Base Pay is the single most-discussed attribute in employee reviews. Across more than two million Glassdoor mentions mapped to a 50-attribute taxonomy, it generates more discussion than any other attribute: more than career growth, more than workload, more than management quality.
In the public employer-brand content of 325 fully-indexed companies on LinkedIn, compensation is close to invisible. Four in five employers have not posted a single piece of content that mentions any compensation topic. Among the minority that have, the typical company has mentioned it once.
The taxonomy tracks five compensation attributes, which fall into two groups: content about the level of pay (Base Pay, Variable Pay, and Long-term Incentives) and content about pay clarity (Pay Progress and Pay Transparency). Together all five account for less than one per cent of all employer-brand attribute mentions. Content about the level of pay accounts for 0.5% of employer messaging. Content about pay clarity accounts for 0.05%.
Employer messaging clusters around a few themes. Thirteen of 50 attributes account for nearly 80% of all content, and four of those (Career Growth, Team Spirit, Innovation Culture, and Social Responsibility) account for more than 40%. Social Responsibility is the only attribute that employers mention more often than employees do in reviews. Compensation is the opposite extreme.
| Category | % |
|---|---|
| Career Growth | 13.8% |
| Team Spirit | 13.1% |
| Innovation Culture | 10.4% |
| Social Responsibility | 6.9% |
| Base Pay | 0.25% |
| Variable Pay | 0.18% |
| Long-term Incentives | 0.10% |
| Pay Progress | 0.03% |
| Pay Transparency | 0.02% |
When employees complain about pay, most aren’t asking for more money
The conventional wisdom says compensation is a hygiene factor. Daniel Pink’s work, building on Herzberg, makes the case: pay is the most important thing until it meets expectations, and then it should come off the table. The assumption that follows is straightforward: if employees bring up compensation, they must want more money.
That assumption is testable. We tested it, and the data does not support it. When employee pay mentions are classified by the role they play in the reviewer’s experience, most are not complaints about the amount. They are complaints about not understanding how pay works.
Here is what that looks like in the data. Base Pay and Variable Pay are roughly evenly divided between positive mentions (“the pay is good”) and negative ones flagging it as needing improvement. Long-term Incentives skews positive. This is the hygiene factor working as described: the conversation is about whether the number is right, and opinions go both ways.
Pay Progress and Pay Transparency tell a different story. Nearly nine in ten Pay Progress mentions are classified as an improvement area: complaints about unclear progression paths, slow raises, or opaque structures. More than eight in ten Pay Transparency mentions flag it as needing improvement. These employees are not asking for higher pay. They are asking how pay works: how it progresses, how pay decisions are made, what the pay structure looks like.
| Attribute | Positive (Identity) | Improvement Area |
|---|---|---|
| Long-term Incentives | 70.9% | 28% |
| Base Pay | 53.2% | 46% |
| Variable Pay | 50.3% | 48% |
| Pay Transparency | 11.4% | 86% |
| Pay Progress | 9.3% | 90% |
Employers are most silent on the pay topics employees most want answered
The gap between what employees want to know and what employers communicate is widest on pay clarity. Pay Progress and Pay Transparency rank 47th and 49th of 50 attributes in employer-brand content. Only 18 companies across the entire 325-employer panel have published anything about either topic. The 0.05% is not a rounding error on a small priority. It is near-total silence on the compensation questions where employee demand is overwhelmingly one-directional.
| Share of employer-brand content | Complaint rate |
|---|---|
| 13.79 | 52.9 |
| 13.07 | 9.4 |
| 10.36 | 30.5 |
| 6.92 | 21.5 |
| 5.41 | 59.9 |
| 5.35 | 41.1 |
| 4.97 | 64.7 |
| 4.37 | 14.9 |
| 4.01 | 42.9 |
| 3.77 | 8.5 |
| 3.55 | 42.1 |
| 1.89 | 16.8 |
| 1.88 | 31 |
| 1.84 | 70 |
| 1.74 | 23.8 |
| 1.72 | 7.3 |
| 1.64 | 65.9 |
| 1.5 | 39.4 |
| 1.15 | 60.9 |
| 0.95 | 69.8 |
| 0.85 | 3.6 |
| 0.82 | 79.8 |
| 0.65 | 13 |
| 0.59 | 39.4 |
| 0.58 | 64.8 |
| 0.56 | 23.9 |
| 0.54 | 71.3 |
| 0.53 | 38.6 |
| 0.46 | 38.9 |
| 0.45 | 67.1 |
| 0.44 | 17.5 |
| 0.43 | 41.3 |
| 0.41 | 30 |
| 0.33 | 56.1 |
| 0.3 | 59.8 |
| 0.28 | 79.1 |
| 0.27 | 13.2 |
| 0.25 | 45.7 |
| 0.24 | 39.1 |
| 0.21 | 81.8 |
| 0.21 | 7.9 |
| 0.21 | 91.2 |
| 0.18 | 48.4 |
| 0.1 | 27.5 |
| 0.09 | 12.1 |
| 0.09 | 24.3 |
| 0.03 | 89.5 |
| 0.03 | 18.3 |
| 0.02 | 86.4 |
| 0.01 | 7.2 |
Base Pay is what employees think about most: it dominates the review headline, the one-line title each reviewer writes unprompted. But Pay Progress is the compensation topic they most want changed. The employer’s public content addresses neither.
Candidates confirm the priority. When jobseekers in Singapore, Malaysia, and Hong Kong are asked which monetary factors matter most, Pay Progress ranks among the top choices in all three markets. In Hong Kong the signal is strongest: Pay Progress is the second-most selected monetary factor, chosen by more than half of respondents, nearly level with Base Pay. Pay Transparency also appears in all three markets, though at lower rates.
These are not questions only the payroll department can answer. They are questions your content can.
Without employer content on pay, Glassdoor sets the narrative
When an employer says nothing about compensation, someone else defines what candidates and employees hear. For pay, that someone is Glassdoor.
For Base Pay, the balance of positive and negative mentions in reviews is roughly even: some companies pay well and some don’t. For Pay Progress and Pay Transparency, the picture is one-sided. In the cons section of reviews specifically, nearly three-quarters of Pay Progress mentions are negative, and for Pay Transparency, negative mentions outnumber positive by roughly six to one. The employer’s public content offers no counterweight.
| Attribute | Cons | Pros |
|---|---|---|
| Base Pay | 65.9k | 85.3k |
| Variable Pay | 13.9k | 16.2k |
| Long-term Incentives | 1.8k | 4.9k |
| Pay Progress | 33.8k | 4.3k |
| Pay Transparency | 4.0k | 0.7k |
The absence of employer content does not keep compensation out of the conversation. It means the conversation happens without you.
Employees want to know how pay works, not just what it is
The data shows a gap. It does not prove that closing the gap will improve employer-brand outcomes. And the hygiene defence remains partly valid: for employers whose pay genuinely meets expectations, silence about compensation amount might be the right strategy.
But the near-total silence on compensation clarity, when the loudest employee demand is “tell me how this works” rather than “pay me more,” suggests that the hygiene framework is being applied too broadly. Taking pay off the table is only correct if the underlying expectations are met. The data shows that for Pay Progress and Pay Transparency, they overwhelmingly are not.
Your employees are not waiting for you to publish salary bands. They want to know how pay progresses, how pay decisions are made, and what the pay structure looks like. Most employers’ content answers none of those questions.
About the data
This analysis draws on three sources from the EBI database, examined in August 2026.
Employer content (“say”). Attribute mentions extracted from LinkedIn posts published by 325 multinational employers whose LinkedIn post data is fully indexed (gated on the database’s own completeness flag). Posts are classified by content type using an LLM pipeline; only those classified as Employer Branding or Recruitment Marketing are then mapped to the 50-attribute taxonomy. Approximately 58% of LinkedIn posts (those classified as Corporate Branding) are excluded from attribute mapping by design. Some compensation mentions may live in that excluded content. Instagram and YouTube data for the same employers is included in the broader dataset but adds fewer than 30 compensation mentions per channel, consistent with the same pattern.
Employee reviews (“do”). Attribute mentions extracted from Glassdoor reviews for the broader EBI employer panel (approximately 375 companies, including some whose LinkedIn data is still being indexed). Reviews are mapped to the same 50-attribute taxonomy with an additional dimension classification (Identity, Improvement Area, Ambition, Differentiator, Image, Join, and Stay) capturing the role each mention plays. “Improvement Area” is the dimension label for mentions where the reviewer flags something as needing to change; “complaints” in this article refers to mentions carrying that classification. Glassdoor reviewers are self-selected; the population likely over-represents strong opinions. The platform’s review structure (pros, cons, a summary headline, and advice to management) provides a framework that may surface opinions that would not appear in unstructured conversation. Base Pay’s dominance survives the most spontaneous signal available: it is the most common topic in the reviewer-written headline, ahead of Career Growth and Work-Life Balance.
Candidate preferences (“want”). Dynata primary research: Singapore (n=200, fielded May 2026), Malaysia (n=200, fielded June 2026), and Hong Kong (n=200, fielded August 2026). The survey uses a two-stage forced-choice design: Stage 1 asks respondents to choose up to 3 of 9 monetary attributes; Stage 2 asks them to choose up to 5 of 41 non-monetary attributes. Mention rates are not comparable across stages. All candidate preference findings in this article are reported within the monetary stage only. The data covers working adults aged 22-55 in three APAC markets and does not represent global candidate preferences.
What we are not claiming. This silence is not uniformly wrong. Some employers may communicate compensation effectively through channels outside this analysis: job descriptions, recruiter conversations, careers pages, internal portals. Employers may also be rationally responding to platform incentives: social media algorithms reward engagement-friendly content. The LLM pipeline that classifies posts and maps attributes is not human-validated. This is a cross-section, not a trend. The employer-content panel is 325 large multinational employers, not a random sample. “Blackout” is editorial shorthand for near-silence; 66 of 325 companies do mention compensation, at a median of one mention. Figures will be frozen to a publication-date snapshot; the live panel continues to grow.
How these articles are written — where automated analysis ends and human editing begins — and full detail on how the database is built and what it records are on the methodology page.