Absolutely Accurate and Completely Wrong
The world of compensation is filled with odd inconsistencies. Are you motivating someone to retain them, or are you retaining people, so you can motivate them? Are you looking to hire “world-class talent”, or do you want to pay more like your peers? Perhaps the most frustrating is the dilemma of compensation data being absolutely accurate and almost completely wrong.
Now, before you go into a tizzy, I know that pay data for many positions is pretty darned close to perfect. If you want to know how much to pay an average accountant at an average company, the data will be pretty close. But if you want to hire an extraordinary engineer at an unusual company, you may find yourself blending data, creating composites, extrapolating levels that do not exist in the data and applying premiums that change the data from science to art.
Nearly everyone has had to do this. Some compensation professionals spend far more time on art than science. And yet, the final product is handled as if it is just as fact-based as the data coming directly from a survey. And, of course, even pure survey data is suspect. Did that other person apply the correct job code? Do they judge talent and measure experience the same way you do? (Or at all?) Do they actually understand the workings of their incentive plans and report on them properly? Your guess is as good as mine.
Increasingly, many companies have embraced “market pricing.” It gives them the feeling of moving away from artful pay structures and artificial restrictions. Market pricing seems relatively easy, but don’t be fooled. Job descriptions can remain semi-formal. There is no need to build an internal Job Worth Hierarchy. Inconsistencies base pay are balanced by a bigger focus on variable pay. Data is plentiful and not too expensive. But the long-term results often look like a crazy person set pay for everyone. Art is often in the eye of the beholder. Market pricing is often like painting by number, but you never use the same color twice.
The process and rational of “old-school” approaches to pay seem to make sense. There is an order and structure that makes detail-oriented people feel more comfortable. But many of these rules were create for a different age. Only 10 years ago the world of business was much slower than it is today. Most of our legacy process were created 30 or more years ago! Three decades ago email was basically science fiction. Spreadsheets were still something used only by super-geeks. Almost nothing we used in 1988 is still useful today, but, for many companies (and some of their survey providers and consultants), compensation processes have barely changed.
We need dispel the myth that pay data is specifically accurate. It’s just silly to express pay down to single dollars (ex. $54,327) because that’s what is shown in your data source. We also need to agree that market pricing is only effective if there is a functional framework to support and verify decisions with thin or suspect data. We are speeding toward a day when surveys that report on data several months old will be a comical concept. We need a new approach.
In theory the data will be more accurate than ever, and perhaps even more incorrect for a specific purpose. A recent (great) guest post by Joe Thompson touches upon the impact Artificial Intelligence may have on our profession. I am excited. I personally think the data will be amazing, but, like so many other advances, it may not mean better decisions. To succeed in the next quarter-century, we will need to become even more comfortable being both scientists and artists. How have you begun to future-proof your approach to pay?