In today's data-driven world, the importance of the right data and the correct analysis cannot be overstated. This holds especially true to compensation data, a topic that directly impacts the lives of individuals and the success of businesses. In fact the significance of accurate compensation data analysis is highlighted by incidents such as the recent Google salary data breach and the subsequent explanations of the structure shown — many of which were inaccurate.
In cases like the Google oopsie, most of the online analysis after the leak were pointing to a pay gap — which isn't to say a pay gap isn’t likely present. However, this case showed how using the wrong compensation data, or not understanding the structure or type of data, makes it harder to identify the real reasons behind these gaps. Let's take a step back to learn why the correct analysis and the correct compensation data is crucial, as well as what the most common data analysis mistakes are.
Compensation: More Than Just Numbers
Compensation isn't just about numbers; it shapes people's lives, affecting what they can afford for essentials like food, housing, and healthcare. It also has a critical impact on businesses. The ability to hire and keep the right talent at the right cost can determine a company's future success. When companies try to “cut costs”, they often target their most expensive asset – their employees. This is why companies need not only to find the right talent at the right cost, but also to create a flexible plan that can adapt to changing business needs, potentially avoiding layoffs.
Not to mention, misaligned incentives can also lead companies down a rocky path, allowing competitors to poach your most talented individuals, ultimately damaging your company's ability to compete and thrive in the industry. In essence, how you attract, inspire, and retain talent is absolutely crucial.
Common Data Analysis Mistakes in Compensation
1. Misunderstanding Survey Submission
One common mistake is not understanding how compensation surveys are submitted.
Most surveys invest a lot of time and resources into collecting and cleaning the data, and in particular matching similar roles to one another. It’s important to understand their process because if they do not match accurately across roles, levels and locations, it will drastically impact the data – and therefore your bottom lines.
It's important to know how surveys aggregate the data even if it’s accurately matched. For example: in a survey, you might have columns for Salary, Bonus, and Total Cash. However, if you sum Salary and Bonus, in most surveys it won't match the Total Cash column. This happens because of how companies provide their data and how the survey represents that same data in aggregate, e.g. some companies will have a bonus while others won’t, but if both submit to the survey, it can skew the amounts.
2. Lack of Data Context
Failing to consider how many data points are tied to each role and compensation element is another common error.
Let's say you purchase a survey that contains data from 10 companies. However you notice that, for a particular role, only 3 companies have provided data, with a total of 5 employees. In this situation, you'll have to decide if this data is sufficient for your needs. This doesn’t mean figuring out an exact statistical significance in all cases, but more agreeing on what is practical for your business. For example, having just 3 data points for CEO pay may be acceptable as each company typically has only one CEO. However, if a tech company provides only 3 data points for a software engineering role, you might want to consider using additional data sources.
3. Equity Reporting Confusion
Equity is commonly reported in terms of dollar value or as a percentage, sometimes even as an annual dollar amount. This is an important context, especially since the value of equity is highly dependent on the stage and valuation of a company. For example, a 1% equity grant at a $10 million company is very different from 1% at a $20 million company. Which means, if you pull a percentage from a survey, you may not have a clear picture of that equity value.
To get even more complicated, when a company goes through a funding round, such as raising $10-20 million, it might be considered a seed round or a Series A. In cases like this, determining the true value of equity can become even trickier when you start to filter down. Not to mention, many surveys are unable to split equity types apart. A compensation analyst will tell you the importance of knowing if something was a new hire, promotion, or equity grant; yet most surveys will aggregate these together. This can drastically increase the equity values and lead you to over granting equity pools.
It gets even more complicated when we add publicly traded companies to the mix. Public companies face an added set of challenges when they issue dollar-value shares and compete against non-public equity that is illiquid.
4. Neglecting Internal Employee Data
Sometimes what you were looking for was right in front of you all along. No, this isn't just a rom-com scenario, it can also be true with compensation data.
Many organizations overlook the value of their own employee data. It can come from existing employees, new hire offers, previous hires and even counteroffers. This internal data can provide valuable insights that external surveys just can't.
The reason the data can be more valuable is because internal company data doesn’t have to be handled the same as a survey. Surveys are required to anonymize data to a certain extent to comply with antitrust laws. However, companies can pull and track their own data much more granularly, which can be even more useful than a survey when it comes to going against your talent competitors.
One way to think of this is that surveys contain hundreds if not thousands of employees you would never hire. However, for candidates and employees, they already passed your hiring bars to some degree, and in some cases are already performing in your organization. This can make that data more relevant to your organization.
The Consequences of Analyzing the Wrong Data
Analyzing the wrong compensation data can lead to a domino effect from hell.
If your compensation ranges feel off, your leaders and recruiters may stop trusting them and using them. When this happens, it could lead to talent loss or recruitment difficulties, or pay disparities. Additionally, inaccurate compensation data can cause budgeting problems for your finance team if you don’t have a consistent pay approach. Not to mention, when your compensation plans aren't motivating, it can affect employee engagement and performance.
Advice for Accurate Compensation Data Analysis
1. Get really good at Excel
There are online tutorials for literally everything, including Excel. When you know what you want from your data and how to use Excel for it, you'll see a remarkable improvement in your data quality.
This is my favorite Excel resource to share.
2. Become a Data Detective
Use clues and trace data back to its source, including how it was collected, and how it was aggregated.
3. Treat Data as a Signal, Not a Silver Bullet
Understand that data provides direction, not instant solutions. It's a valuable compass for decision-making.
4. Treasure Your Internal Compensation Data
Your data is a goldmine. Don't minimize your team's internal resources and insights!
5. Create a Pay-for-Performance Strategy
Having data is one thing; how to use it is another. Don’t neglect a strategy.
Measure Twice, Cut Once
As you can see, there are some pretty common data slip-ups that can seriously influence our analysis and decision-making. Always keep in mind that handling your data carefully is crucial, as it can have a significant impact on both your team and your business.
Comments