Retention & Attrition Baseline

Leadership wants a straight answer: is mentorship actually helping people stay? The Retention (Directional) report section gives you a number to work with — but numbers about people leaving are easy to overstate. This article covers how to set your baseline and how to read the section honestly.

Setting your attrition baseline

1. Go to Settings → Organization and scroll to the Reporting section.
2. Enter your Annual Attrition Baseline (%) — your organization's overall annual attrition rate, not a number specific to mentorship participants.
3. Click Save Reporting Settings.

Where to get this number

• Your HRIS's built-in turnover or attrition report (Workday, BambooHR, and SAP SuccessFactors all have one).
• Your People/HR team — they likely already report this number to leadership on a regular cadence.
• If your organization defines attrition a specific way (e.g. excluding retirements, or measured by fiscal year), use that same definition here so the comparison is apples-to-apples.

If you don't know the number yet, leave the field blank. The report will explain what's missing instead of showing a misleading result.

Reading the Retention (Directional) section

Retention (Directional) is one of the toggleable sections on the Reports page. When included, it shows:

• Annualized participant attrition — computed only from members whose employment status is confirmed by your HRIS sync (a recorded termination date). The rate observed over your report's date range is annualized and capped at 100%, so a short window doesn't produce an unreadable spike.
• Retention delta vs. baseline — your admin-entered baseline minus the annualized participant attrition rate, showing whether participants are retained at a higher or lower rate than the org overall.
• Program departures, shown separately — members who left the mentorship program (ended a match, opted out, were deactivated) but did not leave the company are labeled "Left the program (not counted as attrition)" and are excluded from the attrition number entirely.

Reading it honestly

• This is directional, not causal. Participants aren't a randomized control group — people who opt into mentorship may already be more engaged, more senior, or more tenured than the average employee, and any of those could independently affect retention. A gap between your baseline and the participant number is worth investigating with your People team, not worth quoting on its own as "mentorship reduced attrition by X%."
• Watch the sample size. Orgs with fewer than 20 HRIS-covered participants get a low-sample caution on the report — at that size, one or two departures can swing the annualized rate dramatically.
• Check the coverage caveat. Attrition is computed only from members with HRIS employment data. If part of your participant base isn't HRIS-covered, the report tells you how many members are excluded — the number reflects that covered subset, not your entire program.