Back
Hooray! You’ve turned on Ambition’s AI Assistant for your team. Now what?
The assistant is an extremely powerful tool for users at every level of a revenue organization, but, as with any AI chat tool, finding the prompts that best surface the unique insights you need can take some practice.
Here, I’ll walk through some of the tips, prompts, and best practices to keep in mind when using the AI Assistant to make sure you’re getting the most comprehensive information as efficiently as possible.
Share this article with the rest of your team to help them level up their prompting, too!
Best Practice 1: Be As Specific As Possible
When prompting the AI Assistant, it’s important to be as specific as you can. If you’re not getting the results you want, it’s likely because the prompt was vague and needs some refinement. The Assistant is smart, but it can’t read your mind!
For example, let’s say you want to see all outbound calls for your team this week, so you type in “Calls” and hit enter. The assistant only knows that you’re interested in information about calls. It will take a guess and might pull your team’s call metrics for a certain timeframe, but it might not be the specific data or timeframe you had in mind.
To be more effective, try to include what, when, and who you’re looking for with your first question. Using the example of outbound calls, you might include:
- Who: “My team” or a specific team member like “Madison”
- What: “Outbound call metrics,” or if there is a specific goal associated with the calls, write the name of the goal you are interested in
- When: “This week”
So, the prompt becomes, “Show me my team’s outbound call metrics for this week,” instead of just, “Calls.” Now, the Assistant knows exactly what you’re looking for and can answer your question comprehensively on its first try.
Best Practice #2: Use Prompt Chaining
Rather than putting multiple questions in one message to the AI Assistant, try using a prompt chaining approach. Prompt chaining is a method that breaks complex questions down into simple steps by using the output from one question as the input for the next. It’s a great way to improve accuracy and control when leveraging conversational AI.
To best use prompt chaining, always start with a question that lays the data groundwork for the overall topic you’re asking about. Then, let your curiosity take the lead and dive into follow up questions about the gathered data.
For example, imagine you want to surface data about your team’s goals, and then go deeper to understand how it may impact the business or how your team compares to other teams. Here’s how that might look for a manager who wants to understand how their team is pacing alongside another manager’s team:
- Question 1: “How is my team pacing towards their revenue goals for this year?”
- Question 2: “How is Madison’s team pacing towards their revenue goals for this year?”
- Question 3: “What can I do to match Madison’s team’s performance this year?”
- Question 4: “What opportunities are coming down the pipeline that will impact revenue goals for my team and Madison’s team?”
In this example, the first two questions gather the foundational data (how your team and Madison’s teams are pacing towards goal), so that the AI Assistant can build off of it to answer the questions you really want answered (what your team needs to do to match or outperform Madison’s team). Laying groundwork in this way helps the Assistant understand the context of your question so that it can give the most helpful answers.
Best Practice 3: Be Mindful of Data Blind Spots
When interacting with the AI Assistant, it’s important to remember that it doesn’t necessarily have all of the relevant data to your business or team. So if, for example, you find that the Assistant can’t provide answers to questions about your revenue goals for the year, check the Goals tab to see if any have been built yet. If they’re not there, reach out to your Ambition manager or, if you are the manager, set them up.
If the Assistant doesn’t have the information it needs to answer your questions, it might make educated guesses. For example, you might ask if an employee is on target for their calls this week, but there’s no specific goal around call targets set up in Ambition.
In this instance, the Assistant will look at past history to make an informed guess—maybe by averaging the employee’s call volume from the past several weeks to make an assumption about the target. But if the employee’s call volume had been unusually low over the time period the Assistant draws from, it may return inaccurate answers like:
“Yes, Daniel is on track for his call target with eight of 10 calls made for this time period.”
Here, it’s looking at Daniel’s abnormally low call volume and making a wrong assumption about the target.
The answer should have been:
“No, Daniel is significantly behind on his call target with eight of 50 calls made for this time period.”
Because the Assistant doesn’t know that the actual target is 50 calls, it doesn’t have the context to give this correct answer.
Bottom line: To get the most accurate, context-rich answers, make sure Ambition knows as much about your team’s goals and targets as possible.
Best Practice #4: If You Don’t Know What You’re Looking For, Start with Basic Questions
You can use the AI Assistant to surface the basic information you need to ask your questions. For example, if you don’t know the specific name of a metric or goal you want to ask about, start by asking the Assistant to list the ones available before you ask your follow up question.
Let’s say you want to ask about recent closed/won trends but aren’t sure exactly how it’s labeled. You could use a simple chain prompt to ask:
- Question 1: “What metrics are available?”
The AI Assistant will generate a list of all the metrics it has related to your Ambition environment. You spot the one you want (Closed Won Opportunities) and follow up with:
- Question 2: “Will you show me the trends in total “Closed Won Opportunities” over the last three months for Daniel and Madison?”
This way, instead of guessing at the name of the metric you’re interested in and potentially confusing the AI Assistant, you can start with one simple question that ensures your prompt is specific and returns a correct answer on the first try.
Best Practice #5: Be Creative
The beauty of Ambition’s AI Assistant is that it lets you ideate through data and get immediate visibility into how reps and teams are doing. Don’t be afraid to experiment with its abilities and explore any limitations. If you have an idea for a question, just try it! Tell it your goals and ask away.
For example, a self-motivated rep might say: “I want to understand how I am performing compared to my peers and what I can do to improve. My peers are Madison, Daniel, and Corinne. Look at our current active competitions and metrics for call volume, appointments set, and open pipeline.”
This is a fairly complex prompt, but because the rep provided the appropriate context (specific metrics and peer names), the AI Assistant should be able to tackle it. Even if you’re not sure if the Assistant will be able to answer your question, offer it as much detail as you can and give it a whirl. You might be surprised at the insights it can offer.
That’s it! You’re ready to prompt like a pro. If you’re a customer and have additional questions about Ambition’s AI Assistant or would like to have it turned on for your team, reach out to your account team.
Not a user already? Schedule a demo to learn how Ambition can transform your org’s sales performance and see the AI Assistant in action.
