Use Case: Smart Sales Intelligence

In modern sales, every second counts. Imagine you're a salesperson about to make an important call to an unknown (at least to you) prospect. You wonder:

• Has this company ever been a customer?

• Did someone speak to them recently?

• How big is the company?

Finding answers has traditionally meant diving into time-consuming manual research that often leaves you with even more questions. But thanks to AI augmentation, there's a smarter way.

Let's see how your sales reps can easily access relevant customer info so they can start a quality sales process without delay!

Looking for daily AI news? Check out our sponsor:

Learn how to make AI work for you.

AI won’t take your job, but a person using AI might. That’s why 500,000+ professionals read The Rundown – the free newsletter that keeps you updated on the latest AI news and teaches you how to apply it in just 5 minutes a day.

Subscribe with one click.

Problem Statement

Converting B2B prospects into customers can take forever, and engaging prospects in sales calls is no piece of cake. Your sales reps need to be prepared. That means:

• Searching the CRM and finding the right account: Even in a midsize company, a simple CRM search for a company like "GM" can return dozens of results.

• Crawling through call notes: Figuring out who contacted whom can take forever and create awkward moments if you don't have that info.

• Keeping up with the news: Some topics should or shouldn’t be brought up in small talk. But who has time to read all the news?

• Complicated analytics: Sales reps often lack the skills, data access, or time to use analytics to gain deeper customer insights before a call.

These challenges are not just small problems. When added together, they create big obstacles to successful sales.

Either the sales rep shows up unprepared for a call, or they spend way too much time researching a prospect before picking up the phone - only to find that no one is answering or the person they need to talk to is on vacation.

The bottom line is missed opportunities, wasted effort, and poor performance.

What sales teams need is not more data but the right data at the right time.

Solution Overview

The Smart Sales Intelligence chatbot is designed to address these challenges. Here's how it works at a high level:

User layer

The user interacts with the system through a simple chat interface powered by an LLM fine-tuned to custom chat instructions. This chatbot can be integrated into any application that sales reps use: the home page of the CRM system, an Outlook email client, or even a WhatsApp chat on their phone.

Analytics layer

The solution essentially consists of two AI services. An AI chatbot that handles the conversation with the user, and an AI agent that is fine-tuned to retrieve data from other internal or external services, given some input it receives from the chatbot.

Data layer

All data sources required for this use case must be available as internal or external APIs that the agent can access.

The solution architecture is designed to be modular and scalable, making it easy to integrate with your existing systems and adapt to your organization's evolving needs.

Solution Breakdown

To understand how this solution works, let's consider an example and see what happens under the hood:

Try this out!

If you want to follow along, check out this live demo of the Smart Sales Intelligence chatbot.

The demo has access to the following “CRM” data:

• Acme

• Vertex

• AquaPure

• DynaTech

• Beacon

Walkthrough:

Imagine you're a sales rep about to call a potential customer, Acme Corporation. You've never dealt with them before, so you turn to the Quick Sales Intelligence chatbot for help.

You start by throwing "Acme Corp" into the bot.

What happens then is that the chatbot (first AI service) takes that input and passes it to an agent (second AI service), which performs a search across some internal data sources.

There are three things that can happen:

1. There's no hit

2. There's one hit

3. There are multiple hits

For cases 1 and 3 the chatbot would ask you to provide more details to narrow the search down (or let you know there’s no match respectively and suggest similar company names that exist).

In case 2, the agent will pull a unique identifier such as the CRM account ID and, if available, an external identifier such as a DUNS number and return it to the chat.

When you ask for more information, this time the chatbot will not only give the company name, but also the newly retrieved IDs to the agent, so the agent can use these IDs to directly query your CRM system and retrieve more information, such as opportunity histories or call notes:

(For the techies: Yes, there's a JSON exchange between the two AIs, which might look like this:)

    {
      "source": "CRM",
      "salesforceAccountId": "1000001",
      "DUNSNumber": "5-048-3782",
      "stats": {
        "customerLifetimeValue": 250000,
        "wonOpportunities": 15,
        "numberOfContacts": 5
      },
      "callNotes": {
        "salesNotes": [
          {
            "contactID": "C001",
            "note": "Discussed new product features, very interested in the upcoming release.",
            "date": "2024-01-10"
          },
...

Similarly, the external ID could be used to search a commercial B2B database for up-to-date information on this company:

With this wealth of information at your fingertips you could start your sales call, or even go one step further.

Instead of just pulling static CRM or external company data, you could hook this up with internal analytics services to get even more insights.

For example, if you have a churn prediction model in production, you could use this solution architecture to get the churn prediction for that particular account right out of the chat conversation.

This is powerful stuff and makes advanced, data-driven insights available to a large base of users.

So, in a matter of minutes, without any manual research, you've gained a comprehensive understanding of Acme Corp and your likelihood of retaining them as a customer.

You're now well-prepared for your sales call, increasing your chances of success.

Key Outcomes

Key outcomes from this use case include:

• Efficient sales-call preparation: The system drastically reduces the time that sales reps spend on manual research, allowing them to approach more clients in the same time.

• Better customer interaction: It allows sales teams to approach calls with a comprehensive understanding of the prospect, leading to more informed conversations and better sales outcomes.

• Access to analytics: Data-driven insights help sales reps proactively address potential problems and capitalize on opportunities.

• Easy-to-use interface: The conversational interface makes this solution accessible even to those with limited technical or analytical expertise.

Conclusion

The Smart Sales Intelligence chatbot is an effective way to augment your sales process with AI, addresses core challenges during call preparation and unlock the outcomes discussed above. It allows your sales team to operate with greater intelligence, agility, and effectiveness.

I hope you enjoyed this deep dive today. If so, please leave a testimonial or forward this article to a colleague who might benefit from it.

If you need help implementing something like this for your own organization, just reply to this email.

Stay augmented, stay ahead.

See you next Friday!

Tobias