AI Dashboard to Improve Sales Performance

AI Dashboard to Improve Sales Performance

AI Dashboard to Improve Sales Performance

Project Overview

Designing data visualization for a video review tool to create a performance score based on a customer's negative and positive facial expressions. The redesigned dashboard reduced search time by 60% through key metrics and saved filters.

My Role

UX Designer, UI Designer, UX Researcher

Strategies

Design Systems
Product Strategy

My Team

2 UX designers
Sr UX designer

Project Manager

Timeline

May - Aug 2023
3 months

Problem

Managers need a way to track their sales team's performance on an individual basis.

Car dealerships rely on a repeatable and efficient sales process. While managers can track performance metrics, they lack insight into how customers respond to pitches or their overall sentiment during the interaction.

Car dealerships rely on a repeatable and efficient sales process. While managers can track performance metrics, they lack insight into how customers respond to pitches or their overall sentiment during the interaction.

From a survey of 20+ sales managers, our team found:

From a survey of 20+ sales managers, our team found:

📌 Manually reviewing videos is subjective, making it difficult to provide constructive feedback.

📌 Manually reviewing videos is subjective, making it difficult to provide constructive feedback.

📌 Managers need to quickly find specific videos and see sales results at a glance.

📌 Managers need to quickly find specific videos and see sales results at a glance.

📌 Managers avoid manually reviewing sales videos and when they do, often miss key information.

📌 Managers avoid manually reviewing sales videos and when they do, often miss key information.

Approach

We optimized onboarding for scalable adoption by supporting comprehension and recall.

A tool to view recorded sales videos with emotional insights using AI.

The UX design team collaborated with software engineers to determine the types of statistics the software would utilize. Our task was to design an interface that clearly displays these performance metrics.

Reinforce key actions

This helps support comprehension and recall

Define complex terminology

Users need an accessible way to understand complex terms

Re-access onboarding at any time

A tooltip icon explaining a technical term

A mockup illustrating the camera setup within the sales office

A mockup illustrating the camera setup within the sales office

A mockup illustrating the camera setup within the sales office

Our team developed software to analyze salesroom footage and map buyers' facial emotions.

Our team developed software to analyze salesroom footage and map buyers' facial emotions.

This facial data is turned into quantifiable output that can be used for performance reviews. This data tells managers how the customer is responding to the pitch and their overall sentiment.

This facial data is turned into quantifiable output that can be used for performance reviews. This data tells managers how the customer is responding to the pitch and their overall sentiment.

A variety of facial expressions

A variety of facial expressions

A variety of facial expressions

Capturing microexpressions

Capturing microexpressions

The system categorizes each facial expression as positive, negative, or neutral.

The system categorizes each facial expression as positive, negative, or neutral.

Solution

We optimized onboarding for scalable adoption by supporting comprehension and recall.

A dashboard to see recorded sales videos with at-a-glance stats.

The sales manager can efficiently locate, review, and analyze each video to meet their specific needs.

Summary View

Summary View

Users can preview the sales recording by clicking the information icon on each recording card, eliminating the need to navigate to a new tab. Key information includes camera used, deal summary, products sold, and sentiment score. This structured breakdown ensures users can quickly find the video they need.

Users can preview the sales recording by clicking the information icon on each recording card, eliminating the need to navigate to a new tab. Key information includes camera used, deal summary, products sold, and sentiment score. This structured breakdown ensures users can quickly find the video they need.

A summary view of a selected video

Filtering

Filtering

Filters enable quick and precise access to relevant sales interactions. The user can save filters for the future.

Filters enable quick and precise access to relevant sales interactions. The user can save filters for the future.

Filtering to show videos with specific features

Transcript View

Transcript View

In the video view, managers can watch the recorded sales interaction. This view features a dashboard displaying sales information and a transcript that analyzes the customer's sentiment in response to what the salesperson says, paragraph by paragraph.

In the video view, managers can watch the recorded sales interaction. This view features a dashboard displaying sales information and a transcript that analyzes the customer's sentiment in response to what the salesperson says, paragraph by paragraph.

Video view- Transcript panel

Sentiment Score

Sentiment Score

In the Insights panel, managers can see overall performance in the AI-generated insights panel. This includes an overview of keywords used, a percentage of who talked and when, and a break down of the customer's emotions during the exchange.

In the Insights panel, managers can see overall performance in the AI-generated insights panel. This includes an overview of keywords used, a percentage of who talked and when, and a break down of the customer's emotions during the exchange.

Video view- Insights panel

Research

Research

Surveys with 4 dealership sales managers and 2 salespeople about their experiences with the current process of watching sales videos yielded 120 insights, which we plotted on a feasibility/impact matrix to find 3 key insights.

Surveys with 4 dealership sales managers and 2 salespeople about their experiences with the current process of watching sales videos yielded 120 insights, which we plotted on a feasibility/impact matrix to find 3 key insights.

Figma board with sticky notes insights

1.

1.

Provide actionable feedback on sales performance.

Provide actionable feedback on sales performance.

2.

2.

Create filtering system and summarized overviews to reduce the time spent on video reviews.

Create filtering system and summarized overviews to reduce the time spent on video reviews.

3.

3.

Provide a transcript for each video highlighting key moments to speed up the review process.

Provide a transcript for each video highlighting key moments to speed up the review process.

Based on these insights, I led a focus session with 5 key stakeholders in the project to further refine project expectations and problems to solve.

Based on these insights, I led a focus session with 5 key stakeholders in the project to further refine project expectations and problems to solve.

Zoom call with stakeholders

Key Insights

Key Insights

Glanceable Information

Managers need instant insight into which sales performed well or poorly.

Managers need instant insight into which sales performed well or poorly.

Sentiment Analysis

Managers wanted a data visualization of how customers respond to the salesperson.

Managers wanted a data visualization of how customers respond to the salesperson.

Filtering

Managers review the same reports daily and need to quickly swap filters and save them for future use.

Managers review the same reports daily and need to quickly swap filters and save them for future use.

Usability Testing

The redesigned dashboard reduced search time by 60% through key metrics and saved filters.

The redesigned dashboard reduced search time by 60% through key metrics and saved filters.

The team conducted usability tests with sales managers to evaluate task completion and gather feedback on the tool's performance.

The team conducted usability tests with sales managers to evaluate task completion and gather feedback on the tool's performance.

5/5 completed all tasks successfully

5/5 completed all tasks successfully

The task included finding and reviewing specific videos, adding a combination of filters, and finding the sentiment score.

The task included finding and reviewing specific videos, adding a combination of filters, and finding the sentiment score.

4/5 said they would use the software daily

4/5 said they would use the software daily

Managers expressed a desire to switch to this tool over their current software.

Managers expressed a desire to switch to this tool over their current software.

5/5 said the software would improve their workflow

5/5 said the software would improve their workflow

The redesigned dashboard reduced search time by 60% by showing key metrics and saved filters.

The redesigned dashboard reduced search time by 60% by showing key metrics and saved filters.

Transcript from usability test with sales manager

Feedback from My Team

Feedback from My Team

Jocelyn Martinez

User Experience Designer @ Ethos Group

"I've had the privilege of mentoring Sierra this past summer, and I'm continually impressed by her attention to detail, adaptability, and kindness.

As we worked on her team project together, I really admired how much she valued everyone's input on her team. She made space to listen for ideas from every developer and see how to make the pieces fit, while also recognizing the quick turnaround in timeline. The designers as a whole were very impressed with her leadership.

Following this, Sierra was hired on part-time to work on other upcoming projects. Sierra's commitment to personal and professional development is inspiring, and I have no doubt she's destined for great things."

"I've had the privilege of mentoring Sierra this past summer, and I'm continually impressed by her attention to detail, adaptability, and kindness.

As we worked on her team project together, I really admired how much she valued everyone's input on her team. She made space to listen for ideas from every developer and see how to make the pieces fit, while also recognizing the quick turnaround in timeline. The designers as a whole were very impressed with her leadership.

Following this, Sierra was hired on part-time to work on other upcoming projects. Sierra's commitment to personal and professional development is inspiring, and I have no doubt she's destined for great things."

Tiffany Bavier

Product Owner/User Experience Designer @ Ethos Group

"I highly recommend Sierra for her outstanding performance as a UX design intern. During her time with us, she demonstrated a keen eye for detail, a strong skill set, and passion for creating intuitive user interfaces.

Her ability to collaborate effectively with designers, project managers, and developers was impressive. Sierra consistently delivered high-quality work that showed great understanding of design principles. She is a quick learner and always willing to go the extra mile to achieve excellent results."

"I highly recommend Sierra for her outstanding performance as a UX design intern. During her time with us, she demonstrated a keen eye for detail, a strong skill set, and passion for creating intuitive user interfaces.

Her ability to collaborate effectively with designers, project managers, and developers was impressive. Sierra consistently delivered high-quality work that showed great understanding of design principles. She is a quick learner and always willing to go the extra mile to achieve excellent results."

Conclusion

Risks and critique

Risks and critique

There are potential ethical concerns about using AI to quantify salespeople's preferences. If the tool produces inaccurate reports, sales managers may give misleading feedback. To mitigate this risk, the software bookmarks key areas of the video for review so the sales manager can verify the AI's accuracy.

Additionally, the sentiment score generated by AI may not fully capture the nuances of human emotion. For example, a salesperson’s true feelings may not align with visible expressions like smiling or frowning, especially in high-stakes situations like car sales, where human interpretation is crucial.

There are potential ethical concerns about using AI to quantify salespeople's preferences. If the tool produces inaccurate reports, sales managers may give misleading feedback. To mitigate this risk, the software bookmarks key areas of the video for review so the sales manager can verify the AI's accuracy.

Additionally, the sentiment score generated by AI may not fully capture the nuances of human emotion. For example, a salesperson’s true feelings may not align with visible expressions like smiling or frowning, especially in high-stakes situations like car sales, where human interpretation is crucial.

Long-Term Vision

Long-Term Vision

In the future, I would expand the product's features to include an AI-generated recommendation report for salespeople on areas of improvement. This report would allow sales managers to review, approve, and edit the suggestions before sending them to the salesperson. This addition would provide a more structured and actionable way for sales managers to guide their teams in enhancing performance.

In the future, I would expand the product's features to include an AI-generated recommendation report for salespeople on areas of improvement. This report would allow sales managers to review, approve, and edit the suggestions before sending them to the salesperson. This addition would provide a more structured and actionable way for sales managers to guide their teams in enhancing performance.