Knowledge Management System - Pharmaceutical Industry

Challenge:

To design a disruptive experience that will support several thousand scientists with the physical process of pharmaceutical research and development by collectively learning and sharing knowledge possessed over the years. Various research data was stored in silos in several systems, emails, storages, etc. Thus, searching for essential information was a big obstruction. Due to this, many scientists kept re-inventing the wheel, which impacted the overall efficiency. Moreover, a unique challenge was the baby boomers, as all the priceless knowledge would be gone after they retired if not transferred appropriately.

Solution:

My team decided to design and develop an enterprise knowledge management platform for scientists that will host all the knowledge and documents. Specialized scientific systems would integrate into this platform to show the required information, thus making it easy to search for all enterprise knowledge. Furthermore, visualization was implemented to aid the scientists in understanding the relationship between the different knowledge and documents, especially for regulatory processes.

My Role and Deliverables:

I worked as a Lead UX Designer and Researcher. The engagement was split into 2 phases - the discovery phase followed by the build phase (agile sprints). In the discovery phase, I performed stakeholder interviews, contextual inquiries with the scientists, and a diary study in their labs. This allowed the creation of feature requirements, personas, and IA (app map). After capturing the requirements, the entire set of product features was divided into agile sprints during the build phase. The features and epics were converted into wireframe and low-fidelity prototypes for usability testing. Approved wireframes were then passed on to the visual design and development teams. I delivered end-to-end UX – from research to validating the implemented system.

Deliverables:

• Stakeholder and contextual interview read-out

• Prioritized feature list

• Personas

• Information Architecture

• Wireframes

• Invision prototype

• Usability testing and research read-out

personas

Interviews with business stakeholders encouraged me to understand the business motives and KPIs. The stakeholders sought to create a tool with problem-solving ability that would enable product development and collaboration across a broad pharmaceutical development population (scientists). Furthermore, BMS wanted to increase productivity and help the leaders to make informed decisions. The contextual inquiries and end-user shadowing supported empathizing with the users to understand their needs and frustrations. I was able to identify three types of personas as below.

• Bench Scientist

• Team Lead

• Executive/Higher Management

These personas served as tools throughout the design and development process. At every design stage, the personas resolved the pain points and frustration.

Multi-Disciplinary Collaboration for Ideas:

From the research interviews, I realized that a lot of scientific knowledge is hard to learn by sitting in a room alone. So I orchestrated collaborative workshops to learn the domain knowledge at a faster pace and at the same time capture ideas from several viewpoints. Three co-creation workshops with 12-15 participants (onsite and remote) divided into four different groups solved diverse challenges. Collective brainstorming between the design team and the BMS scientists helped to co-own the ideas and efforts. The deep knowledge of scientists guided the team to make informed decisions during the design phase. These workshops helped to capture around 70+ ideas and features categorized into five themes.

KPI workshop output:

Feature Matrix and Prioritization:

The ideas generated from the workshops were captured as features in a shared spreadsheet that allowed easy collaboration between designers, developers, and BAs and between the POs and business stakeholders. This live and dynamic document was used as one source of truth to generate a feature matrix and prioritize the requests based on the level of effort (LOE) from both the design and development teams.

As seen above, each feature was mapped to a theme and categorized into ”Must Have” and “Nice to Have”. The value and justification were noted to keep all the stakeholders and product owners on the same page. This document was quintessential during the agile sprint planning phase and creating a future product roadmap.

Identifying use cases and journeys:

The discussions identified two distinct types of users: knowledge creators and consumers. Even though, as a team, we captured 70+ features, I wanted to make sure that we were tackling the correct problems using these features. Hence, I created a user journey to embrace an actual use case and how the features would be used daily. This exercise allowed all the team members, including the stakeholders, to align on the right features. Below are a few of the use cases. (Click the below links to open in a new tab)

• Use Case 1 - Finding a specific workspace where I know it lives within the structure and adding a new KA.

• Use Case 2 - Finding and using KAs when I don’t know the names and where they are precise.

• Use Case 3 - Continuation of previous use case – I am working on the document (report) and need to find related information.

• Use Case 4 - Identifying gaps in a current project and creating required workspaces to add knowledge and solve the problem statements.

Discovering and defining use cases helped me focus and prioritize the challenges we were solving.

Task Flows:

Before diving deeper into the wireframes, I like creating task flows or "cupcake" wireframes (minimalistic visual representation) for all team members to align with the flow and solution. Mapping visual task flows also allows one to grasp the breadth and depth of the product. The visual task flows brought everyone involved with the product design and development on the same page. The task flows also assisted during feature prioritization and sprint planning.

Concept Sketches and Wireframes:

After validating the task flows, I began working on the concept sketches and low-fidelity wireframes. Wire-framing in each sprint allowed me to illustrate the early-stage ideas to the client and create low-fidelity prototypes. Failing early and often with different ideas committed me to building a robust design. The low-fidelity prototypes were used for the layout and interaction flow approval.

Once accepted, high fidelity comps were designed and replaced in the prototype that was later used for client presentations and user testing.

Below are the prototype demo videos to show a few uses cases.

Usability Testing:

I conducted 80+ usability tests in 5 different rounds, with varying levels of fidelity of prototypes and research focus. During the last round, I tested a fully implemented product with integrations. The overall attitude was positive, and all users mentioned the application had progressed in each iteration. The participants particularly appreciated the dashboard, which allowed them to find their recent projects and documents quickly. The below images show a usability testing read-out summary.

• The first round was validation testing of the product.

• The second and third rounds helped to identify problems with interactions, labels, and information being added to the system.

• The last couple of rounds helped test the product's usability and find the overall satisfaction score based on a scale of 1 to 5. The average score was 3.75, and the median score was 4.05, with a standard deviation of 0.72

The image below shows the participants' first reaction during one of the usability testing sessions.

Outcomes:

• This product was genuinely disruptive, and the first of a kind in the pharmaceutical industry, and it was recognized at various industry-specific conferences and symposiums.

• The product demanded a process change and had initial adoption challenges, but a series of training and change management assisted with better adoption. The product was rolled out globally enterprise-wide in batches to about 3000 scientists.

• A post-launch survey showed that scientists are in favor of the product as it takes less than half their time to search a document and also understand the context compared to traditional practices.

• The above images show usage analytics and growing patterns of collaboration using the product.

• The client now wants to tackle the remaining features of the future product roadmap.