Last night I attended and made some brief remarks at the Chicago Comp+Cocktails event sponsored by Payfactors. Payfactors is doing some interesting things in the world of compensation management, with the overall message of a compensation revolution. I’ll have more to say about that at some point in the future.

My remarks last night were focused on how the compensation function can/should/needs to change. In preparation for the event, and through conversations at the event, I drafted the following about how the comp function needs to evolve. Of course, not all comp functions are identical, and many have already made a transition in mindset and strategy. Yet as I scan across my network and speak with compensation leaders, it seems that we all (including myself) have some work to do.

The comp profession needs to be less about answers and more about questions.

My distaste for “best practices” is known – it’s at the core of why the GetNerdy blog exists. As organizational context becomes ever more defined by the interplay between business strategy, culture, and social cause, we have to approach problems assuming the answer will be an innovation. What works at Company A won’t work the same way at Company B, because those companies are probably more different than similar. We like to focus on other practices within our industry, but we forget about how local market dynamics or distinct employee experiences shape how reward solutions actually make an impact.

As compensation professionals, we need to become experts in solving problems, not providing peer group data and “best practices.” We need to question our assumptions, challenge existing practices, and press for change.

Survey data needs to yield to bigger data.

Advances in data tech show that we can be more precise and unbiased in terms of benchmarking. There’s a method to LinkedIn suggesting to me my market value… Amazon recognizing that my buying a certain two books means they’ll be able to entice me to buy the third… or Google serving up an ad for a data science degree as soon as I search the word tuition. Matching algorithms are among the most mature in the machine learning space. So the day is not far off when the machines look at descriptions of your roles and data available about the people in those roles, and smashes it up against the same from hundreds of companies. The resulting benchmark will be as accurate and likely less biased than a trained compensation professional.

That might be the scary one. It sounds like the compensation analyst job will go away – and maybe it will as we know it today. It’s incumbent upon us to provide value in other ways… like the ideas below.

Benchmarking needs to yield to “valuing productivity.”

With flexible roles, gig economies, and pivotally productive roles like engineers, what others pay on average or the top 25% is less important than how to value the productivity of effort and results. We need to own the conversation about sourcing skills and output, knowing the different forms that can take. Our unit of analysis today is “the employee.” In the future it needs to become “the work.”

This is one area where the compensation technology/consulting ecosystem needs to step up, too. The world of surveys has existed because the business problem has been framed as “market pricing.” But the real business problem is estimating the cost to acquire skills and productivity. We need tools that think this way. When a seasoned software engineer chooses to “retire” but still freelance via Upwork at $60 per hour, she is still part of the market that will go completely uncaptured by traditional measurement. The compensation function needs to understand this dynamic and guide the conversation.

Structures and processes need to yield to stories.

In an era of transparency and progressive expectations, traditional salary management is becoming pointless. Employees don’t care that the range midpoint is $X when they see on Glassdoor that their value is $X+Y. Managers can’t articulate why a 4% raise is good, let alone keep someone with a 2% raise fully engaged in that conversation. We force difficult conversations upon managers, without a clear story to tell. One manager I spoke to recently described the annual compensation review process as “my annual opportunity to disappoint most of my team.” Ouch.

What needs to matter more is how career progression and development align with pay opportunity progression, weaving a holistic story of their employee experience. In an era of pay transparency, we need to take the focus from “where do I stand” to “where am I headed.” The compensation function needs to challenge our assumptions and discover ways to deliver career moments that reward progression and contribution, rather than stress-filled annual salary conversations.

Variable pay needs to be distinguished from incentives.

Okay, time to stop picking on salary management and benchmarking. Let’s close with variable pay. Aon Hewitt has tracked the rise of variable pay spending, and there is no doubt that more companies are using variable pay and pushing money into variable awards. Yet, as Aon Hewitt also notes, we largely don’t find our variable pay programs all that effective.

The root of the issue, in my mind, is that we confuse “variable pay” and “incentives.” We use the terms interchangeably, but they are not the same thing and confusing them leads to disengagement. Variable pay doesn’t work because the incentive structures within them don’t make sense. Often times we don’t need variable pay to create incentives. Variable pay still plays a valuable role in managing costs and funding incentives (sometimes), but likely not how we’ve been doing it for year. We need to be expert diagnosticians of incentives, natural or created, and use that within the variable pay structures the business requires.

* * *

What are still open issues for me are what do the solutions look like, and how do we invest in ourselves to become the talent we need to be. I believe that the core capabilities of problem solving, data sourcing and analysis, and behavioral science are essentials. But what else?

I’d love to hear your thoughts on these hypotheses and ideas. Please comment below, or send me an email.