Hi, welcome back. Data can be extremely helpful in figuring out what would work. If you're lucky to have a lot of internal compensation data and performance data at individual, team and organizational levels, you can conduct analyses of various complexity. This analysis, as always, starts with the right questions. For example, what is the relationship between the performance-based portion of compensation, and performance? How did the recent changes in the compensation model affect performance? If we increased commissions, would our salespeople sell more? What is the effect of stock option plans on employee retention and performance, depending on vesting schedules, percentages relative to employees' fixed compensation, levels of hierarchy, length of tenure, and so on. And obviously you have to keep an eye for confounding variables that may explain the correlations, and above all you have to be careful not to interpret correlations as causation. While your internal data can help you understand how your incentive systems affect performance, external compensation data can be quite useful to benchmark compensation levels. Quantitative research can help here. The good news, there are many sources of data that you can use to benchmark salaries. The bad news, you still have to calculate, or this is good news if you like math. There are three major sources. Traditional surveys is one. They are published by the government, associations and consultants. They are usually very general and give you an aggregate view and may be somewhat stale depending on the location where you are at. This website, for example, gives you a wealth of information on all sorts of jobs tracked by the US Department of Labor. Most of the information is dated 2014 and you can get state-level data. Such sources are the cheapest and arguably the fastest method to get salary benchmarking. The second source is online data from sites like Payscale and Glassdoor. In places with strong government data transparency practices, you can find public sector employee salary data openly available. Another source of data, particularly for traded companies, is their filings. Websites like Bloomberg aggregate this data, mostly on executives. The best part about this data is that it gives you information on specific people, anonymous, identifiable, or identified, but not averages. The third source of compensation data is custom surveys, which you can conduct on your own or have a consultant do it for you. Custom surveys entail contacting competitor or similar companies to learn about compensation levels for the jobs you are interested in. The strength of this method is being able to tailor it to your task. Regarding of which source of data you will pick, you will have to go through the following steps. Identify the jobs that you are benchmarking. Concentrate on job substance rather than titles. A salesperson may be called client director in one company, a sales executive in another and a grand master of customer happiness in yet another one. It's critical to understand that you're comparing green apples with green apples. Figure out which sources would fit your purposes best. Do you need very specific data, or just ranges would do? Are you analyzing particular job market which is very different from others? Is this a unique role, or a very commoditized position? Are your competitors big companies with publicly available data? These questions can help you pick the right sources. Weight your data. If you're doing a salary survey and one of your responders provided information on three employees and another on ten, taking simple averages may not be a good idea. The same applies to different sources. How many salary reviews was your Glassdoor figure based? How much do you trust the numbers you got from salary.com? By weighting, you are basically deciding how much influence a particular source will have on your final results. Compensation is a critical topic which is hard to avoid when talking about human capital. But as we've seen, it is very hard to generalize. Eventually, it all boils down to your unique situation. But unfortunately, reward systems are seemingly the easiest to manipulate. In the world where conventional wisdom is highly influenced by the concept of the rational economic man, consultants and managers often resort to fixing compensation models to resolve problems that lie much deeper. As Jeffrey Pfeffer put it, it's simpler for managers to tinker with the compensation system than to change an organization's culture. Today with so much internal data one has access to, one could gather it's important to look at compensation and performance relationship without forgetting a myriad of other variables.
