Most employers treat a bad Glassdoor review as a reputation problem. A negative Facebook post about workplace conditions gets escalated to leadership, someone gets terminated for it, and the incident is filed under “social media misconduct.” The legal exposure rarely enters the conversation until the US Equal Employment Opportunity Commission (“EEOC”) charge arrives.
The EEOC treats social media evidence differently than employers expect. Public posts about workplace discrimination, hostile conditions, or employer retaliation can constitute protected activity under federal anti-discrimination law. In some cases, they also qualify as protected concerted activity under the National Labor Relations Act (“NLRA”). Terminating an employee over a post that touches on either category creates a retaliation claim that is separate from, and sometimes stronger than, the underlying discrimination allegation.
The cases below come from a recent engagement with a former EEOC investigator with Association of Workplace Investigators (“AWI”) credentials who worked these investigations directly.
The Glassdoor Case: When the CEO Explains It Himself
A tech company employee filed an EEOC charge alleging discrimination based on gender identity and disability. The employee had also posted on Glassdoor before filing, criticizing the company’s culture and alleging that promotions and desirable assignments went to a narrow demographic. The company terminated the employee.
When the former investigator interviewed the CEO, she asked why the employee was terminated. He said it was because of the Glassdoor post. She asked again. He confirmed it.
The EEOC made a finding of retaliation. The CEO’s candor about the reason for the termination made the investigation straightforward, but the legal result would have been the same either way. Terminating an employee because they publicly complained about allegedly discriminatory workplace conditions can constitute retaliation under Title VII, regardless of how the employer characterizes the decision internally or what the employee handbook says about social media use.
The error was not that the employer had a social media policy. It was that whoever made the termination decision did not stop to ask whether that Glassdoor post constituted protected activity. That question should be asked before every termination tied to something an employee said publicly, not after an EEOC charge arrives and a lawyer is asking it for the first time. By the time an investigator is asking the CEO why someone was fired, the answer is already locked in, and there’s no going back to reframe the decision. The safeguard has to sit upstream of the termination, not downstream of the charge.
The Facebook Post That Produced Two Retaliation Claims
A pregnant employee complained internally about discriminatory treatment. A coworker, aware of what was happening, posted on Facebook that the employer was discriminating against the pregnant employee. The employer found the post and fired both of them.
The EEOC found that the coworker’s Facebook post constituted protected activity and that terminating her was retaliation. The termination of the pregnant employee created a separate retaliation claim on top of the original discrimination complaint. Because the Facebook post described working conditions and involved a coworker acting in solidarity with a colleague, it also may have constituted protected concerted activity under the National Labor Relations Act (“NLRA”), creating an additional source of potential liability that many employers do not anticipate when they are focused solely on the EEOC charge.
Two terminations. Multiple findings. One Facebook post that an employer decided was a firing offense. What looked like a single, contained decision produced liability on two separate legal theories from two different agencies, and both terminations trace back to the same failure to ask a basic question before anyone hit send on a termination letter. A company that thinks it is managing one EEOC charge can find itself managing an EEOC charge and an National Labor Relations Board (“NLRB”) charge at the same time, born from the same set of facts and the same missed step.
Commissioner’s Charges and the Wider Net
Most EEOC charges begin with an individual complainant. Commissioner’s charges operate differently. Under federal law, EEOC commissioners can initiate investigations based on publicly available information without an individual first filing a charge. The mechanism is used selectively and generally reserved for matters with broader implications rather than isolated disputes.
The former investigator described the Uber investigation as a documented example of how this works. The agency initiated a Commissioner’s charge based in part on Susan Fowler’s publicly published account of her experience as a software engineer at the company, supplemented by anonymous accounts from other employees and published research regarding harassment in the tech industry. The former investigator was assigned to that case. It involved interviews with dozens of Uber employees across multiple roles and resulted in a pre-determination settlement.
That level of scrutiny is not a realistic concern for most employers. But the underlying principle matters for any organization generating visible, public patterns of conduct. A single Glassdoor review is unlikely to attract a Commissioner’s charge. A pattern of reviews describing similar allegations, combined with social media activity and industry reporting, presents a different risk profile. The Uber investigation began with public information, not a filed charge. An employer’s reputation accumulates evidence whether or not anyone is watching for it.
Third-Party Sources and Anonymous Tips
The former investigator described another case in which a third party, with no formal connection to the employer, began posting publicly that the company discriminated against young women. The third party also contacted current and former employees through direct messages, gathering accounts of their experiences. Investigators used that information not as direct evidence but as a starting point for determining where to focus the investigation.
The EEOC’s authority to use nontraditional sources reflects a broader point: the agency is not limited to reviewing documents the employer provides. Publicly available information, anonymous complaints, social media posts, and even information supplied by union representatives can all influence the scope of an EEOC investigation. Employers who assume the investigation is limited to what they choose to submit are often surprised by what emerges from other sources.The employer is never the only source of the record, and it rarely knows who else is building one.
Internal Communications Are Usually the Real Problem
Social media posts typically enter the EEOC record because the charging party provides them. Investigators are not routinely subpoenaing social media platforms. Internal communications are different, and they are where employers most reliably create their own worst evidence. That distinction matters for investigation strategy: a company can’t control what charging party posts, but it can control what its own managers put in writing.
The former investigator referenced a hiring case in which a manager emailed HR that a candidate was qualified but “seemed a little old for our culture.” That email became the central piece of evidence in an age discrimination charge. In hiring and promotion investigations, investigators are specifically trained to request communications between hiring officials, HR, and supervisors because those records have a documented history of containing admissions that nobody intended to preserve.The practical fix isn’t avoiding email altogether; it is training decision-makers to document job-related qualifications only, and to flag any commentary about age, culture fit, or similar language before its sent, not after it’s subpoenaed.
Interview notes present the same problem in a different form. If interview questions vary across candidates, notes reflect off-topic observations about personal characteristics, or scoring is inconsistent among similarly qualified applicants, the investigation record itself can become evidence of bias. Conducting consistent, documented interviews is not simply good HR practice. It is part of what makes a position statement defensible. Inconsistent notes let an investigator infer discriminatory intent even without a single explicit statement, which is why interview documentation deserves the same scrutiny as any other piece of potential evidence.
What to Do Before the Post Becomes a Charge
The analysis that should happen when a social media post surfaces is not, “Does this reflect badly on the company?”
It is, “Does this constitute protected activity?”
Those are different questions with different answers—and different legal consequences. Answering the wrong question first (reputational instead of legal) is how companies end up building a retaliation claim while trying to manage a PR problem.
A post complaining about workplace conditions, wages, hours, or treatment based on a protected characteristic will frequently fall on the protected side of that line. The employer’s read of tone or professionalism doesn’t change that classification; the content and the protected characteristic are what control.
Social media policies should address legitimate conduct standards without attempting to prohibit employees from discussing wages, hours, or working conditions. The NLRA protects that activity regardless of what the handbook says. Policies that prohibit protected activity may be unenforceable, and enforcing them can create additional retaliation exposure. A policy that survives scrutiny addresses things like harassment, confidentiality, and disparagement of specific coworkers, while leaving wage and working-condition discussions alone entirely.
Before any adverse action is taken against an employee who has made a public complaint, the decision should be reviewed specifically for retaliation exposure. That review is considerably cheaper than the alternative. At minimum, that review should confirm the adverse action was already planned or justified on independent grounds before the protected post surfaced, since timing alone is often the strongest evidence a charging party has.