One Model Blog

Beyond the Hype: Making AI and People Analytics work for HR

Written by The One Model Team | Jan 26, 2026 6:39:55 PM

From Excel to enterprise AI: How does People Analytics actually scale?
For many HR analytics teams, the journey starts the same way:
A spreadsheet, a pivot table, and a heroic analyst holding it all together.
And for a while...it works.


But as One Model Chief Product Officer, Tony Ashton, joked on stage at Gartner ReImagine Sydney,

“If there’s something wrong in cell A512, that formula is going to screw up your day.”

 

The laugh lands because it’s painfully true. Excel is a powerful tool (even when inserted into PowerBI or Tableau), but it was never designed to scale people analytics across an enterprise. It was not designed to protect sensitive data with permissions. To scale, enterprises need to move from individual effort to systems that support consistent, trusted decisions.

Watch Tony's Full Presentation

A quick note from the video: Tony mentions that the data is kept in Australia. This was true for the audience watching this presentation live in Australia. One Model has servers across the US, Canada, and Ireland as well.


The Familiar Starting Point: Excel and Hero Analysts

Tony’s career mirrors the path many HR analytics leaders have taken.

He started long before pivot tables existed in Excel, building HR analytics in government roles using spreadsheets because that’s what was available. Later, he joined a small analytics firm that eventually became part of SuccessFactors and SAP — moving from a 150-person company to a 100,000-person enterprise almost overnight.

That experience taught a hard lesson:
What works for analysis does not work for scale.

In Excel-based HR analytics:

  • Logic lives in individual files

  • Knowledge lives in people’s heads

  • Risk lives everywhere

When one person leaves, goes on holiday, or updates the wrong formula, confidence collapses. And once leaders stop trusting the numbers, analytics stops influencing decisions. That's true no matter how good the insight is.

 

Why Excel Breaks at Scale in HR Analytics

HR analytics is uniquely difficult to scale because it sits at the intersection of:

  • People’s lives and careers
  • Ethical and merit-based decisions
  • High privacy and security requirements

At enterprise scale, HR leaders ask questions like:

In Excel, every one of those questions becomes a debate — not because the analyst is wrong, but because the system has no single source of truth. Executives don’t argue with dashboards because the insight is uncomfortable;  they argue with the data so they don’t have to deal with the message.
Or as Tony put it:

"If your data is not very clean and very accurate and repeatable, then your executive team is going to nail you and you're going to have a bad day."

 

The Real Shift: From Data Assembly to Insight Generation

Scaling people analytics isn’t just about replacing Excel with another tool. It’s about changing where time and energy are spent. In immature analytics environments, teams spend most of their effort on:

  • Extracting data from multiple systems
  • Cleaning and reconciling inconsistencies
  • Rebuilding the same metrics repeatedly
  • Defending numbers in meetings

In mature environments, that work is handled by the platform. That’s when teams can finally focus on:

  • Interpreting trends
  • Understanding drivers
  • Predicting outcomes
  • Advising leaders on what to do next

This shift, from data assembly to decision support, is what separates analytics that report from analytics that actually change behavior.

Why Stitching Tools (and Teams) Together Doesn’t Work

A common response to scaling pressure is to add more:

  • More tools
  • More dashboards
  • More integrations
  • More analysts

But stitching together technologies and teams creates new problems:

  • Tools need to be connected and data refreshed regularly
  • Metrics are defined differently across tools
  • Lineage is unclear when numbers are questioned
  • Security and access rules become fragile
  • AI layers amplify inconsistencies instead of fixing them

Tony made a critical point:

 "If you want to start using AI and you've got a bunch of really badly managed systems underneath, then you're going to have a really badly managed AI on top of it." 

True scale requires a unified system where:

  • Metrics are defined once and reused everywhere
  • Users can drill from insight to individual records with appropriate permissions
  • Lineage back to source systems is transparent
  • Trust is built into the platform, not patched on later

What HR Analytics Maturity Actually Looks Like

Mature people analytics environments share a few defining traits:

  • Consistency: The same question returns the same answer, every time
  • Traceability: Every number can be explained and validated
  • Accessibility: Leaders don’t need to be analysts to get insights
  • Governance: Security, privacy, and ethics are built in, not bolted on

This is what enables AI, predictive modeling, and self-service analytics to work responsibly — without relying on hero analysts to translate, defend, or fix the outputs.

Scaling Is a Systems Problem, Not a Talent Problem

The biggest myth in HR analytics is that scale comes from hiring smarter people. In reality, scale comes from removing friction (especially in the data orchestration phase):

  • Removing manual reconciliation
  • Removing duplicated logic
  • Removing fragile spreadsheets
  • Removing dependence on individuals

Excel will always have a place, but enterprise people analytics demands systems designed for repeatability, trust, and growth. Or, as Tony’s story shows, the real evolution isn’t from Excel to AI — it’s from individual effort to enterprise capability.

 

Can we show you what enterprise capability looks like?

Let's connect today.