Why HR teams don't trust their own numbers.
Trust is built from documentation, not dashboards. Three habits that move HR analytics from spreadsheets-in-the-meeting to numbers-that-decide.
There's a moment that happens in almost every HR analytics meeting I've sat in. Someone shares a slide. A senior leader frowns. They say, "that doesn't match what I'm seeing in the system." The room goes quiet. The number was probably right. The trust just wasn't there to defend it.
Trust in HR analytics isn't a function of the dashboard. It's a function of the documentation, the data lineage, and the conversations that happened before the dashboard was ever opened.
Three habits separate the teams whose numbers get acted on from the teams whose numbers get questioned.
1. Write the business logic down, in plain English, before any DAX
Before I write a single measure, I write a one-page document. What is this metric? What does it include? What does it exclude? Why? What's the sign convention? What's the grain?
That page goes to the HR ops lead and the finance partner for sign-off. By the time the measure exists in Power BI, the conversation about what it means is already done.
When the number gets questioned in the meeting, you don't argue about the value. You pull up the page, point at the rule, and the conversation shifts from "is this right?" to "is the rule still right?" That's a much healthier place to land.
2. Reconcile to the source every refresh
The dashboard should agree with the source system — to the penny, or to a tolerance you've explicitly chosen and documented. If the HMRC levy account says the balance is £119,510 and your model says £119,507, you should know why. Rounding? Timing? A missing transaction?
I keep a tiny reconciliation tab on every report. Three rows. Source value. Model value. Variance. If the variance is zero, the dashboard inherits the credibility of the source. If it's not zero, the variance has a comment explaining why and a date when it'll be resolved.
That tab is the most powerful trust-building thing on the whole report.
3. Treat changes like a software change, not an Excel edit
When the formula changes, the version goes up. When the business rule changes, the change log captures it. When a measure is deprecated, it's left visible with a note for one quarter before being removed.
This sounds heavy. It isn't. A change log is a Notion page. A version is a string in a header. The discipline isn't in the tooling — it's in the habit of being clear about what changed and when.
The payoff is that when someone asks "did this number used to be different?" you don't say "let me check." You point them at the log.
None of this is technically difficult. All of it is socially difficult, because it means slowing down to write things down when everyone wants the answer now. But the slowdown is paid back tenfold the first time a finance director defends your number in a board meeting because they read the rule before the dashboard.
The dashboard is the easy part. The trust is the work.