The Importance of Reading Online Reviews Before Hiring

Why Online Reviews Have Become the Closest Thing Most People Have to a Background Check

Before you let a stranger into your house, work on your boiler, care for your kid, or handle several thousand dollars of your money, you want to know something about them. For most people, most of the time, that knowledge comes from a page of star ratings written by other strangers.

It’s a strange system when you look at it directly. You’re outsourcing a judgment about risk to an anonymous crowd, on a platform with commercial incentives of its own, about a business that has every reason to shape what appears there.

And yet reviews are genuinely useful — if you read them properly. The skill isn’t in finding a high star rating. It’s in reading the shape of the feedback, knowing which patterns are meaningful, and understanding what reviews structurally cannot tell you no matter how many you read.

This guide covers all three, including the federal rules that now govern what businesses are allowed to do with reviews, which changed more in the last two years than in the previous twenty.

Understanding the Federal Rule That Now Governs What Businesses Can Legally Do With Online Consumer Reviews

Most guides to reading reviews skip this entirely, which is odd, because it’s the part that tells you what the baseline actually is.

In August 2024 the Federal Trade Commission finalized a trade regulation rule — 16 CFR Part 465, the Rule on the Use of Consumer Reviews and Testimonials — that took effect on October 21, 2024. It doesn’t offer businesses guidance. It defines specific practices as unfair or deceptive, which means the FTC can go straight to civil penalties rather than building a case from scratch each time.

That distinction matters more than it sounds. Before the rule, the FTC had to argue in each individual case that a given review practice was deceptive under Section 5 of the FTC Act. Now the conduct is enumerated, and the agency has said explicitly that codifying it was intended to let them move faster.

Fifty-three thousand dollars per violation is not a rounding error, and “per violation” in a review context can mean per review. That’s the deterrent the rule was designed to create.

The Specific Categories of Deceptive Review Practices That Federal Regulators Have Now Prohibited Outright

The rule breaks the prohibited conduct into distinct provisions. Knowing them is genuinely useful as a reader, because it tells you exactly which manipulations a business has a legal reason to avoid — and therefore which ones, when you spot them, are a serious signal rather than a minor annoyance.

Prohibited practice	What it looks like from the outside
Fake or false reviews and testimonials	Reviews from people who don’t exist, including AI-generated ones, or from people who never used the service. Creating them, selling them, or buying them are all covered.
Buying reviews with a required sentiment	Compensation or incentives conditioned on the review being positive — or negative, if aimed at a competitor. The rule covers conditions conveyed implicitly, not just stated outright.
Undisclosed insider reviews	Reviews by officers, managers, employees, or agents that don’t clearly disclose the connection to the business.
Company-controlled review sites	A site that presents itself as offering independent reviews while actually being controlled by the business it reviews.
Review suppression	Using threats, intimidation, or false public accusations to get a negative review removed or prevented, or presenting a curated subset as if it were the full picture.
Fake social media influence indicators	Bought or bot-generated followers, likes, and views, where they’re used to misrepresent influence for commercial purposes.

One nuance worth knowing: the rule permits businesses to flag or remove reviews they reasonably believe are fake, but the FTC has signaled that they should be able to substantiate that belief. “We took it down because we thought it was fake” is not a blanket defense for deleting criticism.

Another: the proposed prohibition on review hijacking — repurposing reviews from one product onto a substantially different one — was not finalized. The Commission couldn’t settle on a workable definition and left the door open to revisiting it. So that particular practice sits in a grayer zone than the six above.

Why the Distribution of Star Ratings Tells You More Than the Average Score Displayed at the Top

The average is the least informative number on the page. It’s a single figure compressing wildly different realities.

A 4.3 built from a smooth slope — lots of fives, a good number of fours, a few threes, a scattering below — describes a business that’s generally good with normal variation. A 4.3 built from a wall of fives and a cluster of ones describes something else entirely: a business that either delivers brilliantly or fails badly, with very little middle. That’s a volatility profile, and for anything high-stakes, volatility is the thing you actually care about.

Most platforms show you this breakdown. Almost nobody clicks it.

Sample size matters too, in a way people consistently underrate. A 5.0 from six reviews carries almost no information. A 4.4 from four hundred carries a lot. If you’d like an intuition for it: with a handful of reviews, one enthusiastic friend or one bad day moves the whole number. With hundreds, individual noise cancels out and you’re seeing something closer to the real pattern.

Recency is the third axis. A business that was excellent three years ago and mediocre for the last eight months will still show a healthy lifetime average. Ownership changes. Key staff leave. Sort by most recent and read the last three to six months separately from everything else. If the recent picture diverges from the lifetime picture, trust the recent one.

What Negative Reviews Actually Tell You and Why an Unbroken Record of Five Stars Should Make You Suspicious

Negative reviews are the most useful reviews on the page, and most people skim past them looking for reassurance.

Read them for pattern, not for outrage. One person furious about a scheduling mix-up tells you nothing. Six people over eighteen months describing the same specific failure — no-shows, invoices that don’t match the quote, work that had to be redone — tells you what this business is actually like when things go wrong. Which is the situation you’re really buying insurance against.

Sort negative reviews into two buckets. Complaints about outcome and competence sit in one: work quality, damage, safety, billing disputes. Complaints about experience and preference sit in the other: brusque manner, arrived at 8:10 instead of 8:00, didn’t like the paint color they recommended. The first bucket should weigh heavily. The second usually shouldn’t.

Then read the responses. How a business replies to criticism is one of the most predictive signals available to you, and it’s free. Responses that address the specific issue, acknowledge what went wrong, and describe a remedy suggest an organization that handles problems. Responses that attack the reviewer’s character, dispute that they were ever a customer, or hint at legal consequences suggest the opposite — and depending on how far they go, may brush against the rule’s review suppression provisions.

A perfect record deserves scrutiny rather than confidence. Every business that has operated for any length of time has disappointed someone. A wall of unbroken fives means either a very small sample, a very new business, or curation.

None of these proves anything on its own. Small businesses genuinely do ask happy customers to leave reviews, and a batch after a busy month is normal. But several of these together, especially the burst-after-criticism pattern, is worth weighing.

How Different Review Platforms Verify Their Contributors and Why That Changes How Much Weight to Give Them

Not all review sources are built the same way, and the differences are structural rather than cosmetic.

Source type	Strength	Limitation
Open platforms	Large volume, easy to find	Anyone can post without proving they were a customer
Transaction-verified platforms	Reviewer demonstrably paid for the service	Only covers jobs booked through that platform
Testimonials on a business’s own site	Often detailed and specific	Selected by the business; you never see what was left out
Licensing board records	Formal complaints and disciplinary findings	Only captures matters serious enough to be filed
Nonprofit complaint databases	Shows dispute handling, not just sentiment	No enforcement power; participation is voluntary

Cross-referencing is the whole technique. A business with strong reviews on one platform and nothing anywhere else is a thinner signal than one with consistent patterns across three. Divergence is informative in itself — if a provider is excellent on the site where they control the listing and mediocre everywhere else, you’ve learned something.

Curated testimonials on a company’s own website aren’t dishonest by default. Displaying only your best feedback is ordinary marketing. But under the rule, presenting a curated selection in a way that misrepresents overall sentiment can cross into review suppression. As a reader, treat a testimonials page as advertising copy, because that’s what it is.

Why Reviews Cannot Substitute for Verifying Licensing, Insurance, and Regulatory Standing Before You Hire

This is the limit that matters most, and it’s the one people most often ignore.

Reviews measure customer satisfaction. They do not measure whether someone holds a valid license, carries liability insurance, is bonded where the state requires it, or has an open disciplinary matter. Those are separate records held by separate bodies, and a five-star average tells you nothing about any of them.

People are frequently delighted by unlicensed, uninsured work. The reviewer had a good experience, the job looked fine, and they had no way to know what wasn’t in place. The exposure only surfaces when something fails — and by then the reviews that reassured you are irrelevant.

For anything touching health, money, children, or structural safety, the sequence should be: verify credentials first, read reviews second. Reviews decide between qualified providers. They don’t establish that someone is qualified.

Every one of these is free or close to it, and none of them takes long. The federal CareerOneStop tool, run by the Department of Labor, will point you to the correct licensing authority for a given occupation and state if you don’t know where to start.

How to Weigh Online Reviews Differently for Higher-Stakes Services Where the Consequences of Getting It Wrong Are Severe

Choosing a coffee shop from reviews and choosing a home health aide from reviews are not the same activity, and the same reading strategy shouldn’t apply to both.

As stakes rise — anything involving health, safety, significant money, or the care of someone vulnerable — three adjustments are worth making.

First, weight negative reviews more heavily than positive ones. When failure is expensive or dangerous, the downside tail is what you’re actually assessing. Two credible reports of a safety issue outweigh fifty reports of pleasant service.

Second, discount vividness. Emotionally intense reviews, positive or negative, are memorable out of proportion to their informational value. A single furious account will stick in your head longer than twenty measured ones. Notice that happening and correct for it.

Third, treat clinical and technical claims in reviews as unreliable. A reviewer saying a treatment cured their condition, or that a repair was done to code, is reporting a belief, not a verified fact. They usually have no way to assess either. Warmth, punctuality, communication, and whether they felt respected are things reviewers genuinely can report on — and those are worth reading closely.

For regulated healthcare and childcare specifically, the licensing and inspection records maintained by state agencies carry far more weight than any review platform, because they reflect trained inspectors applying defined standards rather than customers applying personal ones.

What to Do If You Believe You Have Encountered Fake Reviews or Deceptive Testimonial Practices

You can report it, and the reporting route now has teeth behind it.

Complaints go to the FTC at ReportFraud.ftc.gov. The agency doesn’t intervene in individual disputes, but reports feed the Consumer Sentinel Network, which is what enforcement sweeps are built from. The December 2025 warning letters didn’t come out of nowhere — the FTC described spending the rule’s first year letting the complaint pipeline mature before acting on it.

Your state attorney general’s consumer protection division is a parallel route, and some states have their own review and advertising statutes that go beyond the federal rule.

Review platforms themselves also have reporting mechanisms, with varying degrees of responsiveness. It costs nothing to use both channels.

If you’ve already been harmed financially — you hired someone on the strength of reviews you now believe were manufactured — the review issue and the underlying dispute are separate matters. The first is a regulatory complaint. The second usually runs through the licensing board, small claims court, or your payment provider’s dispute process, depending on the amount and how you paid.

Frequently Asked Questions About Reading and Interpreting Online Reviews Before Hiring

Can a business legally ask customers to leave reviews? Yes. What the rule prohibits is conditioning compensation or incentives on the review expressing a particular sentiment. Asking for honest feedback is fine; paying for five stars is not.

Can a business delete negative reviews? On platforms it controls, it has more latitude, but presenting a curated set as though it reflects overall sentiment can constitute review suppression under the rule. Using threats or false accusations to get a review removed is prohibited outright.

Are AI-generated reviews covered? Yes. The rule addresses reviews misrepresenting that they come from someone who doesn’t exist, which the FTC has explicitly connected to AI-generated content.

How many reviews should I read? Read the most recent ten to fifteen, then all the one and two-star reviews from the past year regardless of number. That’s usually enough to see the pattern.

Should I trust a business with no reviews at all? Not automatically distrust it — new and small businesses exist. But with no review history, verification of licensing, insurance, and references does all the work, so do it thoroughly.

Do employee reviews of the business tell me anything? Sometimes, indirectly. Persistent reports of high turnover or unpaid wages can bear on whether the crew showing up at your house is experienced or stable. Treat it as weak supplementary signal, not evidence.

References and Citations