Intel Creation
2019 | User Research, Interaction design, Visual design, Prototyping and User testing
Overview
The goal of this project was to improve the creation workflow within EclecticIQ platform. I worked as a product designer alongside a UX architect and project manager during the discovery phase. During testing and implementation, I worked within a multi-functional team with FE and QA by providing assistance and reviewing at different stages of the project.
Project context
EclecticIQ is a desktop platform for threat analysts to collect, investigate, and disseminate intelligence. Analysts relate intel objects during case investigations, which are either ingested or manually created, leading to the creation of reports distributed to other organizations. To address negative feedback and feature requests, I was tasked with developing new functionality enabling analysts to create intel objects within the graph canvas tool. Despite an initial belief in the obviousness of the problem and solution, we advocated for a dedicated discovery process upon taking over the project to understand user problems and goals thoroughly.
If you'd prefer to skip the process and go straight to the final results, click the button below
Opportunities
Get to know and study the users and their main workflows
Lead and draw the overall discovery and design process
Opportunity to collaborate with different expertise
Challenges
No single source of truth for user feedback
Hard to get buy-in from stakeholders to run a proper discovery for this project
Lack of direction and participation from stakeholders during the discovery process
Understanding the users and their problems
Given the substantial existent user feedback on the topic, I initiated desk research to consolidate scattered information from various channels into a unified repository. This proved time-consuming due to the unstructured nature of the feedback. After analyzing the requests, I categorized them into eight groups, assigning different urgency levels: red for high, yellow for medium, and grey for low urgency. This categorization process revealed additional user problems within the creation workflow.
To delve deeper into these issues, I conducted shadowing sessions with in-house analysts and reviewed past user interviews. Extracting relevant feedback, I organized it by clustering common 'keywords' (tags) to gain a comprehensive understanding of the problems at hand.
Prioritization
To provide the product managers and engineers with visibility into the key areas of the platform that needed to be addressed from a usability standpoint I categorize the user problems into 2 big themes: create and relate and visually map them. This not only helped to prioritize usability issues in order of need but also helped to shape the product roadmap.
This made it easier to synthesize the feedback and narrow it down to a few pain points within the creation workflow:
The leght and structure of the creation forms were causing frustration to the users
In the different contexts of the platform, the creation flow was either inconsistent or missing
Creating relationships between objects could not be done from all contexts (only from the detail panes and the pattern was not very optimal)
Users were not able to complete crucial bulk operations
To illustrate and communicate to my team how the pain points affected user workflow I draw Allan journey when structuring the intelligence for a report. The journey map highlighted the problems around create and edit intelligence objects and their relationships.
Ideation
Based on the above problems identified, I worked towards addressing these pains by coming up with potential solutions based on the following principles
Consistency
Enable creation functionality from any context, to avoid context switching and make the behavior consistent across the platform
Efficiency
Enable users to create fast without being confronted with many decision points
Scalable
When possible, the solutions would have to work for multiple objects so users could perform actions over simultaneous objects
Intuitive
Hide complexity (when possible) from the users to make the experience more intuitive
Designs
I started with low-fidelity designs and after a couple of critique sessions with fellow designers and the product team I created a clickable prototype that accommodated the new features in an attempt to reach more users, I ended up creating a video, where individuals could comment and leave feedback, navigating through the different parts of the platform and illustrating new ways of completing some of the analyst's daily tasks.
The high fidelity level of the design was a conscious decision so we could tell a complete & self-explanatory story so the users didn't have to think abstractly.
Validation
Given the volume of complaints, we prioritized resolving issues in the graph canvas. Before release, we tested solutions for creating intel objects and relationships through an interactive prototype with five users. Our goals were to evaluate the discoverability and usability of the newly introduced features and understand analysts' behavior in structuring intelligence. Based on the findings following, we iterated on designs and upgraded the prototype.
Finding 1: The number on the 'unpublished' icon was not clear to everyone
Solution: Add tooltip with the summary
Finding 2: While creating relationships some of the visual elements in place were not accelerating the experience and creating some indecision: "don't see the arrow until I hover the object"
Solution: Highlight the arrow (instead of being grey) and make the arrow following the ‘user mouse". Also the object that you start creating the relationship from should be highlighted.
Finding 3: removing relationships was not always straightforward - half of the participants couldn’t understand if the arrows could be selected/not and tried to delete with back key without success
Solution: allow back on the key on the keyboard to delete. While hovering the relationship arrow add a hover effect.
Finding 4: While creating relationships between objects users don’t always consider direction. They fail to undaerstand why the system doesn’t allow to create a relationship between a report and ttp/indicator.
solution: remove the technical constraint and enable users to create relationships regardless of the direction, after clicking ‘save’ system can apply the correct direction.
Few takeaways
The most important takeaway was the importance of jumping into immediate solutions to seek to understand first and get to the root of the problems. Going beyond the initial scope, observing users, and understanding their behaviors, helped us understand some of the underlying problems around intel creation workflow which otherwise would never have been spotted.
On the first release, we launched the creation of intel in the graph which had a massive positive impact on the analyst’s workflow. The remaining ideas were added to the product roadmap in an effort of continuing to simplify analysts’ workflow inside the platform.