Data Management
SaaS Web
B2B
Our web application helps owners manage massive amounts of property data.
A typical client has hundreds of sites; within each, there are at least 10 buildings scattered everywhere. Everything related to the property needs to be evaluated and recorded in the system, from the architectural structure to a light bulb.
The geolocation data of properties is a crucial part of the assessment process, and the Map Module is introduced as an essential feature to record and organize it.
However, the geolocation data was found inaccurate, and it is affecting the business delivery.
The goal is to identify the cause of the problematic data, and revamp the Map Module feature to reduce errors.
Owners need to monitor their buildings' current condition and be prepared for future maintenance plans with cost estimation. They hire a professional assessment team to evaluate onsite and then record, organize and calculate the assessment data.
Usually, there are thousands of buildings scattered in hundred sites.
Map Module is an important feature to support the assessment workflow because:
Here is the current map module. The geolocation data in the Map Module were found:
The data is only valuable when it is accurate
Clients count on the data we bring to line up their maintenance and budgeting plan. The problematic data can lead to more errors in following steps and flawed deliverables. Besides, the extra time and efforts consumed on QA/QC.
The process of evaluating assets require users to go onsite and exam the buildings system by system, floor by floor, room by room. Any deficiencies spotted need to be marked down according to their location, as the maintenance plan is based on grouping the adjacent and relevant items to be fixed together.
Previously, only building’s outlines were marked in the map, but other data such as detailed floor plans and deficiencies notes were saved in PDF or text format.
We assume the comprehensive dataset could reduce the errors because:
INITIAL DESIGN TASK
Enhance the functionality to include comprehensive data
Through research, it became evident that the data chain's incompletion is not the only cause of problematic data. I found the problem was occurred in the end-user's workflow, unidentified errors rolling into bigger mistakes in the consecutive assessment process.
In the end, I focused on improving usability to boost productivity and efficiency for users. It will help them to reduce time on tasks and possible errors.
MAIN DESIGN OBJECTIVES I FOCUSED ON
Design is approved and under internal testing along with other system updates. The current results of testing are:
* Creating asset's profile time increased due to more data involved. But the benefit of it can be seen from the rest workflow.
TIME ON FIRST STAGE WORK
~35% more
ERROR RATE ON SAME SCENARIO
~30-50% less
TIME ON SAME TASKS
~40-60% less
As my task also included rearranging the backend data structure, it is critical to comprehend the domain knowledge and workflow before designing.
I asked myself some questions to begin with and mapped out the initial findings in the framework:
What data should we add?
How does it work?
The initial findings validated that adding more data might be useful in reducing the errors. It also became clear that mistakes might slip through the cracks when teams are processing data in the consecutive workflow.
To find out where the root cause is, I pushed for an opportunity to conduct interviews with multi-type core users and here are some key findings:
LINKED Data Content
Nesting structure can help users to identify and locate related data easier.
However, it was not reflected in the old design.
CONSECUTIVE Workflow
The workflow of assessing data follows a strict sequence, and each team relies on the previous team's result to move forward. One mistake will lead to the following problems.
So, where did the errors start?
The challenge here is that we are designing for over 10 roles, and each of them works with the Map Module differently. To maximize the design effectiveness, I need to find the core users who impact the process mostly.
I categorized the original workflow into three distinct types based on the use pattern and context. It allowed me to translate the diverse stages of work and map it into three key use cases and quickly decide on the interviewed users.
Previously, the insights were treated by individuals without putting the entire picture of the team's work together. After interviewing the users who represent the entire asset assessment's key process, I developed a storyboard to visualize the entire workflow. So now, the problems are getting clear:
Unmet needs of end-users and errors occurred and accumulated through the process
Users can't complete their maps productively and consumed extra time to fix problems from earlier steps.
Time is limited, so quality and accuracy are often ignored: users rushed to finish the required task without following the regulation, leaving no time to check errors.
The system is also flawed with unregulated guidance, not assisting users in detecting and preventing errors.
TWISTING FOCUS
Only completing the data is not enough - users are the key to the problem
Redefine objective through reframed problems
We discovered three main usability issues that hindered the speed and process of working with data. To improve usability to boost productivity and reduce errors, more refined challenges arise:
3
Insufficient system feedback & assistance leaving users in the mist of errors
1
Incomplete Database dramatically increased the effort for evaluating assets
2
Problematic Workflow
in both individual and cross-team stages creates chain mistakes
to
How might we...
complete the data & make it easier to navigate and manage?
How might we...
optimize the process to reduce time consumed on tasks?
How might we...
better prevent and guide users out of errors?
Time reduction
Error decrease
I took part current system and audited UI elements and interaction flow. Here are the major improvements that can be made:
How to arrange geolocation info & interactive elements
Adding new data require redesigning how information to be presented and layered. To get some inspirations, I looked at three type of tools that each has a distinct aspect I could learn from:
How to manage complex data property & scalability issue?
How to simplify editing function for user’s easier adoption?
After the workshop, we finalized three main design objectives. Within each, I mapped out ideas from users' insights, lessons learned from the past, and inspirations from other tools. Then we evaluated with the product supervisor and developers based on feasibility & impacts and determined the first round of design details.
Moving forward, we decide to work on the flow for the Assessment Lead first, as Onsite Evaluators are using an iPad application, and the Project planner’s tasks are limited to review the data that has already been recorded. Furthermore, the data input is the first step of the assessment process, and the results are a key foundation for the following work.
Reorganize the backbone of data structure is by far the most important challenge as this will further determine workflow and UI elements for a happy end of the task. One key question is:
How do we organize the hierarchy to help users easily locate a subcategory and find related data?
Based on users tasks analysis and mental modal, I uncluttered the data into 5 categories and divided them into 3 tiers:
I fleshed out an overall user flow to communicate with the team about integrating the new design. Meantime, it helped me to identify the major tasks that users need to take and interface I should design for:
Layer & visibility control
Text info edit (input & review)
Vector object edit (upload & draw)
Overview of properties
I sketched out lo-fi design options for internal user testing and discussion. Here are the key design decisions based on three detail levels:
With the major tasks and features in mind, I referred back to some inspiration from other tools. From each, I extracted useful design references to help me brainstorm UI solutions.
KEY LEARNINGS
Grouping the relevance to reduce the information overload
KEY OUTCOME 1
Comprehensive & Organized Data
KEY OUTCOME 2
Streamlined & Regulated Workflow
KEY OUTCOME 3
Clear & Smart System Feedback
WE WANT USER FEEL
“I can easily locate and navigate through related data”
USERS INSIGHT RECAP
“Data are in a different format, and I can’t review them on the same page.”
Previously, users can only access the building as editable objects in Map Module. After empowering the function, users can seamlessly review site, building, floor plan with room and deficiencies detail in the interactive map.
DESIGN SOLUTION FOR ASSETS' MAIN PROFILE PAGE
Previously, users can only search through a unique index number to look for the data. Although the data has a location connection, it was not visualized in the system. I reorganized the information architecture, using the location and scale as the logic behind it. Now users can quickly locate any assets based on the location and review all the data within a preferred range.
Information Architecture Before & After
WE WANT USER FEEL
“I can finish my task with less time consuming and simplified steps.”
“I can focus on required tasks.”
USERS INSIGHT RECAP
“I need to go back and forth between pages to add data.”
“I can't start until I fix other errors.”
Previously, users need many pages jumps with inputting duplicated information to add buildings into a site. After reorganizing the workflow, I used the modal pop-up page to collect building info. Users can create multiple buildings profiles at once without going back to the main page. Besides, users can't add the building's shape before they created a corresponding profile.
Design impacts on "Adding buildigns" workflow
Flow of "Adding new buildings to a site"
With the new function of adding floor plans and detailed rooms, more work is created. From a quick prototyping testing, I found two major tasks are taking a long time and easily making mistakes: 1 Building with multi-story has a typical floor plan, and users want to duplicate them instead of recreating. 2. Users need to number a large number of rooms.
So I added duplicated selected object function and auto label function to reduce the effort.
Floating toolbars enhance editing functionality
WE WANT USER FEEL
“I know what I did and what I should do”
USERS INSIGHT RECAP
“There are so much to digest and I don't know if I missed anything”
I grouped similar into to small panel and progressively disclose users the non-primary info:
Rooms are organized under each floor in accordion
Layer panel control current layer and detailed display option
Lacking system feedback is causing extra efforts for users to avoid errors. We can't prevent operational mistakes, but we can help users realize and correct them before the following work. So I analyzed the typical errors and went one step further within feasible technical support, provided users a smarter and clear status update through distinct UI indication.
Floating toolbars enhance editing functionality
The data is complete, the workflow is smooth, and the system feedback has been improved. We found that providing everything to users is not always the best option through continuous user testing and prototyping. At the end of the day, they can only handle limited information and limited choices. But on the business level, only doing "minus" creates risk or at the cost of accurate and regulated results.
HOW MIGHT WE
Find the essential needs and make them available?
From user testing feedback, we realized because of the nature of data, the [Site] page and [Building] page has very different content and tasks operation. So I further optimized the interface to distinct the information that relevant for using context.
We realized sometimes mistake caused by freedom of choice. So I worked with the development team on a feasible reconcile of steps users can take.
Task flow of normal process & trouble shooting (click to enlarge)
What I learned:
What can be better:
