AI Coaching Simulator

Pooja Junnuri

Interaction Designer
UX Researcher
UX Designer
Adobe Analytics
Dovetail
Figma
ACTO

ACTO, a B2B software company, the sales training process was becoming a bottleneck. With a growing team of 150+ sales reps and only 15 coaches ratio, manual coaching was proving inefficient and inconsistent.

Key issues:

Coaches spent an average of 15 hours per week on one-on-one training

New reps waited up to 2 weeks for their first coaching session

Inconsistent feedback led to varying levels of performance across the team

Research

We conducted a mixed-method research approach:

Quantitative data:

Analyzed 6 months of sales performance data

Surveyed 45 sales reps and all 5 coaches

Qualitative data:

Conducted in-depth interviews with 10 sales reps and 3 coaches

Observed 5 live coaching sessions

Key findings:

78% of reps felt they needed more frequent feedback

Coaches reported spending 60% of their time on repetitive training scenarios

Top-performing reps practiced their pitches 3x more than average performers

Fig:  Feature Task Flow
Fig: Feature Task Flow

User Personas

We created two primary personas

a) Sarah the Sales Rep

2 years of experience

Struggles with objection handling

Wants more immediate feedback on her performance

Fig: User Persona for Sales Representative
Fig: User Persona for Sales Representative

b) Mike the Sales Coach

10+ years of sales experience

Overwhelmed by the number of reps he needs to train

Wants to focus on high-impact, strategic coachin

Ideation and Design Process:

We used a design thinking approach, starting with a workshop that generated 50+ ideas. We narrowed these down to 3 concepts, which we developed into low-fidelity wireframes.

After internal reviews, we created an interactive prototype using Figma and conducted 3 rounds of usability testing with 5 reps and 2 coaches in each round.

Fig: User Story (Left) User Journey (Right)
Fig: User Story (Left) User Journey (Right)

Key Features of the AI Coaching Simulator

a) Personalized AI-powered sales scenarios

b) Real-time feedback on tone, pacing, and content

c) Performance analytics dashboard

d) Automated improvement suggestions

Fig: Brainstorming Layouts for conversation experience
Fig: Brainstorming Layouts for conversation experience

Final Design:

v1 Prototype

Key user flows:

Rep logs in and selects a scenario

AI generates a custom sales call simulation

Rep completes the call, receiving real-time feedback

System provides a detailed performance report

Coach reviews the report and adds personalized comments

Usability Testing Results:

We conducted usability testing with 15 sales reps and 4 coaches. Key metrics:

Task completion rate: 92%

System Usability Scale (SUS) score: 84/100

93% of users rated the AI feedback as "extremely helpful"

OUTCOME

Weekly Coaching Hours (Pre vs Post Implementation)

Our AI Coaching Simulator dramatically reduced the time coaches spend on manual training. Before implementation, coaches dedicated an average of 15 hours per week to individual coaching. After just 3 months, this dropped to 5.7 hours - a significant 62% reduction. This time savings allows coaches to focus on more strategic tasks and support a larger number of sales reps.

Rep Confidence Levels

Perhaps one of the most encouraging outcomes was the boost in sales rep confidence. Based on self-reported data, we saw a 35% increase in confidence levels among our sales team. This graph shows the baseline confidence (100%) compared to the new level (135%) after using the AI Coaching Simulator. Higher confidence often translates to better performance and job satisfaction, creating a positive cycle of improvement.

User Feedback:

Sarah (Sales Rep): "The AI simulator has been a game-changer. I can practice difficult scenarios anytime, and the instant feedback helps me improve quickly."

Mike (Sales Coach): "This tool has freed up so much of my time. Now I can focus on strategic coaching rather than repeating the same basic scenarios."

Next steps:

Expand the scenario database to cover more complex sales situations

Implement natural language processing for more nuanced feedback on conversation flow

Enhance the mobile app version for on-the-go practice

Conclusion:

By leveraging thorough research methodologies, cross-functional collaboration, and core design principles, we created a solution that addresses real user needs and business challenges. The AI Coaching Simulator, born from a deep understanding of both sales reps' and managers' pain points, has proven to be a game-changer. It delivers measurable improvements across multiple dimensions of our sales operations, from time efficiency to performance and confidence levels.

Partner With Pooja
View Services

More Projects by Pooja