Start the Conversation:
Cold Outreach Simulation
Start the Conversation:
Cold Outreach Simulation
Embedded Files
Cold Outreach Simulation
Powered by AI, step into the role of a sales rep and practice cold outreach with realistic LinkedIn leads.
This CustomGPT helps newly-hired sales professionals start better conversations as learners practice writing value-based cold emails, receive coaching feedback, and refine their approach.
Client: Signal, a fictitious SaaS company
Audience: Newly hired Business Development Representatives (BDRs) at Signal
Responsibilities: Instructional Design, eLearning Development
Tools Used: ChatGPT
The Problem
The Problem
This is an instructional design concept project for a fictitious B2B SaaS company called Signal. I imagined that Signal’s sales team is scaling rapidly, and many new hires enter the role with little experience in cold outreach. Without structured training or a safe space to practice, BDRs report low confidence, outreach anxiety, and inconsistent message quality.
Despite high outreach volume, reply rates hover around 3%, highlighting a need for stronger value propositions and more personalized messaging. Leadership needed a scalable eLearning solution they could deploy within a month, without the cost or complexity of a six-figure custom software investment.
The Solution
The Solution
To address these challenges, I designed a scalable, scenario-based simulation that allows BDRs to practice cold outreach in a low-pressure environment before reaching out to actual prospects.
Using a Custom GPT, learners engage in open-ended, chat-based outreach scenarios that mirror real sales conversations. The simulation provides personalized, coaching-style feedback focused on value articulation, personalization, and clarity—directly targeting the behaviors contributing to low reply rates. This approach gives new hires repeated opportunities to iterate, build confidence, and refine their messaging, without requiring custom software development or live facilitation.
Anticipated Outcomes
Anticipated Outcomes
Increased learner confidence in cold outreach
Stronger value-based messaging in initial emails
Improved ability to ask discovery-focused questions
Reduced anxiety around outbound communication
→
Increase overall reply rate to 5% or more within 60 days by improving value-based messaging and personalization.
My Process
My Process
Identified the Performance Gap
I focused on the real task BDRs struggle with: starting a conversation without pitching too early.Defined Success Criteria
Learners succeed by writing calm, professional, value-based outreach that uses open-ended questions and avoids premature selling.Designed Scenario Framework
Each practice round includes simulated LinkedIn profiles with contextual clues learners must use to personalize outreach.Built Coaching Feedback Loops
After each attempt, learners receive feedback from a sales coach persona, reinforcing what worked and guiding improvement when needed.Applied Learner Control
Learners can retry, move on, or follow-up—supporting autonomy and motivation.
Testing and Iteration
Testing and Iteration
Through an iterative design process supported by weekly feedback sessions, I continuously tested and refined the experience to ensure focused practice on the core learning goal: crafting response-worthy outreach.
During testing, I noticed...
inconsistent variation in generated LinkedIn profiles.
So I decided to...
Refine the prompt to ensure a more reliable experience.
Initially, the AI responded directly as a prospect, simulating objections and replies. While realistic, this distracted from the primary learning goal: writing outreach that earns a response.
To reduce cognitive load, I redesigned the feedback to focus on whether the learner’s message would likely earn a response based on predefined criteria, rather than simulating a full conversation.
If the learner struggles, AI prompts the learner to try their email again.
I gave learners the option to retry or move on. Some learners aim to refine a single email, while others prefer applying feedback across new scenarios—both paths support learning.
Results and Takeaways
Results and Takeaways
AI is most effective for learning when it supports practice and feedback, not just content delivery.
CustomGPTs are especially effective training tools when timelines are tight or budgets are limited.
With AI, learners aren’t forced into a one-size-fits-all path; they can get help when they need it and move on when they’re ready.
AI can act as both the coach and the client, enabling realistic practice with conversational decision-making.
Page updated
Google Sites
Report abuse