Decisions, Not Dashboards

Most HR teams judge themselves on how well they help people thrive, through retention, engagement, and a healthy culture. However, the best HR teams do that and also focus on moving the needle on business outcomes through the implementation of HR policies and programs. The bridge between the two is HR analytics. When we use data to explain and predict outcomes, we can choose actions that can lift both people outcomes and business results. Senior leadership of the organization already expect this, the rest of the business functions already provide this, and the vast majority of CHROs understand that people analytics is core to the implementation of HR strategy.

However, when it comes to implementing People Analytics, there are still too many HR functions that rely on dashboards as the way to deliver value. This falls way short of what is needed, and often creates the wrong impression in the organization that people analytics is just about reports on HR variables with no insights, recommendations, or direct links to business KPIs

Descriptive vs Predictive Analytcs

Descriptive analytics summarises the past. Predictive analytics estimates the future and the drivers you can pull.

Think of it this way. Descriptive monitors, predictive decides. Use descriptive to know where you stand, use predictive to choose what to do next.

Descriptive analytics

• Answers “what happened.”

• Dashboards, scorecards, monthly reports, and heatmaps belong here.

• Good descriptive work brings clarity, exposes gaps, and keeps us honest.

Predictive analytics

• Estimates “what is likely to happen and why.”

• Uses patterns in your data to forecast outcomes and quantify the levers that matter.

• Turns HR into a strategic partner because it informs choices before outcomes show up in the P&L.

Here are some examples of predictive analytics that show how HR’s can add value to the business:

• Coaching and sales results. Model how coaching hours, content, and manager quality relate to conversion rate, average deal size, and time to quota.

  Decision: fund coaching Module B for Team East; expected +3 pts conversion, +$1.2M/quarter.

• Training and product cost or service quality. Forecast how training modules affect defect rates, rework, or first contact resolution.

  Decision: prioritise QA Module 3; expected −12% rework, −$280k/quarter refunds.

• Compensation and growth. Simulate how market ratio, pay mix, and equity grants influence quota attainment, ramp time, and regretted attrition in critical roles.

  Decision: target equity for Level 4 ICs in Growth Unit; expected −2 pts regretted attrition, +$3.1M ARR.

• Retention, engagement, and business outcomes. Link engagement to store sales, NPS, safety incidents, or SLA adherence.

  Decision: deploy manager enablement in bottom quartile sites; expected +6 NPS, +$900k/month revenue.

How the Workflow Actually Runs

• Start with a business question that matters this quarter.

• Generate multiple hypotheses and feature ideas. Expect most to be rejected; the survivors are your levers.

• Test the hypotheses with simple models; keep the few that explain the outcome and quantify effect sizes.

• Turn those into clear recommendations aimed at moving the needle.

• Track the result against a baseline.

Statistics You Will Actually Use

You do not need a PhD or a degree in mathematics to do People Analytics. But you do need to understand basic data management and practical statistical thinking: Distributions, sampling, statistical significance, confidence, correlation versus causation, T-tests, and regressions will get you most of the way there. 

Later, add forecasting to your toolkit. Learn multiple regression, Montecarlo simulations, and other causal models that include drivers such as hiring rate, tenure mix, or pay position. Use them to forecast attrition in critical roles, hiring demand, training throughput, sales compensation targets, absenteeism trends or overtime risk. 

I will publish a series on forecasting techniques in the coming months.

Finance Fluency for HR

The phrase “I am a people person, not a numbers person” is a false choice. HR must be both people savvy AND numbers savvy. Everyone in HR is also a business person, and thus should be able to understand the P&L and balance sheet, connect people actions to profit margin, revenue, and labor productivity. Business cases for HR programs should be built on ROI, not just cost savings. Learn to use Activity Based Costing (ABC) to trace true costs to activities, for example cost to hire per qualified applicant, cost per resolved service ticket, cost to serve per employee transaction. ABC clarifies where process redesign or automation will pay back.

A Practical Model of Partnership: Standard Chartered

Standard Chartered offers a helpful reference. Group HR leader Tanuj Kapilashrami has repeatedly emphasised commercial and data acumen across HR, including ensuring teams can read financials. This created credibility for HR in business trade-offs. In parallel, Steve Scott, the bank’s People Insight and Analytics head, has described how HR leadership and people analytics work together to make HR more data driven, with a focus on behaviours, shared language, and putting people data into everyday decisions. Their teams have also shared methods to partner with HRBPs and democratise insights so non-specialists can act.  

You can learn more about them here and here.

Where ChatGPT Fits in Predictive Analytics

ChatGPT and other LLMs can accelerate the work of People Analytics in many ways. Her are a few:

• Hypothesis generation: translate a business question into many testable hypotheses and feature ideas.

• Method guidance: choose the right model family (time series vs classification) and get a step-by-step plan.

• Excel help: generate forecasting formulas, regression setups, and sensitivity tables your team can run today.

• Interpretation: draft non-technical explanations, risk notes, and executive summaries with alternatives and caveats.

• Scenario planning: outline what-if narratives tied to numbers; validate them in your BI tool.

• Calculations: Ask the model to run a 5,000x simulation of sales outcomes or attendance forecasts. 

• Visualization: Request for step charts or other tools to help understand and explain the data

• Produce reports: Develop insights and CEO/CFO ready summaries 

• Privacy guardrail: use synthetic or aggregated data; avoid pasting personal data into prompts.

Do This Within This Month

• Pick one HR decision that matters this quarter, for example store service levels, time to fill, or pipeline coverage. One that has been difficult to solve and that directly impacts business results. 

• Write at least three to five hypotheses that can help explain why this is happening or what can be done to fix it from an HR perspective. 

• Identify the data you will need to test these hypotheses. Make sure it is clean and ready to use.

• Select the statistical method that will be required to test these (it might be as simple as calculating percentages!).

• Build a model in Excel or your BI tool

• Run your analyses to determine which hypotheses hold, and the impact they have on the business results you seek to improve.

• State a recommendation with an expected business effect and a people effect, including a finance view.

• Prepare a short before and after presentation (using visuals!) so that leaders see the link between the analysis and the outcome 

• Set a review date in 30 days to compare predicted vs actual and refine.

Adoption KPIs to Track

• KPI 1: % of HR decisions with a stated hypothesis, evidence, and recommended action.

• KPI 2: Ratio of predictive to descriptive artifacts produced.

• KPI 3: Estimated $ impact from actions taken (rolling 90 days).

Bottom Line

Data management and governance should sit within the people analytics team. However, Analytics is everyone’s job in HR. HR business partners, COEs, shared services, and leaders should bring evidence to every important decision. The business leaders already expect it, most CHROs understand that people analytics is central to HR’s mission, and the business and people outcomes will improve as a result.

Choose one decision, move it from descriptive to predictive, and make the choice visible. Build your statistics and finance fluency, including forecasting and ABC costing. Challenge the false “people versus numbers” mindset. The organisations that combine empathy with evidence will retain better, serve customers better, and grow faster. That is the standard we should set for ourselves.

Evidence that CHROs emphasise people analytics and that role-modelling drives HR adoption is from Insight222’s research on upskilling HR in data literacy.