Prepare your data for Pave Agent

What is this?

Pave Agent answers questions from the data your company has in Pave. The more of your compensation data lives in Pave, the more Pave Agent can do: an analysis that would mean pulling from your HRIS, your bands, your last merit cycle, and market data can come back from a single question.

Use this article as a checklist. It covers the data Pave Agent needs for the most common questions, the data that opens up more of them, and what to expect if something is missing.

What your Pave plan includes

Which Pave products your company uses decides what kind of data Pave Agent can work with.

• Market Data only. Pave Agent answers from Pave market data: benchmarks, peer groups, market insights, and public job postings. It does not analyze your own employees, pay, bands, or merit cycles. If you ask a question that needs your company's data, it tells you your plan does not include it and offers what it can do with market data instead.
• Market Pricing, Compensation Planning, Team View, Total Rewards, or Visual Offer Letter. Pave Agent can also work with your company's data in Pave, such as employees, jobs, pay, and bands. Pave market data benchmarks are available when your plan also includes Market Data Pro.

The rest of this article applies to companies with company data in Pave.

Data Pave Agent needs

These four are the foundation. Most questions about pay, such as how a team compares to its bands or to market, depend on all of them.

Data	Why it matters	Where to set it up
Job architecture	Every comparison starts from what job an employee is in. Job families and levels should reflect the jobs you actually have, not placeholder jobs.	Understand job architecture
A live band set	Pave Agent compares pay against your live band set by default. Draft and archived band sets are only used when you ask for them by name.	Create and manage band sets
Employees matched to bands	Compa ratio, range penetration, and who sits above or below their band all depend on each employee being matched to a band.	Match employees to bands
A market data source	Pave market data benchmarks, survey data you have uploaded into Market Pricing, or both. Pave Agent uses these for any question about how your pay compares to market.	Review benchmarks

If you do not have bands in Pave yet

You can still use Pave Agent for questions about your employees, your organization, and pay history, as long as that data is in Pave. Questions about bands, compa ratio, or range penetration depend on a live band set with employees matched to it, and may not return results without one. If you have a band set in draft, you can ask Pave Agent to use it by name.

Data worth adding

Teams that get the most out of Pave Agent have added more of their compensation data to Pave. Each of these opens up a new kind of question.

Data	What it lets you ask	How it gets into Pave
Equity	"How much unvested equity does the sales team hold?" Pave Agent works with equity values such as vested, unvested, and upcoming vesting.	An equity platform integration, such as Connect Carta
Equity and variable pay ranges in your bands	"Which engineers are below the bonus target for their band?" Bands that include bonus, variable pay, and equity values, not only base salary, let Pave Agent compare total compensation against your ranges.	Your band sets in Market Pricing
Compensation history	"Who hasn't had a raise in two years?" Past base salary, bonus, and variable pay changes with their effective dates.	Your HRIS integration
Managers	"What is the average span of control for directors?" Reporting lines let Pave Agent roll up teams and analyze organizational structure.	Your HRIS integration
Gender and ethnicity	"Is there a pay gap by gender within each level?" Demographic fields make pay equity analysis possible.	Your HRIS integration
Performance ratings	"How does pay positioning differ by performance rating?" Ratings loaded into a merit cycle in Compensation Planning.	A merit cycle in Compensation Planning

If a field is not in Pave, Pave Agent cannot use it. Data in a system that is not integrated with Pave is not available.

Things to know

Fix the data rather than writing around it. Guidance can tell Pave Agent how to interpret your data, such as your target percentile or how you define tech roles. It cannot stand in for data that is not in Pave. See the article on setting up Guidance.

A file can fill a gap. If data lives in a spreadsheet rather than in Pave, you can upload it to your file library and ask Pave Agent to work from it. Library files stay available across conversations, but a file does not update on its own. Data in Pave stays current through your integrations and the changes you make in Pave. See the article on uploading files and exporting results.

Pave Agent does not change your data. It cannot create jobs, publish a band set, or match employees for you. Make those changes in Pave. Changes can take up to a day to show up in Pave Agent's answers.

For a full list of the sources Pave Agent can draw on, see the article on what data Pave Agent can see.