Henshaw Prompting Guide & Best Practices

The Henshaw Assistant helps you explore workforce data using natural language with no technical skills required

But to get the best results, you need to ask the right kind of question.

This guide shows you how.

warning

Henshaw Assistant is not designed to process personal data. Do not input any personal information. Be aware that outputs generated by Henshaw may contain inaccuracies, so always verify the results.

Why Prompting Matters

Henshaw interprets your natural language prompt and turns it into a structured query behind the scenes.

That means the way you phrase your question has a big impact on the results you get.

Think of it like talking to a skilled analyst: the clearer you are, the more useful your answer will be.

Prompting principles to follow

Be clear and precise about the insight you want. Henshaw performs best with direct, focused questions.

Tips for clear prompts:

• Focus on one question at a time.
• Use simple, descriptive language.
• Avoid combining unrelated metrics in a single prompt.

Use natural language; you don’t need technical terms or formulas.

Iterate and refine your questions if the first answer isn’t what you expected.

Validate results by checking key figures or summaries against your Orgvue data.

Be specific about what you want to know.

• Poor: “Show workforce data”
• Better: “Summarize total headcount by department”

Mention the metric, group, or filter you want to focus on.

• Example: “Compare average salary by department in this pack”

Use specific property names in prompts when possible

• Poor: "Show cost breakdown across the organization"
• Better: Provide a breakdown showing total current salary by function"

Tackle one thing at a time. If you have a complex question, break it down.

• Start with: “Show average span of control by function”
• Then: “Now show the same data by level”
• Finally: “Compare this across regions”

Use shorter, more specific questions and clarify key details like the properties, inputs, or groupings.

Example: Instead of “Show spans in my workforce,” try “Show the average span of control for managers at each level of the organization.”

Use Case	Poor Prompt	Improved Prompt
Headcount	“Show headcount”	“Show total headcount by department”
Span of Control	“What’s my span of control?”	“Show average span of control for managers by level”
Attrition	“Who has left?”	“Show number of leavers by department”
Gender Breakdown	“Show gender data”	“Show headcount by gender across each department”
Performance Comparison	“Which function is best?”	“Compare total cost per employee across functions”

Advanced prompting tips

Once you’re comfortable with the basics, try these techniques to go deeper:

Refine your results to follow up on a previous prompt.

• “Now show only departments with fewer than 50 employees”
• “Filter to include only permanent staff”

Explore relationships between metrics in your data.

• “Compare headcount and total cost by department”
• “Show average salary and span of control by level”

Need help?

If you get stuck or receive unexpected results, try rephrasing your question

Be aware that outputs generated by Henshaw may contain inaccuracies, so always verify the results

Still unsure? Use the in-app feedback or reach out to support@orgvue.com