AI for Workday: Prompting 101

So, here's the thing.

You keep hearing about how AI can help with everything. You decide to try it out on Workday config. Makes sense, right?

Maybe your company has Mando AI or you're just using ChatGPT on your own time. Everyone's talking about AI like it's going to change your life. "It's like having a Workday consultant available 24/7! It should be able to explain configurations, write reports, catch mistakes you'd never see coming!"

You're pumped. Finally, something that might make your job easier.

You open your AI tool of choice, fingers ready, and type the first thing that comes to mind: "Someone got promoted but their salary didn't update."

And then... nothing. Not nothing—but so simply not helpful, you might as well have asked your coffee mug. The AI has no context on what kind of data you want, for which employees, or what you're even trying to accomplish.

Sound familiar?

Here's what I bet you're thinking: "Am I completely screwing this up?"

Here's the thing - it's not because you don't know Workday. You know Workday better than most people know their own email inbox. The problem is simpler than that: You've never had to explain Workday logic to a machine before.

Think about it. When you're configuring a calculated field or setting up a business process, you're clicking through menus, selecting from dropdowns, navigating familiar paths. You're not narrating what you're doing like some kind of Workday sports commentator.

But that's exactly what helps AI when you're asking about Workday.

Here's the Deal with AI Prompts

AI isn't Google. It's not a search bar where you throw some keywords and hope for the best. It's more like having a conversation with someone who's smart about technology but needs you to be super clear about what you're asking for - especially when it comes to something as specific as Workday.

The more specific you get, the better it performs. And it all comes down to one thing that'll change how you think about AI prompts forever:

Start with the object you're working with. Always.

This is a mental shift—and it’s what I call the Decision Tree Approach. Once you get this down, you'll be pulling calculated fields, business processes, domain security setups, and reports out of AI tool like you've been doing it for years.

Let me walk you through it.

What Even Is a Prompt Anyway?

Look, a prompt is just you telling the AI what you want. It's that simple. But the difference between a good prompt and a terrible one? Specificity.

Good prompts give the AI three things:

1. A clear object to focus on

2. A specific action you want it to take

3. The exact output you're expecting

But you've got to speak the AI's language first when it comes to Workday. And that language starts with objects.

The Object Is Your North Star

Here's what I mean by object - it's the "thing" in Workday you're trying to work with. It’s the starting point for basically every useful prompt you'll ever write.

We're talking about:

• Worker

• Job Requisition

• Position

• Compensation Plan

• Organization

• Business Process

When you lead with the object, you're grounding your question in how Workday thinks about data. You're giving the AI the context it needs to go from high-level generic responses to specific and useful answers.

The Decision Tree That Changes Everything

Okay, so you've named your object. Now what? This is where most people get lost, but it's straightforward once you see the pattern.:

Here's your decision tree:

Start with: OBJECT (let's say Worker)

Then ask yourself:

• Need a field? → Ask for a Calculated Field

• Need a process or workflow? → Ask about the Business Process

• Need to understand permissions? → Ask about the Security Domain

• Need actual data output? → Ask for a Report or data list

This isn't rocket science, but it's how Workday experts think about problems. Clear, structured, and actionable.

Let's Fix Some Real Prompts

Nothing beats seeing this in action, so let's go over some before-and-after examples that'll make this click. In this blog post series, we’ll categorize different prompting techniques where relatively small adjustments will go a long way in improving the quality of your AI’s responses.

The Vague Disaster
"The Vague Disaster" is what happens when you ask a question without adequate context and leave the AI hanging without enough information about what you’re trying to solve. As mentioned above, starting with the Object goes a long way in addressing these cases.

• What you may have typed: "How do I report on pay history?"

• Why this doesn’t work: The AI has no context for what you're talking about. What kind of compensation data? For how long? In what format?

• What you should type instead: "As a Workday Compensation Analyst, create a report showing salary progression over the last 3 years by employee."

• Why this works: You started with the object (Employee Compensation Event), specified what you want (custom report), and listed exactly what data you care about. Boom.

Here’s an example of a question that Mando AI answered perfectly because of the way it was prompted.

When You Don't Know the Object

What if you're not sure which object to use? Just ask the AI to help you figure it out first.

Try something like:

• "What object stores base pay and grade profile for employees?"

• "What's the best object to use for tracking job applications and recruiter activity?"

That one question can save you from going down completely wrong paths. Let the AI guide you to the right starting point, then build your real prompt from there.

Your Quick Reference Guide

Before you fire off your next prompt, run through this checklist:

1. Name the object right up front - don't make the AI guess

2. Use Workday's structure - calculated field, business process, domain, report

3. Get specific - include context, constraints, what you expect back

4. One goal per prompt - don't try to solve everything at once

Try This Right Now

Next time you open your AI tool - whether that's ChatGPT or a Workday-specific AI product like Mando - resist the urge to just ask a question. Start with an object instead.

Try something like:

• "For the Compensation Plan object, what calculated fields show salary range penetration over time?"

• "What Business Process governs onboarding for a Job Requisition in EMEA?"

Feel the difference? You're not just throwing questions at a chatbot anymore. You're having a structured conversation with a system that understands how Workday works.

And the more you practice, the better - and faster - you’ll get.

What's Coming Next

In our next post, we’re going to go over six specific prompting strategies that'll take you from good prompts to absolutely killer ones. We'll talk about using verbs that Workday recognizes, how to stack filters like a pro, and how to guide the AI toward exactly the format and result you need.

Because here's the thing - you're not just chatting with some generic AI. You're working with a system that speaks fluent Workday. And now? You're starting to speak its language right back.