Smart Search

A new way for software sales teams to locate the best content within their sales enablement platform with the integration of artificial intelligence. The goal of this feature is to help return the best, most effective sales documents, case studies, and slide decks to help sales teams win sales.

Team
2 Designers
Duration
8 Weeks
Role
Researcher & Product designer
Time
Summer 2023

Discovery

Within sales teams, there are two activities that consume time, detracting from revenue-generating efforts and direct engagement with buyers: deciding what sales content is best for the scenario, and finding it.

Locate

Smart search enhances the machine's understanding of a user's query by intelligently matching it with content processed by a LLM, that has been trained on the company's extensive database of uploaded materials.

Utilize

Once a user identifies the desired content and selects the materials they wish to utilize, they can quickly distribute them to potential buyers.

Research

Context

Ai can be valuable in several aspects of a Sales Enablement Platform to increase salespeople's productivity. By integrating Ai into various aspects of a Sales Enablement Platform, salespeople can enjoy increased productivity, more effective sales strategies, and ultimately, better sales results.

Challange

How might we discover pain points and challenges within the sales process, identifying opportunities to increase productivity with the integration of Artificial Intelligence into the Bigtincan ecosystem.

Questions &Curiosities

Our People

Who are the people that work on sales teams and what are their different roles/ workflows?
What challenges do sales teams face as a group and as individuals?

Artificial Inteligence

Who are the people that work on sales teams and what are their different roles/ workflows?
What challenges to sales teams face as a group and as individuals?

The Users

Sales Agent

Selling products or services to customers, aiming to maximize revenue and build lasting client relationships.

Sales Enablement

Equipping sales agents with the necessary tools, training, and resources to enhance their effectiveness.

Marketing & Design

Collaborate to create engaging content that effectively communicates the brand's message, and attract customers.

Their Journey

Actions

Readiness
Scheduling and information intake
Preparing content for sales call
Sales engagements  proving ROI
Follow-up communication
Coaching and training

Summary

Sales enablement managers work with junior sales agents to equip them for sales engagements.
A sales lead may come in, next a sales agent will be assigned to follow up to schedule a meeting and get information about the customer.
Sales agents and sales enablement collaborate to prepare for sales engagements. Together they gather marketing content and prepare meeting agenda/ script.
Sales agents meet with buyers to understand they’re interested in and answer questions that they might have to assess/ prove product fit.
Sales agents’ communicate back and forth with buyers until the sale is closed won or lost.
Sales ennoblement managers review agents work and answer questions to help in future opportunities.

Users

Sales Enablement
Sales Agent
Sales Enablement
Sales Agent
Sales Agent
Sales Enablement
Content Design
Sales Agent
Sales Agent
Sales Enablement
Sales Agent

Time spent

Secondary Research

Ai Capabilities

LLM’s

  • Generally anything associated with understanding and generating natural language
  • Translating, summarizing, reframing, questioning natural language
  • Understand and map natural language to other format (voice to text)

Other Ai

  • Consolidating and analyzing big data, forecasting, and predicting
  • Search and recommendation algorithms
  • Recognizing patterns and detecting anomalies

Ai Insights

People

Users don’t need to know there using AI, tools can just be smarter or meaningful.

Ai tools can help teams save time and put more energy toward selling.

Technology

Ensuring that the integration of Ai fits product use case.

Ai isn’t a solution in itself but a technology that can enable a solution.

People

Can a LLM be trained from the uploaded content in a customers tenant?

When generating Ai response, can we sight sources when applicable, and how many?

Primary Research

9 User Interviews

Research Goals

Gather rich, personal narratives about users in the context of the sales cycle, their day to day activities, challenges, and possible frustrations.

  • Discover the daily tasks of Revenue Teams
  • Explore challenges and pain points faced by users
  • Understand user sentiment towards AI

Learnings

User needs

Sales Agent

Selling products or services to customers, aiming to maximize revenue and build lasting client relationships.

Sales Enablement

Equipping sales agents with the necessary tools, training, and resources to enhance their effectiveness.

Marketing & Design

Collaborate to create engaging content that effectively communicates the brand's message, attracts customers

Pain points

Locating Content

Finding the best and applicable sales material to share with buyers

Sales Enablement

Equipping sales agents with the necessary tools, training, and resources to enhance their effectiveness.

Marketing & Design

Collaborate to create engaging content that effectively communicates the brand's message, attracts customers

Affinity Map

28 Feature ideas
3 Solution spaces
2 Viable solutions to explore

Solution Space

How might we

discover pain points and challenges within the sales process, identifying opportunities to increase productivity with the integration of Artificial Intelligence into the product ecosystem.

How might we

create an experience allowing users to find the information and content they are looking for faster, allowing them to spend more time closing sales and doing higher-level tasks like customer-relationship building.

Design
Framework

Content Discovery

Helping users locate the content that they are looking for in a content warehouse

Winning Content

Sales agents can struggle with selecting winning content for a sales engagement

Passive Ai Integration

An Ai integration doesn’t need to be “flashy // new” but can be passive and enhance the user experience

Questions
& Curiosities

Fine Tuning

Can we train a LLM based on company uploaded content?

Current Search

How are people currently managing & finding content?
What are users querying?

Risks

Feasibility of managing unique LLM’s for customer tenants.
Sensitive company information passed into public domain.
Content permissions and ensuring that the right user receives information they have access to.

Semantic Search

Ai powered search experience

Overview

Utilizes advanced Ai algorithms to understand the context and intent behind sales-related queries, providing highly relevant and accurate search results.

  • Trained on company’s unique dataset, including internal documents, communications, and domain-specific knowledge.
  • Ensuring that the LLM is trained on a comprehensive set of company-specific content while maintaining strict data privacy and security protocols.

Experience Overview

Query

  • Users need to locate something
  • Knowledge to share with a buyer
  • Looking for knowledge for themselves
  • Articulate what they’re in need of
  • Choose where you start search
Intuitive integration and utilizes current UX patterns
Intuitive integration and utilizes current UX patterns

Return Results

  • Return useful & applicable content
  • Returned a block of knowledge with source links
  • Review source content
  • Narrow down results w/ filtering
  • Ask a followup to learn or new search
Populate more relevant and better matched search results
   Users need to be able to preview the content.     
Linking generative statement to content limiting Ai source ambiguity

Utilize Findings

  • Use what you found
  • Familiarize yourself
  • Add to a sales presentation
  • Distribute to client
  • Turn into training on search topic
Users need to be able to easily share the content and implement
Useful paths exist to Chat to continue conversation

Experience Map

User Query
Search Results
Extend
Looking for content
Looking for information
Ai optimized results
Review returned content
Ask Questions
Ai to chat with content
Ai trained on company content
Traditional search results
LLM response
Bookmark for future
Generate text
Ai to understand query
Implement content
Ai Integration
Ai Integration

Design Iterations

Side Panel Integration

Feedback

  • Doubles down on current Ai chat Ux Experience that is not fully flushed out
  • Returned Ai search results could be overlooked
  • Creates confusion with traditional vs new Ai results and which one should be used
New Page Section

Feedback

  • Ai generated text wasn’t useful and could be a hinderance to them finding their content
  • Developing and adding a new page section would be time consuming for engineering team

Solution Design

Next Steps

  • Integrate an NPS Feedback system to further train each model
  • User test to identify barriers to entry to create a meaningful first use experience
  • Further develop use cases and integrations of ai