Spreadsheet Purgatory: How to Start Measuring What Matters
👋 This is a free edition of the MPL Build Newsletter—resources for running your people team like a product team. Become a paid member to get access to exclusive guides, templates, and case studies.
Haven’t we all been there? Sat in a meeting where someone shows us a Google-Sheets dashboard with 47 metrics on it. Data painstakingly scraped from the HRIS and ATS and CRM and SaaS dashboards they were paying for, and crammed into our $16.99 Google Sheets subscription with pie charts (I die) and pivot tables. It has tabs. It has drill-downs. It has year-over-year comparisons colour-coded by severity.
Whenever I see a dashboard like this proudly shared by another People or Ops leader, I ask them what decision they’ve made differently because of it.
It’s not entirely uncommon that what I get back is silence.
Then: “Well, it gives us visibility into the business.”
Right. Visibility. That thing we say when we mean…I spent three weeks building this and I’m not entirely sure why.
This is what I find happens when HR Gets Serious about being Data-Driven. We build monuments to measurement. We track everything we can track, which turns out to be quite a lot when you have an HRIS and a bit of extra time. We create dashboards that require their own dashboards to explain them. We send out weekly reports that four people open but folks rarely read.
And somewhere in all of this (buried under the pivot tables and the segmentation analysis and the conversations about whether we should track eNPS or do pulse surveys) we forget to ask the one question that actually matters: What are we going to do differently because of this information?
The Dashboard Factory
Your team decides to Get Serious about Data Based Decisions. Someone (often someone who just came back from a conference, or - and this hurts to type - read my last book) declares that this year, we’re going to be Data-Driven. Seriously. Seriously, team, we’re going to do it this time, we’re going to finally do it. Everyone nods enthusiastically. Yes, Data-Driven. Finally. We’re doing it.
So you start measuring. Headcount by department. Attrition by tenure. Time-to-hire. Offer acceptance rates. Engagement scores. Survey completion rates. Training hours. Performance rating distributions. Promotion rates. Span of control. You measure sick days taken and sick days not taken (both concerning, apparently).
The metrics multiply. You build dashboards. You schedule reports. You export to PDF. You set up alerts for when things hit certain thresholds. Red for bad, amber for concerning, green for good. You refresh them weekly, daily, in some cases hourly if you know SQL.
And then what?
Because this is where most teams get stuck. You have all this data. You have beautiful visualisations. You have trend lines and benchmarks and comparative analyses. You know, with mathematical precision, that June attrition in your product team is 23% higher than the company average was in 2023 annualised.
But you don’t know why Sarah left. You don’t know that Devinah asked for more interesting projects three times and was told “maybe next quarter.” You don’t know that Maria watched two peers get promoted for work she’d been doing for years, and nobody could explain why she was overlooked except with vague language about “leadership presence.”
The dashboard shows you the shape of the problem. It cannot show you the problem itself.
What We’re Actually Doing
When you say you want to be data-driven, what you’re really saying is: I want certainty. I want proof. I want to show my CFO a chart that makes the ROI undeniable. I want insurance against being wrong.
This is human. This is understandable. This is also impossible.
Because here’s what you’re actually building when you successfully build a data-driven People Ops function: a system that treats data as a map first and a compass second. You should use your data to inform your movement, head in the right direction, speak to people, investigate.
It’s easy to feel overrun with metrics. And somewhere in all of this, we lose sight of the people we’re meant to be serving, because we are looking for visibility. Keeping a pulse on things. Being informed.
What’s Actually Useful
This doesn’t mean data is useless. Data is useful. Data can validate your hypothesis. Data can challenge your assumptions. Data can help you make the case for investment or identify patterns you wouldn’t see otherwise.
But data should serve the work, not become the work.
The most useful People Ops metrics I’ve seen are ruthlessly focused, tracking 5-8 things that directly connect to business outcomes. They’re reviewed regularly to inform where to go next, what to look into, what is trending in a way that requires further validation or context. Each metric exists because there is a specific decision it informs.
Generally, I like to start with a quite light and high level view of the “health” of the business, and then when there is a specific project or problem, we can then create a more nuanced view which may include user research or success metrics.
Regrettable attrition. I calculate regrettable attrition by asking three questions:
Did we try to retain this person through their exit?
Would we hire again in the future?
Do we have any regrets with how this exit was handled? Meaning, did we miss a critical opportunity to give feedback, was the exit volatile or disgruntled?
If “Yes” to two of the above, the exit is regrettable. If “Yes” to all three, then the individual is highly regrettable.
Time-to-productivity for new hires, defined as first meaningful contribution to something measurable (first sale, first OKR, first project managed)
Manager effectiveness scores (from direct reports, measuring specific behaviours, using RANDS if you can)
Internal mobility rate (people moving to new roles) and Internal Promotion Retention Rate
Offer acceptance rate (split into cohorts if you can)
Performance scores and goal achievement. Where are we seeing performance, ideally viewed by tenure and team.
Each of these are tied to a specific focus area the business cares about. Just a handful of numbers that matter as a high level view, tracked consistently, with real conversations about what they mean and what to do about them.
The rest of my time? Talking to people. Running interviews and 1:1s. Sitting in team meetings. Watching how managers actually manage. Listening to what employees actually needed.
Because that’s where the real work happens. Not in the dashboard. In the conversations.
A world where we measure what matters
Measure the things that matter. Not the things that are easy to measure, but the things that connect to outcomes you care about. Resist the urge to measure everything just because you can.
Build dashboards with purpose. Each metric should answer a specific question or inform a specific decision.
Talk to people. Not surveys sent via email at 3pm on Friday, but actual conversations. Sit with managers. Run 1:1s, interviews and exit interviews and mid-tenure interviews. Create ways to hear from employees that don’t require them to rate their experience from 1-5.
Combine quantitative with qualitative. Use your data to identify teams worth investigating. Then investigate them properly. Use your metrics as signposts, not solutions.
Accept that ambiguity is unavoidable. Some decisions will be made with imperfect information. You will sometimes be wrong. That’s not a moral failing, that’s a feature of working with humans.
The much laboured ‘seat at the table’ doesn’t come from better dashboards, it comes from solving real problems the business cares about. When a critical team is bleeding talent, leadership doesn’t care about your exit survey breakdown, they care that you understand the candid reality of why people are leaving and have a plan to fix it, that you are ready to understand what will demonstrate the leading indicators that it is trending in a better direction.
The best People leaders I know spend maybe 15% of their time looking at dashboards and 80% of their time in rooms with actual humans. They use data to inform their thinking. But they don’t mistake the dashboard for the work.
Because when someone tells you they’re leaving, the number in your attrition dashboard is not the story. The story is what happened six months ago when they stopped speaking up in meetings. The story is the manager who never asked about their career goals. The story is the accumulation of small disappointments that turned into a resignation letter.
So measure what matters. Track your trends. Build your business cases with proper data. But spend most of your time where the actual work happens: in conversations, in observation, in the messy work of understanding what people need and figuring out how to help them get it.