Creates a dashboard of headcount, attrition, engagement and performance metrics from your HR data, with the formula under every chart.
Build a people analytics dashboard
This skill builds one people analytics dashboard from the data you already have. It finds which of your connected tools hold people data, tells you what each export would unlock, then builds the blocks your data can support, with the formula shown under every chart. If you have no data to hand, it can build a sample dashboard on invented data so you can see what's possible.
When to use it
- Leadership asks for headcount and attrition numbers and you rebuild them in a spreadsheet each time.
- People argue about how turnover was calculated instead of what to do about it.
- You have an HRIS and a review tool but no single view across them.
- Someone asked "why are people leaving" and you want the data in front of you.
- You want to see what a people dashboard looks like before deciding which exports to pull.
What you get
One dashboard you can reopen next month, with up to ten blocks:
- Headcount and composition
- Retention and attrition
- Engagement and sentiment
- Performance and talent
- Career and mobility
- Compensation and pay equity
- Hiring
- Absence and wellbeing
- Diversity and inclusion
- Workforce planning
Each block opens with the question it answers, the data sources, and three or four KPIs with a comparison against the prior period. Under every chart is the formula used and one sentence on why that chart type. It also keeps the code that computed the numbers, so they can be re-run on next month's export.
How it works
- It finds what's reachable. It checks your connected tools for employee, payroll, survey, review, hiring and time-off data, makes one small read of each to confirm the fields are populated, and shows a table: source, which blocks it can feed, what's still missing.
- It asks for data and shows what each kind unlocks. An employee list unlocks headcount and the segmentation every other block uses. An exits file unlocks attrition. Survey results unlock engagement. It profiles each file when it arrives and gives a feasibility line per block.
- It chooses the blocks, computes, and builds. Four full blocks beat ten half-empty ones. A block is included when its top two or three metrics can be computed, and skipped with a reason otherwise. Every metric is computed in code from the files before the page is written.
The rules it holds to. Any survey slice with fewer than five respondents is suppressed and shown as "n<5". Turnover uses one stated denominator across the dashboard. Every KPI carries a comparison, or says "no comparison available". A flight-risk score is shown only with its drivers, and only when the data supports it.
What it never does. It doesn't invent numbers or supply benchmarks from memory. The sample version is labelled as sample on every block. It doesn't infer gender, ethnicity or age from names: demographic blocks use only explicit, consented columns. Pay-equity numbers are reported as findings, and decisions on them need HR, finance or legal.
How to run it
Install the Effy AI plugin for Claude or ChatGPT. One install adds every Effy AI skill, this one included. Then connect the tools you want the skill to read from, such as email, calendar, your drive or your meeting notetaker, and ask for what you need in plain words: the assistant picks the right skill.
Before you start: connect your HR system, review tool or survey tool, or have exports ready as CSV or spreadsheets. An employee list plus exits, reviews and survey results is enough for the first four blocks.
- Ask, for example: "Effy, build a people analytics dashboard."
- Upload the exports, or ask for the sample dashboard.
- Review the dashboard. It ends by naming the one export that would unlock the most next.