This project is a redesign of the Compose Message feature within Bullhorn’s flagship ATS (Applicant Tracking System)/CRM (Customer Relationship Management) product.
Compose Message allows users to draft and send emails to contacts and candidates who exist as records within the database. This can be used for a recruiter to forward a job to a candidate, submit a candidate to a hiring manager, or send out mass mailings about new jobs to a distribution list of potential candidates.
After ongoing user research surfaced usability issues with the Compose Message workflow, a project plan was created to tackle the redesign and implementation of this feature.
Unlike most traditional email composing tools, this feature needed to integrate into a variety of productivity workflows and allow for collection of assets from a variety of sources and records.
Compose message redesign
This project was completed over a span of 6 weeks, before being delivered to the development team for implementation
Tools: Sketch, InVision, Pendo, Survicate
Deliverables: Mockup, Interactive Prototype, Technical Specifications
Focus: Research, Design Strategy, User Journey, Wireframing, Interaction Design
USABILITY TESTING & EVALUATION OF EXISTING WORKFLOW
The usability of Compose Message became a priority after a significant number of users called out this form as a point of frustration. These complaints were captured during an ongoing ATS survey (sent out to all new users 6 weeks after their first log in). Data from these calls was compared and patterns began to emerge.
Previous failures with email delivery caused user anxiety around knowing if a message had been successfully delivered to the recipient
Inability to preview attachments causes users to send the wrong documents
Too much white space and scrolling
Users unclear which fields on the form get sent to the recipient and which fields are for internal data collection
Not enough drafting space in the CK Editor
Compose Message (email) forms are everywhere. With such a common pattern that most users are interacting with every day, it’s typically advisable to leverage common patterns over trying to invent an unfamiliar interaction.
Most common Compose Message products are more component than form. Bullhorn’s original implementation is very much a form.
While most message drafting tools are compact and only block a portion of the user’s screen, Bullhorn’s vertical form layout takes up a lot of vertical space. If a user on a 1080p screen (the most common among Bullhorn’s user base) has their Bullhorn app at full screen size, they would have to scroll over 3 full screen lengths to reach the end of the form.
While Bullhorn’s form does have a lot more actions to be taken than Gmail’s simple format, there is a lot that can be leveraged from these compact/concise designs to inform the redesign of a more complex product.
SYNTHESIZE & DEFINE
LEVERAGING quantitative data
Using Pendo, we were able to track specific events within the Compose Message forms and specifically in the CK Editor.
Actions within the CK Editor were tagged in an attempt to surface seldom used actions that could be placed behind a kebab menu to reduce clutter and allow user to complete their tasks in a streamlined fashion.
DATA USED TO:
Determine less frequently used/edited fields
Track events in the CK Editor to assess customization needs
Identify additional trends in “funneled activity”
While this type of data can be very useful, it’s important to consider all factors. Pendo tracks actions taken; Bullhorn uses the previous workflow to pre-populate form fields when possible. Just because a field is not being clicked on does not necessarily mean that the field is not important or useful to the user. It may simply be that the user is not taking action on preselected data.
Create a new workflow that separates and distinguishes delivered content from data integrity functions. Allow for user-determined actions to streamline workflow by only showing what’s needed and lining up next actions.
Give users confidence that their messages and submissions are being successfully delivered to the intended recipient.
After hearing user feedback that it was unclear what was actually being delivered to the message recipient, I decided that the workflow needed to be altered. In reviewing the current Compose Message form, I determined that all fields either affected the message itself or were related to actions/data integrity within the system. After segmenting these fields by function, I decided to group related fields together to provide the user with an experience that better matched their mental model and provided clarity to the process.
TEST & ITERATE
After creating several versions of wireframes, I presented the designs to the UX team for critique. Based on team feedback, I selected two versions and brought them to high fidelity. These two designs were then user tested with 6 users from 3 different clients.
5/6 users preferred option 1 but mentioned that they would like to have the option to switch to option 2 if they did not need to reference the record behind the slideout. From this feedback I designed an extended version of the slideout that was visually similar to the full form horizontal layout, but did not launch a new tab when opening to keep the experience consistent.
2/6 users expressed confusion over the attachments. Both options included suggested attachments that were visually the same as selected attachments. This feedback led me to redesign the suggested attachments to carry less visual weight and to add an attachment count into the field label.
6/6 users liked the email sent confirmation before completing the record actions and possible next scheduled actions.
Next Steps & Reflection
Currently these designs have been provided to the Product Manager and development team for implementation. The ability to extend to a larger horizontal view was split off for possible future feature enhancement. The ability to create message drafts (a function that already exists for a note adding feature) was also removed from the V1 implementation plan.
While in the process of creating this solution, I suggested the creation of a new component: the attachment selector, and suggested a reconfiguration of the current CK Editor. The attachment component will be the first to be implemented. The CK Editor changes are being reviewed by the development team to determine the level of effort required to implement these customizations.