Understanding the existing platform
Before you fix the problem, you’ve got to understand the chaos.
I started off by looking at how Partitions worked in Workplace Analytics (WPA). This gave me a better understanding of how an administrator approached this feature and what their pain points were. A typical partition assignment included four major steps. I was to focus on the CRUD experience of the feature, keeping usability and design finesse in mind.
Hover over the images to see the heuristics.
1. Create partition
When a new analyst is joining the team, an admin wants to create a customized partition and only provide the required data to them, such that they follow the privacy standards set by their firm.


2. Select employees
To filter the right employees so that it is easy for the admin to section off to the analysts.


3. Select attributes
To select the right attributes and pick how that attribute should be shown to the analyst.


4. Select analysts
To find the right analysts and assign them the partition.


Measuring usability through SUS
Because scoring usability is like judging a dance—every step matters.
To ensure we were creating a seamless and user-friendly experience, we needed to evaluate whether users preferred a step-by-step wizard or a single-pane wizard for creating partitions. Using the System Usability Scale (SUS), we gathered structured feedback from respondents by asking them to score various aspects of their experience with both approaches.
The step-by-step wizard provided a guided, linear flow that some users found intuitive, while the single-pane wizard offered a consolidated, at-a-glance approach that catered to power users who wanted quick access to all information.
By comparing the SUS scores of both designs, we were able to identify clear user preferences, allowing us to refine the wizard experience to balance simplicity, efficiency, and usability. This data-driven approach ensured that the final design catered to a wide range of user needs without compromising on the ease of use.

Step-by-step
Average SUS score of 82.5
| Respondent | ST1 | ST2 | ST3 | ST4 | ST5 | ST6 | ST7 | ST8 | ST9 | ST10 | Odd number | Even number | SUS |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 1. Mei2020 | 3 | 1 | 5 | 1 | 5 | 1 | 4 | 1 | 5 | 4 | 17 | 17 | 85 |
| 2. ElizaMae | 4 | 2 | 4 | 2 | 4 | 2 | 4 | 1 | 2 | 1 | 13 | 17 | 75 |
| 3. schmitozen | 3 | 1 | 5 | 2 | 4 | 1 | 3 | 2 | 4 | 3 | 14 | 16 | 75 |
| 4. TurlowOne | 5 | 2 | 5 | 1 | 5 | 5 | 5 | 1 | 5 | 1 | 20 | 15 | 87.5 |
| 5. 1140 | 5 | 2 | 4 | 2 | 5 | 1 | 5 | 2 | 5 | 1 | 19 | 17 | 90 |
Single Pane
Average SUS score of 92.5
| Respondent | ST1 | ST2 | ST3 | ST4 | ST5 | ST6 | ST7 | ST8 | ST9 | ST10 | Odd number | Even number | SUS |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 1. alpinenepal | 5 | 1 | 5 | 1 | 4 | 1 | 5 | 1 | 5 | 2 | 19 | 19 | 95 |
| 2. Im0102 | 5 | 1 | 5 | 1 | 5 | 1 | 4 | 1 | 5 | 1 | 19 | 20 | 97.5 |
| 3. Odin2020 | 4 | 1 | 5 | 1 | 3 | 2 | 5 | 1 | 4 | 2 | 16 | 18 | 85 |
| 4. ricesj | 4 | 1 | 5 | 1 | 5 | 1 | 4 | 1 | 5 | 2 | 18 | 19 | 92.5 |
| 5. spunk52 | 5 | 1 | 5 | 1 | 3 | 1 | 5 | 1 | 4 | 1 | 17 | 20 | 92.5 |
Partition switcher
Ensuring a seamless way to switch while working.
When an analyst handles multiple partitions, they need an easy way to switch between them. Thus, we needed a switcher. Our surveys indicated the side navigation as the ideal location for the switcher.
A dropdown was the best option as it easily lays out multiple items in a list. Description was also added with every partition so that it's easier for the user to identify what partition it is.
However, the challenge was to differentiate partition-specific pages from global pages to avoid confusion. We had multiple iterations for this. In the end, we decided to go with a simple horizontal line between the pages to make it as easy as possible for the admin to understand without complicating the navigation.

Toggle and onboarding
Feature enablement right at the tip.
The final part of our puzzle was figuring out how to let admins enable Partitions for their tenant. The tricky part here is that once Partitions is enabled, analysts will have to be manually assigned to at least one partition to work on Viva Insights and conduct analysis. This is a shift in paradigm for the admins, as before Partitions, any analyst assigned to Viva Insights in the Microsoft Admin Center would automatically get access to all data.
Educating our admins about this change without adding too much friction (so that our admins don’t get scared of the feature and decide not to use it) was our final hurdle before GA.
The final designs
What got shipped out finally.
The explorations proved to be very crucial to the direction that we took. The final product that we shipped kept that original feel of the WPA, while building on top with newer features.
The Homepage
The homepage encompasses the list of all the partitions created by the admin. The banner takes only adequate space and leaves room for the admin to focus on the list of partitions created by them. The admin can also filter different partitions using the pills above the list, and they can also sort the list from the table. A search is also provided to the admin in case there's a huge list of partitions and they have to find the right one. If the admin wants to create a new partition, they can click on the button above the Search and get started with the Wizard.
Filtering employees
Since from the SUS testing, we found out that a single page experience is better, our Create a Parititon wizard is made to be scrollable. The filters pane now improved, by adding a nesting selector. The visibility of the measured population is also drastically improved by grouping it under the filters itself. The user can also rearrange each filter, which was previously not possible.
Assigning analysts
The last step of the wizard is to select the analysts which will have access to this partition. The pills here is different from the attribute pills to show that these are people pills. Similar to the attributes flyout, the user flyout helps the admin find the right analysts for the partition, by providing identifiable elements like avatars and email addresses.
What did I learn from this?
Reflections from my first Microsoft project
Partitions was my first design project at Microsoft, and it taught me the ropes of user research, wireframing, and prototyping. Collaborating with developers, product managers, and designers helped me grow as a team player. I tackled the challenge of creating an interface that worked for both novice and advanced users, sharpening my problem-solving skills. Iterative testing and user feedback played a key role in refining the tool, making it more intuitive and accessible. This project strengthened my design skills, but more importantly, it showed me the power of empathy and collaboration in crafting impactful solutions.