end goal of finding associates who have billed hours on that matter. Maintaining this mental model and making it more apparent that a connection between the data types was important.

Solutions 

Meeting all Groups in the Middle

Lawyers need searches to be simple up front. Business support needs a lot of control over how searches are shaped and how results are processed.

A multi-sided approach was taken to meet a diverse set of needs:

1. Multi-field search reintroduced. Client, Matter, Timekeeper and Party searches can be combined;

2. Autocomplete introduced and triggers after 3 characters are entered;

3. Tabular formatting allows for ascending or descending sort functions, making it easier to comb through large sets of results;

4. Refiners are available only after a search is performed. The result reduces the number of fields needed to begin a search; and

5. Keyword searches within search results.

Outcome

Impact

Lawyers appreciated the refined product, over 76% of attorneys have the application installed on their work-related mobile devices. Help desk tickets declined sharply after the last and final update. Employees who used CML in addition to similar internal products began to prefer the quick and responsive experience of CML. Deal structuring became easier and more accurate. Between 2018 and 2021, partner billing increased 30%.

Some Mistakes Were Made

The new CML used a cleaner, quieter UI, for users needing to perform a short search before moving on to another task. Therefore, the redesign worked well for them. Unfortunately, it was revealed after release that CML did not work well for users who nearly lived in the application on a daily basis. A minority of power users found that they were unable to work as quickly as before.

The solution, in this case, was to retain the midpoint update, which was an information-dense interface. Users who liked the older style were given a user preference to revert the UI to a hybrid card+grid model.

Cost Savings

New features were introduced to users through a time-sensitive banner above the header. This brought new updates without a need to develop a training curriculum with each update rollout.

Generating Feedback Faster

In my time with the project, I learned that a disconnect existed between users and product development efforts in the firm. I wanted to build a direct channel of communication. I introduced a prominently displayed feedback button in the header. It opened a simple form, which was used to collect user feedback, build trust, and expose obscure use cases. It was an easy feature to implement and users felt heard.

Timeline

The first iteration of CML was released at the end of one year. In its second year, we were able to utilize follow-up sessions with users in response to monitored usage and gathered feedback. Research, design, and development efforts often overlapped. By the last release, the feedback and help desk chatter had slowed to a trickle.