Salesforce is the #1 CRM solution in the world
I got hired as a Lead Product Designer to design a chatbot to support the Sales team in improving their pre-sales process
My role includes interview top sellers from global markets to understand their pre-sales process, analysis of the collected data using a jobs to be done framework, identifying and validating the opportunities to support the pre-sales process using an agent.
The Go-to-Market team hired me to collaborate with the internal Sales Team to lead the implementation of an agent specialized in the pre-sales research process. I used my stack of tools to manage this project, providing daily status to the stakeholders, and review the updates with the Design and Sales team.
I used Lucidspark to create and direct the async interview sessions, but also the 1:1 sessions. I also used Figjam to share results and the affinity exercises from the collected data. And finally, I used Figma to create quick lo-fi wireframes and hi-fidelity prototypes. The team used Slack and Jira to manage the communication and production work.
As part of the design process, I did research and benchmarking for best practices for agentic patterns in Saas tools. Salesforce as a company is pushing for AI implementation across all its teams and processes to improve and become an AI leader in the software development industry. I got to use NotebookLM to speed up my own design process to absorb the information from previous focus group sessions that I was not part of at the beginning of the project. I got inspiration from other chatbots like Microsoft's Copilot implementation to validate best practices for interaction on flows, navigation, responses, suggested actions, and more features.
As part of my role, I grouped the collected data by region to identify similar priorities. Then I used the Jobs to be done framework to set up an affinity map to cluster the information and make a final list of JTBD in order for me to create initial dialogs to review and validate with both Sales and Development teams.
After the affinity mapping, I shared the final list of Jobs To Be Done
After the final list of JTBD, I spoke with the Sales and Development Teams to choose a couple of scenarios to create "dialogs" to validate the implementation of the agent into the pre-sale process
I had the opportunity to work with another Sr. Product Designer (Argentina) to speed up the design process, 1 Design Manager (USA) supervising the work done, 1 Seller (USA) who helped with the interviews, and 1 Product Owner (USA) who validated the expected product. We all worked on this fast-paced project together. This MVP took us 4 weeks from initial interviews to final prototypes.
I did a deck to review the design process with Development, Product, Sales, Content, Branding, Marketing, and Legal teams in order to validate the new experience chatbot. To validate the dialogs I did for the Historical Information Flow I got support from the Product Team to coordinate sessions with the Sales and Development Team to get their feedback on the agent experience using the prototypes before handing off the designs for implementation.
One of the challenges when designing an Agent is to align the Product and Cross-functional teams' expectations
The UX/UI team had a Design System in place for me to use with the Service Seller Agent project, and it made my delivery time faster than expected. Also, I was able to suggest new components for future implementations..
The new agent will be a conversational chatbot able to use sources (input) from the Salesforce platform and provide the sellers with documents (output) ready to use in their pre-sales process
