Yeah, we’re nowhere near fixing HR Analytics

Or — more simply — we need to stop congratulating ourselves until we figure out performance assessment data.

Jessica Zwaan
Incompass Labs
9 min readOct 1, 2023

--

If you’re the kind of person who looks at HR strategy documents, something I’ve been known to do, you’ll see hundreds of the same. Groundhog day with exuberant copywriting. “Here at Sprookle, our team will: Nurture the optimal culture. Hire the best people. Facilitate, no — demand!, high performance!

Almost all of these strategy documents I’ve seen (and built myself, god help me) share a nebulous and terrifying homogeneity: describing some egalitarian, productivity paradise — if only we had another tool, more headcount — a better annual survey! Faithfully, we run into the fray ready to implement the tools that will tell us how we can fix everything — ideally by end of financial year, ready for the new wave of hiring.

We’re only one annual survey away from Utopia, guys!

Within most of these plans is the central quest to implement the perfect end-to-end performance process. The keystone to the whole shebang. We’ll hire the best because we can, you know, tell who they are. We’ll keep them in the team because, umm — we’ll have all the right apps and forms telling their manager they’re killing it! And then… well, then I guess we’ll sit in a calibration meeting for a few days, arguing about whether they really are (because you know what Louisa’s team is like; they all just say they’re exceeding expectations).

Rubbish in, rubbish out.

People ask me why I have for so long advocated for traditional, spreadsheet-driven performance calibrations while your size allows. Simple, but not exactly revolutionary. I get it — while they get the job done, they lack the finesse of modern tools.

The performance assessment offering suffers from self-reinforcing pessimism (mine included). Organizations have struggled with performance assessments for years, and promised solutions have often fallen short, leading to a collective eye-roll at the mention of a new tool.

Many modern performance and engagement tools, those digital wizards promising to streamline employee evaluations, have come a long way. They’ve evolved from rigid, paper-based forms to sophisticated software platforms. Yet, much like the university lecturer trying to figure out BlackBoard in April 2020, many managers find themselves disengaged with these tools, extracting little of their value and using a google-doc instead.

Actual footage.

The platforms promise to transform performance assessments into a delightful experience, something we’ll want to use. They offer features that make your head spin — 360-degree feedback, competency ladders, engagement surveys, AI-driven L&D, real-time goal-tracking!

It’s a story that’s been told a thousand times, and it goes something like this:

  1. Buy a tool that covers every base.
  2. Convince yourself that this tool will magically solve all performance-related issues.
  3. Pitch it as a game-changer for your business, promising efficiency, productivity, and all-around excellence.
  4. Wait for the performance problems to disappear into oblivion, and the profits to roll in. 💸

…Oh. Wait. What do you mean Benji says his team don’t want to do the surveys? Well. Okay, look, just ask them to slack their feedback. No, I don’t think it’s fair that Sandra’s team all use ChatGPT to write their comments, but I guess we can’t stop them, and I suppose they’re at least prompting it (right??)

To get good data is to use this new tool all of the time for all of the things. Literally, please always put data in. Yes, step away from your Sprint Planning and input your qualitative 360 feedback for your team for last quarter thank you.

Despite the offerings, individuals barely scratch the surface of what these tools require. They might use them for data collection when it comes to time for the annual “HR yells at us if we don’t fill these forms in” but then overlook them entirely when it matters most. No one has quite cracked real-time feedback yet, so we’re stuck gathering data once or twice a year (or once a quarter if you really want to bother everyone). The real issue here is that they have to use the tools effectively to get good quality data out of them, and then managers have to generally synthesise these huge quantities of information to actually learn anything meaningful.

Here’s the thing: Performance data is the centre of everything. More important than attrition. More important than engagement. It is the most important thing you can collect in HR analytics, because without it the rest of your data tells you very little. It is as if, in product teams, we collected acquisition data, retention data, and customer engagement — but no one ever thought to ask how much they’re spending. Are they even good customers?

So, in order to work out our value and loss statement for our people investments, we need to know the value part. We need to understand where and why great work is happening, and nothing is quite helping us do that right now.

Yeah we need to revisit this whole thing

It’s a tale that captivates the imagination. “Performance assessment” makes it sound straightforward, even elegant — the solution is to simply complete these few forms, have your manager read them, and then we’ll see where you stand in the grand architecture of a high-performing team, our strategy will be realised — we’re going to be a unicorn in no time. After all, who wouldn’t want a tool that simplifies everything and transforms chaos into order?

But here’s the harsh reality: This story rarely unfolds as planned. For every HR leader that believes this approach will be their golden ticket to the fabled “seat at the table” People Teams long for, there are numerous reasons why it won’t be the panacea they imagine.

Complexity Beyond Measure

Companies are networks of humans. Humans are multifaceted, confusing creatures who can be both exceptionally capable and unbelievably stubborn. More so, we’re distracted by a multitude of tasks every single day, it’s impossible to take a year, or even a quarter of qualitative information and package it up. Believing a tool can capture all of this, and then a single manager synthesise all these complexities is akin to thinking a single wrench can fix an entire car. It’s an oversimplification that ignores the intricate web of inputs that human beings just, like, are. Navigating this landscape can feel more like a mosh pit than a tango.

Jack of All Trades, Master of None

End-to-end HRIS platforms may cover a broad spectrum of areas, but they often lack depth in specific domains. They try to be a jack of all trades but end up as masters of none. In the world of performance, depth matters. A tool that excels in one particular aspect can be far more effective than a one-size-fits-all solution. The platform is usually built around optimising workflows, and rarely around breaking down the barriers of calibration.

The Human Factor

Performance isn’t solely about technology; it’s also about people and processes. No performance tool out there (well, I’m trying, but we’ll see!) has yet solved synthesis or calibration. The data is wonky, loopy, riddled with bias. A group of people have to actually sit down, read it all, and make sure it all makes sense, that it’s fair. If you have a team of fifty it’s almost impossible, let alone fifteen-thousand.

Calibration is, like, really hard man. The way we’re doing things at the moment is something close to broken and we’ve only just really started talking about how this is so important.

Putting another form or survey on top simply won’t guide us closer to the truth unless we somehow change this process entirely.

The Illusion of Control

These tools create the illusion of control. We put bad data in, we calibrate it manually, and then we look at pre-built dashboards to tell us if we’re on track towards our strategy. Sometimes, if you have a really good HR Analyst (and most of us don’t!) we’ll give them the bad data to interpret — nice.

People leaders need to implement something that allows their executive teams to be focussed on the insights, rather than running around gathering the inputs. The time we spend reminding, calibrating, nudging, templating, should be better spent analysing the outputs and making plans for our people.

People Data of the future

For years I’ve been reading the future of HR is people analytics. Literally. Here is a Deloitte article from 2014 telling us we need to, “implement HR data analytics to achieve business goals” as one of their big trends in the space. We’ve been talking about it for so long, so why aren’t we actually doing anything groundbreaking yet?

Well, for one, most of our leaders don’t think like commercial or product leaders. We’re still clinging to the Ulrich model and centres of excellence of 1995, convinced if we build admin teams to de-admin ourselves, we’ll eventually transcend to the strategic. Hint: we won’t.

Our methods for collecting data are similarly stuck in 1995, we’ve taken our paper forms online, but we’re still doing things with the same processes and philosophy. Other functions are using powerful BI tools to analyse trends, test their solutions, inform decisions. While most HR data is bad data.

Are managers capable of calibrating qualitative data and turning it into performance assessments? On this answer we have constructed our entire Rube Goldberg machine of HR analytics. Sure you can analyze pay, age, geographic information, or job details, but what do we have to use this against if we don’t have reliable performance data?

We’re quickly hurtling towards salary transparency, but this exciting momentum leads me very quickly into the intellectual black hole of, “what are we basing all of this on if not just more opacity?”

It’s time to look elsewhere than HR.

This won’t come as a surprise to you, but I think we need to look beyond HR in order to find a solution to this problem. I don’t need another survey tool, I need a market intelligence tool. I don’t need another long form with a nice UX, I need a machine learning model. I want to be able to treat my employee investment with the respect it deserves — heck, it’s about 60% of my company’s monthly spend.

I know it sounds inhuman, but bare with me, it really isn’t. What is actually inhuman is thinking that people can make all this work without bias, fairly, and for the better of us all. Humans should be there to validate insights provided by machines, not supply the inputs to some data visualisation tool. I’m suggesting an adaptation that addresses the fact we are humans, and suggesting we should use computers to give us back time for our incredible abilities to interpret and come up with ideas. Besides, putting feedback in some big form twice a year isn’t exactly more human, we’re just more used to it that way. 🤷‍♀

How are other parts of our business collecting valuable data on the quality of their customers, products, leads? These professions have moved beyond pure subjective assessments and into the future of actual business intelligence, and so must we.

Ok that’s all from me, folks. 👋

👉 Buy my book on Amazon! 👈
I talk plenty more about this way of working, and how to use product management methodologies day-to-day, I’ve been told it’s a good read, but I’m never quite sure.

Check out my LinkedIn
Check out the things I have done/do do
Follow me on twitter: @JessicaMayZwaan

Me and my cat, looking professional

--

--

Jessica Zwaan
Incompass Labs

G’day. 🐨 I am a person and I like to think I am good enough to do it professionally. So that’s what I do.