Every year, companies spend real money asking their employees what they want.
Engagement surveys. Pulse checks. Quarterly feedback forms. Annual “state of the company” town halls where someone asks the room to be honest and everyone stares at the floor.
And every year, HR gets the results back and is handed an enormous task: synthesize what hundreds of people said, find the signal in the noise, and translate it into something the company can actually do. With a real budget. On a real timeline.
I’ve sat on both sides of this conversation — as a CFO trying to allocate limited resources across a growing company, and now as a CEO watching HR leaders navigate it every day. After 20+ years in finance and operations, I have deep respect for what HR is being asked to do here. It’s genuinely hard work.
But I’ve also noticed something that doesn’t get talked about enough:
The process HR is working within has a structural problem that no amount of good work can fix: the data going in was always incomplete.
The survey process is broken — but not for the reason you think
Most people, when they talk about fixing engagement surveys, focus on the operations:
Ask better questions. Run surveys more frequently. Use a better tool. Slice the data differently.
These are fine suggestions. But they miss the real problem.
The engagement survey has a structural flaw that no amount of better questions can fix:
It only captures the voice of the people who are already there — and only the ones willing to speak up.
HR does exactly what makes sense with that data. They compile responses, look for themes, group feedback into categories, and try to identify the issues that appear most frequently or most urgently. That’s sound analytical thinking. That’s good HR work.
The problem isn’t the analysis. But good analysis of incomplete data is still incomplete.
It can’t account for the employees who filled out the survey but weren’t fully honest because they didn’t feel safe being candid, or because they didn’t want to seem like complainers, or because they’d already mentally checked out and didn’t think their feedback would lead anywhere anyway.
It can’t account for the employees who didn’t fill it out at all — the ones who are too busy, too burned out, or too disengaged to participate in a process they’ve seen yield little change in previous years.
And it definitely can’t account for the employees whose needs are real and urgent but don’t come through as “dire” in how they phrase things. Someone managing a chronic health condition, navigating perimenopause, or quietly struggling with a caregiving situation at home may describe their needs in measured, professional language on a survey form. That doesn’t make their need any less significant, but it does mean it might not get flagged as a priority when HR is trying to triage a hundred different responses.
HR ends up making decisions based on the loudest, most consistent signals in the data. That’s a reasonable approach given the constraints. But it means entire categories of real employee needs go unaddressed because the process simply never surfaced them in the first place.
Why engagement survey results rarely lead to relevant changes
Here’s what typically happens after an engagement survey lands:
- Step 1: Data comes back. HR or an external analyst spends weeks organizing it into something presentable.
- Step 2: Leadership reviews the report. Notes the themes. Acknowledges that employees want things.
- Step 3: HR is asked to “action the feedback.” Which means: find solutions for a diverse set of needs using a budget that hasn’t grown, on a timeline that doesn’t leave much room for nuance.
- Step 4: One or two benefits get added or updated, usually the ones that showed up most consistently in the data or had the clearest vendor solution available. It’s a reasonable call based on the information HR had.
- Step 5: The next engagement survey goes out. Scores have moved, but not as much as leadership hoped. Some employees feel heard. Many don’t. Leadership wonders what went wrong.
And there’s another problem hiding within these steps: ask employees what benefits they want and they’ll say all of the above — and then add five more. Every time. Like clockwork.
So the list keeps growing with each survey cycle. The budget can’t keep up. And because each new benefit is narrowly scoped to whoever asked loudest, utilization is low across the board. You end up with more programs, more vendors, more administration, and employees who still don’t feel like their actual needs are being met.
I’m not sharing this to be critical of anyone in that process. I’ve been the CFO sitting in that Step 3 meeting, genuinely trying to figure out how to use a finite budget to address what looked like an infinite list of needs. Everyone in that room is trying to do right by their people.
The challenge is that the framework itself sets HR up to fall short because the input they’re working from was incomplete before they ever started.
The deeper problem: you’re missing entire populations
Let me be specific about who is absent from your engagement survey data, because I think this part gets glossed over.
The candidates you didn’t close.
Every offer you’ve extended to a candidate who ultimately went elsewhere represents a person who evaluated your total compensation and benefits package and decided it wasn’t enough. You have no survey data on them. You might have a line in your ATS that says “accepted competing offer,” but you don’t have the detail. You don’t know what tipped it. Benefits? Base comp? Flexibility? Something else entirely?
Your people strategy is being shaped without that input, and that means you’ll keep investing in benefits that attract the employees you already have while losing the candidates you’re trying to bring in.
The employees who left in the last 12 months.
Exit interviews are notoriously unreliable. Most people don’t say what’s actually true — not because they’re dishonest, but because “don’t burn bridges” is standard career advice. It’s what their mentors told them. It’s what they’ve read in every career guide. So they give you a safe, professional answer, and the real reason walks out the door with them.
At one of the companies where I was CFO, we lost a cluster of employees in the same department over eight months. Smart people, good performers, all around the same life stage. It wasn’t until much later that we connected the dots: they were all in their mid-thirties, starting families, and we had nothing in our benefits package that acknowledged that reality. We’d been so focused on the perks that attracted people in their twenties — the fun stuff, the culture stuff — that we’d missed the shift happening right in front of us.
HR wasn’t wrong to prioritize what the data showed. The data just didn’t show us what we most needed to see.
The employees whose needs are changing right now, silently.
This is the one that stays with me. Your current employees are not static. They’re going through things. Their health is changing. Their family situation is changing. What they needed from you two years ago is not what they need today.
An engagement survey gives you a moment in time. A snapshot. It doesn’t capture slow drift. It doesn’t capture the employee who is quietly burning out because the benefits that sounded great in the offer letter don’t map to anything she actually needs in her current life — and who doesn’t describe that on a survey form because she’s not sure how to, or because it feels too personal, or because she’s not even fully aware yet of what’s changed.
What I learned from a pair of headphones
I started Compt because of a moment I still think about.
Years ago, at a company where I was CFO, I decided to do something I was confident the team would love. We were growing fast, culture was good, but the office was loud and I wanted to give everyone something tangible. I ordered beautiful wireless Bose headphones for the entire team: white, premium, the kind of thing employees usually get excited about.
And then I got an earful.
Turns out people have strong opinions about headphones. Some swore by Sony. Others only used Apple. Someone pointed out that white headphones show every mark and they’d never buy white. A few people didn’t use over-ear headphones at all.
I had made the best decision I could with the information I had. I looked around the room, made a judgment call about what would resonate broadly, and acted on it. That’s exactly what HR is doing every time they act on engagement survey data.
The problem wasn’t my intention or my process. It was that I was solving for a version of my team that didn’t fully exist — and by the time I realized it, the very expensive headphones were already on everyone’s desks.
The lesson I took from that experience and everything that came after it is this: you cannot design a benefits program for a diverse workforce based on data that only reflects part of the picture, no matter how carefully you analyze what you have.
The workforce isn’t one person. It’s not even one cohort. It’s people at wildly different life stages, with wildly different health situations, wildly different financial pressures, and wildly different definitions of what it means to feel supported by an employer.
Some of them are training for half marathons. Some of them are going through fertility treatments. Some of them are managing aging parents. Some of them are navigating symptoms that have been dismissed by their doctors for years, wondering if their employer even knows this is something that affects half the population.
A static benefits program — even a generous one — cannot serve all of them well. A one-size program and a diverse workforce are simply in conflict with each other by design.
What happens when listening actually leads somewhere
Here’s where I think the real opportunity lives.
The engagement survey is not going away. It shouldn’t. Asking employees what they think and what they need is the right instinct, and HR’s effort to listen systematically is valuable. The question is what happens next.
Right now, the survey generates data, the data generates a report, the report generates a meeting, the meeting generates an action item, and two quarters later something tangentially related gets launched. HR did everything right. The gap between what employees said and what actually changed is still enormous — and employees feel that gap even when they can’t name it.
What if instead, the output of that survey could be directly connected to flexible, individual support that didn’t require HR to find the one perfect benefit that serves everyone?
What if listening to employees actually led somewhere specific and tangible for each of them, not just to a new vendor relationship that helps some small subset of the team?
This is not a critique of how HR does its job. It’s a critique of the infrastructure HR has been given to do it.
The best HR leaders I know are already thinking this way. They’re not looking for more perks to manage or more surveys to run. They’re looking for a way to turn what employees are telling them into something real — for all of them, not just the loudest voices in the data.
That gap between listening and acting is where trust either gets built or eroded. And closing it is one of the most important things a company can do right now.
