5 Essential Questions Salesforce Admins Must Ask for Effective AI Solutions


You know artificial intelligence (AI) is officially everywhere when your 73-year-young mother asks you if AI can create a birthday card design for one of her friends. Which is another way of saying, it’s most likely crept into your business conversations with stakeholders and users. If not, then let’s be proactive Salesforce Admins and get ahead of the game by looking at five questions you should ask your business to build amazing AI solutions in Salesforce.

But let’s not get ahead of ourselves. First, what are the AI tools admins have at their disposal? 

Let’s start with Prompt Builder in Salesforce, which is used to integrate generative AI into various workflows, enhancing user efficiency and customizing interactions through AI-generated content. Users can create different types of prompt templates, such as Sales Email, Field Generation, and Flex, which are adaptable to various data types and use cases within the Salesforce Platform. 

These templates leverage customer relationship management (CRM) data and can be invoked via Salesforce’s platform tools, allowing users to generate personalized communications and content dynamically. The management and customization of these templates involve a user interface (UI) where admins can revise and perfect templates based on feedback and performance, ensuring the AI outputs meet the desired standards and effectiveness.

Next, we have Einstein Copilot which enables users to interact with the system using natural language to perform tasks and retrieve information, effectively enhancing productivity by drawing answers from secure company data. The tool is embedded within the Salesforce ecosystem, respecting existing access controls and integrating with the Einstein Trust Layer for ethical and secure AI operations. 

Users can engage in conversational exchanges with Copilot, asking questions and receiving context-aware responses, thus simplifying tasks like summarizing deals or drafting emails without needing advanced technical skills. Unlike Einstein Bots, which require extensive setup and operate on rigid rules, Einstein Copilot offers a more user-friendly and flexible approach, utilizing a large language model (LLM) to understand and respond to user requests in a conversational manner.

Now that we have that aside, let’s look at the five questions Salesforce Admins need to ask of their organization and its stakeholders so we can build the best, most powerful tools for our users.

1. Which mundane tasks are driving your team bonkers?

So, why are we asking this question when we could just be building cool prompts?! This is going to help us build the most useful AI solution first. 

Repetitive tasks, like sending emails for sales and service people, can add to frustration. Especially if they need to draw information from the record and offer a personalized voice in that email. Because, let’s be honest, templates work only some of the time, and most users end up customizing the output right away. Look for answers that will guide you to create the most useful prompts in Prompt Builder first, like a post-meeting follow-up thank you message.

2. How painful is it to quickly catch up on a customer record?

You probably know where I’m going with this question, but it’s the details in the answer that we really need. Do they want a summary text field on the record? And if so, which information is important? This will help you correctly ground your prompt to pull the right data in and build a summary they are expecting.

3. Are we missing the mark on collecting important data?

As Salesforce Admins, we can build some pretty sweet prompts. But if the answer to question #2 is “give me a summary of the last five customer interactions based on activities logged”, however, there is no enforcement on logging activities. Well, then the summary isn’t going to give the desired result. 

The goal with this question is for us to help the organization prioritize and enforce good data collection so that our prompts perform!

4. How can similar data turbocharge your decision-making?

Here’s where it gets interesting. If you’re in sales, finding opportunities similar to the one you just nailed can be invaluable. Or, if you’re in customer service, quickly pulling up all records for a frequent caller can save the day. 

This isn’t just about speed; it’s about making connections faster and more accurately.

5. What does success look like, and how do we track it?

It’s super easy as an admin to get caught up in rolling out a new feature or building in the latest technology. But it’s important to think about your rollout plan, your training, and how you want to evaluate the success of your prompts and Copilot actions. 

Even more important is making sure that your stakeholders agree with the plan and what success looks like. Now for some or many of your users, using an AI feature could be intimidating, which means it might not be for everyone at first. A core group of engaged early adopters who represent the larger user base will be key to helping you refine and track each feature.

Solve simple solutions along the way

It’s important to remember that there are some easy solutions we can give our users when having an AI conversation. 

For example, with question #1, let’s say your sales users answer with “I’d ask Einstein to give me my top five opportunities closing this week every day.” This is an opportunity for you to go back and offer up a list view with global variables to your sales org as a quick win. Sure, Einstein can offer that in a prompt, but now you’re saving your user from asking the same prompt every time and getting them right to the answer.

The same can be said if they find themselves asking Einstein for all the opportunities or cases related to a single Account. This could be a signal that a page layout needs to be refreshed. Maybe think about how you can use Dynamic Pages to show that relevant related list higher on the record, or on mobile. That will free up the user to have a deeper, more meaningful interaction with Einstein, so you are building better solutions.

Ready to put Einstein to action? We launched the #BuiltWithEinstein Quest to help you show off your creative AI use cases to the community for a chance to win* a Dreamforce pass or a custom #BuiltWithEinstein T-shirt! *Official rules apply. See the Trailblazer Quests page for full details and restrictions.

It’s simple. You’ll build a prompt with Prompt Builder or an action with Copilot Builder and share your use case in a screen-recorded video on X or LinkedIn with the hashtag #BuiltWithEinstein.

Visit the #BuiltWithEinstein website to view custom Einstein actions and prompts built by Salesforce Admin Evangelists and Developer Advocates. 

Learn more about the #BuiltWithEinstein Quest and be sure to complete it by June 30, 2024.


How Salesforce Einstein Is Supercharging Mobile Experiences.

How Salesforce Einstein Is Supercharging Mobile Experiences

While its impact is widespread, one of the most exciting aspects of artificial intelligence (AI) is its ability to create conversational interactions that generate personalized experiences, supercharging productivity and efficiency. In this blog post, we’ll explore how the implementation of large language models on mobile devices is reshaping the enterprise mobile landscape and how Salesforce […]

Einstein standing next to text that says, "How to Use Generative AI Tools to Write SOQL Queries."

How to Use Generative AI Tools to Write SOQL Queries

Salesforce Object Query Language (SOQL) is a powerful tool that allows you to retrieve data from Salesforce. You can use SOQL to query any Salesforce object, including custom objects, custom fields, and user permissions like profile and permission set perms. As a Salesforce Admin, I know that writing SOQL queries can be a pain. Not […]

Headshot of Tom Hoffman next to text that says, "AI Prompt Writing for Salesforce Professionals."

AI Prompt Writing for Salesforce Professionals

The rise of the machines Machines and artificial intelligence (AI) have been part of popular discussion since Samuel Butler authored Erewhon (1872), where his satirical utopian society explored the morality of conscious machines as a natural evolution of the Industrial Revolution. One-hundred and fifty years later, OpenAI released GPT-4, introducing the world to AI that […]