Best Practices for Conducting a Workplace Pay Equity Analysis

This conversation focuses on how organizations can prepare to do a pay equity analysis, the steps for conducting an analysis and what to do with the results.

Julia Dow has worked with a huge range of organizations doing compensation-related work.  From comp-focused positions at Visa and Raytheon to advising companies like Discord, Reddit, Medium and Figma, Julia has experience with many methodologies used to conduct organizational pay equity analyses.  She was recently interviewed by Gender IDEAL’s Founder & CEO Flory Wilson about how organizations should get started in evaluating their pay equity, what fundamentals need to be in place and how to use the results.  Below is an edited overview of their conversation.

Flory Wilson: Julia, tell us why doing pay equity analyses is so important for an organization?  Are there risks in not doing a pay equity analysis and potential benefits?

Julia Dow: It is important to do regular pay equity analyses because once your organization has the processes in place to strive for equal pay, it can help build employee trust and engagement and it ultimately creates a more productive workforce.  The risks of not doing a pay equity analysis include legal risks, unmotivated workforces that feel a lack of trust with their employer and employees who are unsure if they are getting paid fairly all of which can distract from the work itself and can result in a negative morale around the company.  

The benefit of having a regular cadence of doing pay equity checks is that an organization can anticipate and avoid unexpected costs.  

Flory: You mentioned the cadence of doing these reviews.  What is a regular cadence and how frequently should a company be doing this analysis?

Julia: Most companies opt to do pay equity analyses annually in conjunction with their compensation planning and budgeting process.  They use this as an opportunity to identify if there are any gaps that need to be addressed and they can use the current compensation budget to account for these adjustments.  This also allows them to have smooth and timely conversations about compensation increases with any employees without having to specifically mention that there was a pay equity adjustment. 

Sometimes companies do mid-year review or out-of-cycle promotions, and that can also be a good time to review internal equity to make sure things are balanced as managers are proposing increases as well but doing this annually is a best practice.

Flory: How do organizations get started doing this work?  Who needs to buy-in to the process, what data or systems need to be in place?

Julia: One of the hardest and most important parts of this process is gathering employee data - making sure that data is accurate and that gender and race data is complete.  So getting prepared for the analysis includes checking that your human resources software (HRS) system is accurate and filled out with any other credentials that you want to track for the analysis – this could include  educational degrees attained or years of work experience – are validated.  It also involves making sure that any promotion or performance data that you’re using is included as well.  

The other data that are important to track are the instances when offers have been made outside of normal compensation ranges; having documentation that explains why those exceptions were made is important.  Companies need to be able to  point to why a candidate was hired or promoted at a salary level that is above the typical range with a rationale for why it was a conscious decision that was made.

As far as who is involved in this process, the decision to conduct a pay-equity analysis should come from senior leadership.  Those leaders should care about employees being paid fairly and equitably, so ideally this support exists from the beginning of the process.  Early involvement of leadership is important because there could be a cost involved so having a financial sponsor involved in this work is essential.  You want managers to be involved to give context for why certain employees are paid at different levels and to be involved in communicating any changes in compensation levels and the company’s compensation philosophy to employees.  

Lastly, you need to have legal involvement in this process – either an internal general counsel (GC) or external counsel.  Legal counsel involvement from the beginning of the process is crucial.  We want to make sure the information we (OpenComp or another third-party consultant supporting the analysis) provide is privileged (meaning not subject to disclosure or discovery and cannot be asked about in testimony) so that legal counsel can advise on what remedial steps should be taken if corrective action is needed.  By corrective action, I mean when someone is identified as being underpaid.

Flory: Walk us through the steps of doing a pay equity analysis.  How long does it typically take for an organization to get through the process?

Julia:  Timeline is dependent on data cleanliness, budget and the size of organization.  I’ve seen a pay equity analysis take six weeks from start to finish and I’ve seen it take several months just to get all of a company's data organized.  Once you have all of your data and your employees have been mapped to whatever dataset you’re comparing them against -  whether it’s internal averages or market data - ideally you’ve got data that you’re comparing them to that we call comp ratios or ranges, so that you can look at each employee’s salary data and compare them against the benchmark to identify who is high, who is low,   where the different gender averages lie. 

There are a few different ways to do pay equity analyses. For smaller and mid-size companies, if you don’t have a concentrated number of employees at a specific job and level, you might conduct a cohort analysis, which is a manual comb-through of data where you look at who is below a certain average comp level and then you identify why someone is below or above their peers.

The most legally defensible approach, and one used by larger companies, is a regression analysis.  This is a mathematical approach that takes into account many different factors that generates a trend line that will show whether females or males are below a particular trend line which helps to identify gaps that need to be corrected.

And a mid-range approach is to take an average pay ratio that accounts for compensation variables with factors like scope of role and other variables that influence the range and creates a level playing field.  This establishes what we call an adjusted compensation range, versus unadjusted. Unadjusted would be “all females have X pay and all men have Y pay”.  It doesn’t account for role and experience, it just gives average pay against whatever data source you’re using.  Using the average pay ratio, a company can then identify who is below the average gap and what corrective action should be taken.

The approach a company uses depends on its compensation framework.  Some companies use raw market data to establish pay bands.  Others use external data to inform their compensation ranges to create pay bands leveraging their compensation philosophy.  Either way, you want to look at what is competitive for that role and that level, rather than just taking the average pay for all women and all men, which would omit too much detail.  A company should look at what is competitive for each role at each different level versus just saying what is the average of all female and all male salaries.

Flory: How much data do you need to use the regression analysis approach?

Julia: The regression analysis is used mainly by larger companies with 1000+ employee – you need a good sample size for each role for this method to work.  Smaller companies use a cohort analysis because they don’t have the statistical relevance - the sheer number of workers in each role and at each level.

Flory: Tell us about OpenComp and how it helps companies conduct pay equity analyses.

Julia: Open Comp is first and foremost focused on getting reputable data into the hands of employers so they can make educated decisions on pay.   We want to help organizations use a better approach than “this is what a candidate asked for so that’s what they received”.  We start with a compensation roadmap where we do a competitive analysis looking at all of an organization’s roles and comparing them to market data to understand what percent the company pays at on average for salaries, equity, etc.  After that, we can then help model out different compensation philosophies - so if you want to target the 75 percentile of industry pay you can then shift the costs and understand how close or far your employees are from that target.  Ultimately this should help organizations go to market with offers more consistently, regardless of their gender.   

We also have an equity, diversity and inclusion tool that allows companies to identify gaps in their compensation structures as compared to market averages by gender and ethnicity.  For example, you can see if none of your female employees are getting promoted into more senior levels - you can identify those patterns, overlaying gender and ethnicity onto all data to develop a clear sense of what your internal labor market looks like before you go to the external market.

At Open Comp, we want there to be information symmetry between employers, candidates and employees.  More reputable data will compound and add to the integrity in the market.  The time it would take an organization to benchmark and do the data analysis that we provide at Open Comp is incredibly intensive.  We use machine-learning and AI that does the data-benchmarking for our clients, so we can speed up the process of the analysis so that clients can get to taking corrective actions if needed rather than spending time doing the analysis.  

Open Comp’s sweet spot is private companies that are VC-backed (pre-IPO) because most of these organizations will not have a compensation person in-house.  We see our role as helping the founders and CEOs focus on doing their jobs while knowing their employees are being fairly compensated.

Flory:  So what happens after a pay analysis has been conducted?  What comes next, what does a company do if inconsistencies are identified?

Julia:  Your company should have clarity on what individual compensation adjustments need to be made and can integrate those adjustments into their budget.  It makes the most sense to time this with the budget planning process so that all of the corrections can happen at once.  If a company can not implement full-adjustment increases for all employees who need an adjustment, it can be done over time but an organization  needs to work closely with their counsel to develop that time line.

Flory: What does transparency look like during a pay equity analysis: What’s the best-practice for employers?

Julia: Some employers disclose that a pay equity analysis was completed at the end of the process and share what the results were.  Others just make the necessary adjustments or corrections they need to make without announcing the process to the whole company.  I’d say there isn't really a best practice on this because it is so dependent on the organization and who is involved in the process.  Greater transparency can sometimes raise more questions if shared broadly without a lot of context and training.  We advise companies to train managers on understanding the organization’s compensation philosophy.

Some big companies are really explicit about their results and this tends to be larger companies who have a great headline – we have pay equity! - and want to share that publicly.

Flory: What about for people who have been identified as underpaid? How does the company frame this information for those employees?

Julia:  Most companies just share that the employee will receive a compensation adjustment based on their performance and role.  The company can be transparent that they have conducted a pay equity analysis and identify the adjustment as a pay correction.

Flory: What are the best practices for companies gathering and tracking race and other identity characteristics?  Are companies using race, sexual orientation and other traits in their analyses?

 Julia:  Race data is tracked for EEO (Equal Employment Opportunity) reporting, so most companies should have this information readily available. We also see companies tracking data on veteran status or disability status and that information would be stored in an HRS system.  But often companies have a lot of gaps in this type of data, so that can be a struggle in trying to do a compensation analysis based on these factors.  Sometimes companies just don’t want to push for a lot of this data, sometimes employees don’t want to share it.

The evolution of this work that we typically see is companies start doing this work focusing first on gender.  After doing pay equity through a gender-lens once or twice, a company might then layer race into their analysis.  We haven’t seen too many companies look at other traits such as sexual orientation, but I suspect that will become more common as more companies become more sophisticated working on pay equity issues.  There are still so many companies that haven’t even started this work yet.

Flory: What advice would you give to small companies?  How can they get started doing this work?

Julia:   Ideally a company can compare their employee populations with a reputable market data set.  Comparing their compensation data this way puts into context who is paid fairly and who is underpaid inside of their organization.  Small companies can do a cohort analysis where they can identify information like “this is a group of Level 4 software engineers - why is this one employee an outlier”?  Then you can explore why this outlier exists, negotiate and document pay changes for that employee.  And for compensation negotiations going forward, the organization will have clean data and records of negotiations.  This documentation becomes really important as they grow. For small businesses, in addition to having market data, you want to have clean employee data.  Once an organization starts digging into their data, more questions will emerge.  But having the data organized and understanding how the company has gotten to where it is on compensation will help them as they go forward.

Additional resources:  

Open Comp website

The State of Start-Ups and Pay Equity, Open Comp.

Self-Identification of LGBTQ+ Employees.  The Human Rights Campaign.
Employee Demographics Don’t Have to be at Odds with Employees’ Identities. Harvard Business Review.