When salary benchmarks miss the mark

Better compensation decisions start with current benchmarks that reflect the right role, market, and level of responsibility.

AEC firms don't need much convincing that salary benchmarking is important. According to SNZweig’s 2026 Workforce Policies + Practices Report, 89% already benchmark salaries using industry data or other sources – but having data and using it effectively are two different things.

Salary data is only useful if you're comparing employees with similar roles, responsibilities, and experience. Geography, discipline, job level, and firm characteristics can also influence what constitutes competitive pay. Ignore too many of those factors and a benchmark can create a false sense of precision, potentially leading firms to overpay for some positions while falling behind the market for others.

Here are four common mistakes that can undermine an otherwise sound benchmarking process:

1. Relying too heavily on job titles. Job titles aren't always consistent from one AEC firm to another. A project manager at one firm might oversee a small team and a handful of projects, while someone with the same title elsewhere manages major client relationships, larger teams, and more complex work. Firms should look past the title and consider what the employee actually does, how experienced they are, and which employers compete for their skills.

2. Overlooking geographic differences. AEC salary markets aren't uniform across the country, or even within large regions. According to SNZweig’s compensation data, the median base salary for an architect project manager is about $100,000 in Central Florida, compared with $137,000 in the New York metro area. These figures describe different labor markets, each with its own salary expectations. A national average or broad regional figure can provide useful context, but relying on it alone could lead a firm to underestimate or overestimate what it takes to compete for talent locally.

3. Overlooking differences in experience and responsibility. Two employees in the same discipline can command very different salaries depending on their experience, licensure, responsibilities, and client relationships. Comparing both against one broad occupational average can hide those differences. A general salary range might suggest an experienced technical professional is appropriately paid while a more closely matched comparison reveals the firm is falling behind the market. Even with a well-matched benchmark, firms still need to weigh individual performance, internal equity, firm economics, and career progression before deciding what to pay an employee.

4. Using outdated salary data. Salaries change as firms compete for talent, recruiting pressures shift, and demand grows for particular skills and experience levels. A benchmark that accurately described the market two or three years ago may no longer reflect what employees can command today. By the time a firm realizes its salaries have fallen behind, it may already be struggling to recruit or retain employees. Regularly updating benchmarks helps leaders identify potential gaps and adjust their pay strategies before they become bigger problems.

Before using a benchmark, firms should ask whether the data is current, whether the position is genuinely comparable, and whether geography and experience are appropriately represented.

It's not enough to simply know what other firms are paying. You need to understand what competitive compensation looks like for the specific positions and markets that matter to your firm.

That understanding depends on accurate, detailed, and current industry data. By participating in SNZweig’s compensation benchmarking, AEC firms contribute to a clearer picture of industry pay trends while gaining information they can put to work in their own compensation decisions. Firms that participate before the next participation window closes on November 17 are also eligible for a 50% discount on one of SNZweig's compensation products, a savings of up to $795.

Benchmark your firm’s compensation data and save 50%.