Towers Watson Salary Guide

Compensation benchmarks that actually matter

The Towers Watson Salary Guide — or what the industry now calls it by its successor name, the Willis Towers Watson Total Rewards Database — is one of those tools you either use or someone else uses while negotiating your raise. I've spent roughly a decade looking at these numbers from both sides of the table, and the gap between what the guide says and what people actually get paid is where most disputes happen. Here is how it works in practice. The guide segments compensation data by geography, industry, company size, job family, and seniority level. A senior engineer in Chicago at a 500-person tech firm will have a completely different reference point than the same title in Raleigh at a 2,000-person firm in the same sector. The data comes from self-reported survey responses collected from participating organizations, which is why it tends to skew toward larger employers and Fortune 500-style companies. Smaller startups and regional players show up less consistently.

Getting the Towers Watson Salary Guide data for your use case

The database isn't free. Organizations typically pay an annual subscription that runs from about twelve to forty thousand dollars depending on how many segments and locations you need access to. Individual professionals sometimes gain entry through their company's HR department or through professional networks like SHRM chapters that hold group subscriptions. If you are doing independent research without corporate backing, you can find summarized portions of the data in published articles and compensation blog posts, though these are usually six to twelve months old by the time they appear publicly. The actual download link lives behind a login portal at willistowerswatson.com under the Total Rewards product section. You need a verified corporate email and usually some kind of organizational authority check. There is no public direct URL because the data is commercially licensed. Third-party aggregators like Payscale or Glassdoor sometimes pull comparable segments, but they are not the same dataset and you should not treat them as substitutes when precision matters. What most people miss the first time they use this is how the percentiles actually map to real offers. The guide gives you the 25th, 50th, 75th, and sometimes 90th percentiles for base salary, short-term incentive, and long-term incentive separately. A lot of folks take the median number and present it as "this is what the market pays." That is wrong. The median is where half the data points sit below and half above, but it does not account for the fact that two companies in the same segment might structure total cash differently. One might pay a lower base with a higher bonus target; another does the reverse. If you only look at base salary medians without pulling total cash comp, you will misread the competitive picture by ten to twenty percent in several roles.

I ran into this exact problem in 2019 when I was benchmarking a senior product manager role for a mid-size SaaS company. The guide's median base for our segment was sitting at one hundred eighteen thousand, which looked solid on paper. But when I went back and added the bonus targets and restricted stock units using the incentive columns, the total cash actually came out to roughly one hundred forty-two thousand for the 50th percentile. The hiring manager had been preparing offers around the base figure alone, which meant we were about twenty thousand under market for anyone who had a competing offer. The fix was simple: I built a quick spreadsheet that pulled all three columns — base, target bonus as a percentage of base, and estimated RSU value — and recalculated everything before we went to offer stage. Cuts the negotiation back-and-forth time down to basically nothing because you have the full number ready upfront. Another thing worth knowing: the geographic adjustments in this guide are more granular than most other compensation surveys. They break down metro areas rather than just using state-level or broad region categories. That matters a lot if you are hiring across multiple office locations because a San Francisco number and a Seattle number in the same job family can differ by fifteen to eighteen percent even within the same industry segment. Some other guides flatten this into a single West Coast figure, which quietly undervalues the higher-cost cities. There are real limitations though, and I want to be blunt about them. The data is self-reported, which means companies can and do round numbers, report historical data rather than current offers, or exclude certain compensation components. A company might submit base salary figures but leave out sign-on bonuses, which skews the published medians downward for roles where sign-ons are common. There is also a lag — the surveys are typically collected annually with a six to nine month gap between data collection and publication. In fast-moving markets like AI or cybersecurity, that lag can make the guide feel stale within a quarter of it coming out.

For roles where the market moves quickly, I usually cross-reference the Towers Watson data with two other sources: Radford's technology-specific surveys, which tend to capture startup and tech-sector comp more accurately, and direct salary data from levels.fyi for engineering roles. None of these are perfect on their own. Radford has its own reporting biases toward large tech employers. Levels.fyi is heavily skewed toward FAANG and well-funded startups. The Towers Watson guide sits somewhere in the middle — broad enough to be useful across industries, but not specialized enough for niche roles. If you are using this for internal equity work rather than external hiring benchmarks, there is an additional wrinkle. The guide gives you market reference points, but it does not tell you how to handle internal compression or equity adjustments. I have seen HR teams apply the median directly to every open role in a family and then wonder why existing employees in that family started complaining. The guide is a benchmark, not a compensation philosophy. How you position your offers relative to that benchmark — whether you lead at the 75th, match at the 50th, or trail below — is a strategic decision that the data itself cannot answer. The bottom line is that the Towers Watson Salary Guide is one of the more reliable compensation references available, but it requires careful reading. Pull all the comp components, adjust for your specific geography and company size segment, validate against a secondary source for fast-moving roles, and remember that the numbers are directional rather than prescriptive. The people who use it well are the ones who treat it as a starting framework, not a final answer.