Enter password to view case study
Brand: Paycom Payroll LLC
Creating a 24/7 navigation & help Assistant with a seamless self-service solution.
Creating a 24/7 navigation & help Assistant with a seamless self-service solution.
Creating a 24/7 navigation & help Assistant with a seamless self-service solution.
Generative AI
LLM Implementation
UX Research & Ideation
User Interface Design
Project Summary
Paycom's internal CRM is incredibly complex and hard to navigate without years of training. We need a tool that can allows specialists to quickly find data or information of a client or system.
Team
Kale Atterberry (Lead Design)
Logan Blackburn (Lead PM)
Karthik Abominu (Lead Dev)
Kale Atterberry
Logan Blackburn
Karthik Abominu
Personal Contribution
User Research
Concept Development
Product Growth Strategy
Prototyping
Solution Delivery
Project Brief
Internal employees across TSR, CRR, PSD, and Sales need a tool to quickly understand how to complete a task, surface data points, and status of multiple clients from anywhere in the system.
Our CRM is incredibly complex with many branching pages and sub-experiences. Users need a way to answer questions, internal Specialist onboarding, and on-the-fly client information no matter what page they are on.
Project Brief
Internal employees across TSR, CRR, PSD, and Sales need a tool to quickly understand how to complete a task, surface data points, and status of multiple clients from anywhere in the system.
Our CRM is incredibly complex with many branching pages and sub-experiences. Users need a way to answer questions, internal Specialist onboarding, and on-the-fly client information no matter what page they are on.
Research
We started by collecting data on our users to see where the pain points were currently by testing 100 users. This led to data showing how much time was being spent navigating within CRM or page surfing to find the right data, instead of completing tasks.
We started by collecting data on our users to see where the pain points were currently by testing 100 users. This led to data showing how much time was being spent navigating within CRM or page surfing to find the right data, instead of completing tasks.
Client Satisfaction
68% of Specialist calls are spent with the client waiting for the specialist to find the data or space the client is calling about. 17 of the 25 average minutes per call for a specialist.
10 out of 10 specialists also believe that a simple chatbot can solve most of their daily issues with the system.
Product Complexity
9 out of 10 users find Paycom CRM extremely difficult to navigate.
7 out of 10 users also find it difficult to know where/how to complete a task or where specific data can be found.
Time is Money
Users estimate an average of 7 hours a week spent on just searching within CRM to find specific information.
This equates to roughly $128,000 of wasted time per week.
Common Complaints
A whopping 72% across all internal users think that confusing infrastructure, complex workflows, and no easy search function accounts for a majority of their needs for client calls.
This tool would allow users to communicate from email or message quickly, and focus more on completing more tasks per day.
After many brainstorming sessions (and push for Paycom into the AI sector) we decided on an LLM that would aid primarily in quick data surfacing, navigation help, and status checks across our internal users.
Some of the glaring issues with navigation and organization within the site would be tackled with a different project.
Stakeholder and Executive Decisions led to a rushed timeline where features had to be streamlined to focus on what was achievable in scope for initial launch.
We mapped out focus points for initial project launch and what was achievable from dev in this timeline. The rest of the features would be rolled out after initial build was live and tested.
Research
We started by collecting data on our users to see where the pain points were currently by testing 100 users. This led to data showing how much time was being spent navigating within CRM or page surfing to find the right data, instead of completing tasks.
Client Satisfaction
68% of Specialist calls are spent with the client waiting for the specialist to find the data or space the client is calling about. 17 of the 25 average minutes per call for a specialist.
10 out of 10 specialists also believe that a simple chatbot can solve most of their daily issues with the system.
Product Complexity
9 out of 10 users find Paycom CRM extremely difficult to navigate.
7 out of 10 users also find it difficult to know where/how to complete a task or where specific data can be found.
Time is Money
Users estimate an average of 7 hours a week spent on just searching within CRM to find specific information.
This equates to roughly $128,000 of wasted time per week.
Common Complaints
A whopping 72% across all internal users think that confusing infrastructure, complex workflows, and no easy search function accounts for a majority of their needs for client calls.
This tool would allow users to communicate from email or message quickly, and focus more on completing more tasks per day.
After many brainstorming sessions (and push for Paycom into the AI sector) we decided on an LLM that would aid primarily in quick data surfacing, navigation help, and status checks across our internal users.
Some of the glaring issues with navigation and organization within the site would be tackled with a different project.
Stakeholder and Executive Decisions led to a rushed timeline where features had to be streamlined to focus on what was achievable in scope for initial launch.
We mapped out focus points for initial project launch and what was achievable from dev in this timeline. The rest of the features would be rolled out after initial build was live and tested.
Design Build
Once we decided on primary functions, we iterated until we were satisfied on the Minimal Viable Product.
Once we decided on primary functions, we iterated until we were satisfied on the Minimal Viable Product.
Slide Right to Reveal Dream State User Flow
Slide Right to Reveal Dream State User Flow
Version 4.2 Visual Build
Version 4.2 Visual Build
Version 6.1 Visual Build: Delivered
Version 6.1 Visual Build: Delivered
Dream State
Accessibility should always be a focus. I would add many features to enhance functionality and usability.
Accessibility should always be a focus. I would add many features to enhance functionality and usability.
Not only would the chatbot have enhanced functionality, but our spaces for Help Center, Tasks at a glance, and Company News would be combined into a single assistant interface.
Audio functionality, smarter prompt suggestions, and media attachment would take what was a simple LLM chatbot to a daily tool capable of completing tasks and eliminating many instances of needing expert training to utilize CRM tasks.
Not only would the chatbot have enhanced functionality, but our spaces for Help Center, Tasks at a glance, and Company News would be combined into a single assistant interface.
Audio functionality, smarter prompt suggestions, and media attachment would take what was a simple LLM chatbot to a daily tool capable of completing tasks and eliminating many instances of needing expert training to utilize CRM tasks.
Not only would the chatbot have enhanced functionality, but our spaces for Help Center, Tasks at a glance, and Company News would be combined into a single assistant interface.
Audio functionality, smarter prompt suggestions, and media attachment would take what was a simple LLM chatbot to a daily tool capable of completing tasks and eliminating many instances of needing expert training to utilize CRM tasks.
