What this step is for
The purpose of Base Analysis is to produce a first, structured read of your unadjusted pay gap and a clear set of hypotheses about what is causing it.
In practice, Base Analysis does three things:
It defines who you compare to whom, through the Comparable Jobs setup.
It produces an unadjusted view of the gap on the dashboard - overall mean and median figures, quartile distribution, and a per-comparable-job breakdown.
It gives you the inputs you will need for the next step: a prioritised list of areas to investigate in Drill-Down Analysis and Adjusted Pay Gap analysis.
Base Analysis is not the place where you draw final conclusions or decide on remediation. The figures are unadjusted, meaning they do not yet control for role, seniority, tenure, location or other legitimate pay factors. The aim of this step is to understand the shape of the gap, not to explain it fully.
What is most important in this step
Three things matter more than anything else in Base Analysis:
Get Comparable Jobs right. The way you group employees decides who is compared to whom across the entire analysis and how results will later be reported by category of workers. The grouping should reflect equal work or work of equal value, be defensible against criteria such as responsibility, skills, effort and working conditions, and remain practical (not so granular that groups become too small to analyse).
Read the mean, median and quartile views together. Each of the three measures answers a different question, and they only make sense in combination. Looking at any one of them in isolation almost always leads to the wrong story.
End the step with hypotheses, not conclusions. Base Analysis is a starting point. The output is a structured set of working hypotheses about where the gap comes from representation, within-role pay, a specific pay component, or a mix - that will be confirmed or revised in the next steps.
Step-by-step: how to complete Base Analysis
The Base Analysis collection takes you through this step in three articles. We recommend going through them in this order:
1. Set up and review Comparable Jobs Decide on the grouping logic that defines equal work or work of equal value in your organisation. Test a few scenarios (e.g. Job Level only, Job Level + Job Family), agree on a setup that is both defensible and practical, configure it in Organization Settings, and review the resulting groups in the dashboard. If you need a different setup for a specific legal entity, override it under Analysis → Legal entities analysis settings. 👉 Comparable Jobs - set up and review in PayGap
2. Review your Base Analysis on the dashboard Open the dashboard and work through the three sections in order: the overall mean and median tiles at the top, the quartile distribution in the middle, and the gender pay gaps by comparable job at the bottom. Use the pay component filter to see how individual components contribute, and at the end of the review, formulate a primary driver hypothesis (representation-driven, within-role pay-driven, pay-component-driven, or mixed), a shortlist of comparable jobs to investigate, a list of data concerns, and a list of open questions for stakeholders. 👉 Review your Base Analysis in PayGap
3. Understand the measures behind the dashboard Use this companion article whenever the figures need a deeper interpretation - for example when the mean and median are far apart, when the quartile split looks counterintuitive, when you see a negative gap, or when you are preparing to explain the numbers to leadership, HR business partners or employees. It walks through mean, median and quartile distribution with worked examples and edge cases. 👉 Understanding the measures in Base Analysis
What good output looks like
By the end of Base Analysis you should have:
a Comparable Jobs setup that reflects equal work or work of equal value, is defensible, and gives groups large enough to analyse,
a clear read of your overall unadjusted pay gap: mean and median, with an understanding of why they differ,
a view of representation across the four salary quartiles, read against your overall gender ratio,
a shortlist of comparable jobs to investigate further (typically those above the 5% threshold, and any with extreme values that need verification),
a view of which pay components contribute most to the gap (base salary, variable pay, allowances, total cash),
a primary driver hypothesis: representation-driven, within-role pay-driven, pay-component-driven, or mixed,
a documented list of data concerns and open questions to address before, or during, Drill-Down Analysis.
When you have these, Base Analysis is complete and you are ready to move into Drill-Down Analysis and the Adjusted Pay Gap analysis, where the hypotheses formed here will be tested.
Common questions
Is Base Analysis my final pay gap result?
No. Base Analysis is an unadjusted view. It tells you where the gap is and how it is shaped, but it does not yet control for role, seniority, tenure or other legitimate pay factors. The full picture comes from the Adjusted Pay Gap analysis in the next step.
Should I always use the most granular Comparable Jobs setup possible?
No. More granular is not always better. Very small groups produce unstable averages, can be harder to justify, and may even separate roles that represent work of equal value. Use the broadest grouping you can defend on the basis of responsibility, skills, effort and working conditions.
The mean and the median are very different. Which one should I use?
Both. They answer different questions. The mean reflects total financial disparity (and is required by most reporting frameworks, including the EU Pay Transparency Directive). The median reflects the typical employee. A large gap between the two usually means a small group of high earners - typically men at the top - is pulling the mean up. The article Understanding the measures in Base Analysis covers this in detail.
Can I change Comparable Jobs once Base Analysis has started?
Yes, but with care. Base Analysis itself refreshes automatically when you update Comparable Jobs, but if you have already started justifying outliers in Case Management, PayGap will treat the change as a change to the analysis logic and may ask whether existing justifications should be reset. Avoid changing Comparable Jobs once justification work is under way.