When asking Ask Mo (or any AI) questions, it is important to be precise and specific in your wording. When you provide Ask Mo with a high-quality detailed question (called a prompt) you are more likely to get a more high-quality detailed answer.
The goal of this article is to provide some basic tips for writing effective AI prompts. To do this we will also show some low-quality prompts, an analysis of each, and a more high-quality prompt.
A Basic Guide to Prompting
- Give Clear Direction: Tell Ask Mo what you want it to do: “Provide me a list of...” “Give me the total dollar amount of...” “Make a bar chart of...” Avoid using vague or general phrases like “Tell me about”.
- Use Precise Descriptive Language: Give Ask Mo specific instructions on the types of data you want to be given. Avoid using subjective words and phrases like “what was the biggest event in December 2025?”. Instead be specific and precise. “What event that occurred in December of 2026 had the highest attendance?”
- Provide “Do” and “Don’t” lists: It is important to tell Ask Mo what you want, but it’s also important to specify what you do NOT want. Providing both parameters ensures that you will get targeted responses.
- Refine your prompt: you may not get the answer you’re looking for right away. Tell Ask Mo to refine it’s response with new information that you provide. And in some cases Ask Mo may ask you for clarification. This is Ask Mo ensuring it has the most accurate, precise language it can to ensure a better quality result.
Example 1: Tell me about events in December.
Why is this a poor prompt?
- Lack of specificity of the user's goal. What do they want to know about? Attendance? A list of event names? A list of accounts hosting events?
- Solution: Be specific in your request.
- Not enough definition for key term. There are many kinds of “events” in Momentus. Some are confirmed. Some have been cancelled. Some are still tentative. Artificial Intelligence takes directions very literally. So without specific types of events, Ask Mo may deliver more or less information than you’re expecting.
- Solution: Be specific on what kind of data you want returned. In this case do you want ALL events? Or do you want confirmed events only?
- Vague date range. Again, this is a case of specificity. Is it December of the current year? December of last year? What about events that start in November and run through December?
- Solution: Define date ranges with months and years.
A more fully developed version of this prompt might say:
I want a list of confirmed events in our facility that start and end in December 2026. Tell me the name of the event, the account hosting the event as well as the start and end date for each event.
Example 2: How many service orders were closed in October?
Why is this a poor prompt?
- Lacking business specifics. This prompt may or may not return an expected response because it lacks specificity in the request. In Momentus you may host an event in one month, but not close the service order until a later month. In this case, Ask Mo would pull information for ALL service orders closed in October – regardless of when the event was closed.
- Solution: Define your expectations. If you want Ask Mo to return data within specific parameters, you have to provide those parameters.
- Vague date range. Is the user talking about October 2026? October 2025? Or earlier? Be specific in date requests.
- Solution: Be specific with dates.
A more fully developed version of this prompt might say:
I want to know how many service orders were closed in October of 2024. Limit this to events that started and ended in the October 2024.
Example 3: When is the RBG Conference?
Why is this a poor prompt?
- Terms not used in Momentus. Ask Mo can only utilize information in the way that it is structured in Momentus. In this example, a customer has asked about a conference hosted by “RBEG”. However, RBEG is NOT how the account is named in Momentus. In Momentus, the account is known as “Really Big Example Group”. Ask Mo, currently, cannot make the logical leap from “Really Big Example Group” to RBEG.
- Solution: When writing a prompt be sure to use language and terminology that matches information in Momentus.
- Vague date range. In this case the user asked “When”, but did not provide much specificity. Do they want to know past events? Do they want to know the next event only? Do they want to know the next few events hosted by this group?
A more fully developed version of this prompt might say:
When is the next conference being hosted by the Really Big Example Group? Provide the start and end dates for this conference.
Example 4: Who hasn't paid their invoices yet?
Why is this a poor prompt?
- Lack of specificity of the user's goal. What do they want to know about? Amounts? A list of customers? A list of invoices?
- Solution: Be specific in your request.
- Vague date range. Again, this is a case of specificity. What's the date range that should be considered? Is it all time? Should it be narrowed own to a specific time frame?
- Solution: Define date ranges with months and years with specific dates if applicable.
- Terms not used in Momentus. Ask Mo can only utilize information in the way that it is structured in Momentus. In this example, it is better to specify what you are looking for rather than a generic "who".
- Solution: When writing a prompt be sure to use language and terminology that matches information in Momentus. You could specify a specific account or ask "which accounts" to search across accounts.
A more fully developed version of this prompt might say:
I want a list of invoices for Account ABC, Inc. that are more than 30 days overdue. Tell me the invoice numbers, due dates, and amounts due.