How to Spot a Low-Quality Online Course Before You Pay

The online course market has a strange property: some of the best educational products ever made share shelf space with some of the most cynical, and they use the same fonts, the same platforms, and the same five-star badges. The difference rarely shows up until after you’ve paid — unless you know exactly where to look. This checklist is where to look. It takes about thirty minutes, costs nothing, and will save you from almost every bad course purchase you’d otherwise make.

A disclosure that’s also the point of this article: we earn affiliate commissions when readers buy courses through some of our links. That’s exactly why this page exists — our only durable asset is your trust, and the fastest way to lose it is pointing you at junk. So here is, in public, the full checklist we run before any course earns a recommendation on this site. Use it on our recommendations too.

Step 1: Read the marketing before the course — it’s the most honest thing on the page

Not the claims themselves — the style. How a course sells itself is the one signal a low-quality seller can’t hide, because the marketing is the part they invested in.

The income-claim tell. “Students earning $10k/month.” “Land a six-figure role in 12 weeks.” “The skill that made me financially free.” A course can control what it teaches; it cannot control hiring markets, your effort, or your outcomes — and sellers know this. Income-first marketing means the product being sold is hope, and hope-sellers systematically underinvest in curriculum because the curriculum isn’t what converts. This isn’t just our cynicism: income-claim and business-opportunity marketing is a long-running enforcement priority for regulators like the FTC, precisely because the pattern so reliably marks deception. [TODO: verify current FTC guidance link at ftc.gov before publish.]

The transformation-vs-content ratio. Read the sales page and count: how many words describe what you will learn versus who you will become? Legitimate courses lead with content — modules, skills, tools, projects — and mention outcomes carefully. Junk courses invert the ratio: lifestyle photography, testimonial montages, “imagine a year from now,” and somewhere near the bottom, a vague bullet list of what’s actually inside.

The manufactured-enemy tell. “Universities don’t want you to know this.” “Degrees are dead.” “While everyone else follows the old rules…” Framing education as a rebellion you can buy is a conversion tactic aimed at people the seller assumes won’t check credentials. Which brings us to —

Step 2: Demand the full syllabus — no syllabus, no sale

This is the cleanest binary check on the list. Legitimate courses tell you exactly what they teach before you pay. University-lineage platforms show full module-by-module syllabi, lesson counts, and time estimates for free — you can read every unit of a Coursera or edX program, and even audit much of the content, before spending anything.

So when a course won’t show you a syllabus — when the curriculum section reads “Module 3: The Scaling Secret 🔥” with no substance, or the page sells 40 hours of “value” without saying 40 hours of what — the inference is simple: content that could survive being seen would be shown. Hidden syllabi hide something.

While you’re checking, note scope honesty too. A real syllabus for a real skill acknowledges difficulty and prerequisites. Any course promising to take you from zero to professional in a complex field in a few weekends is lying about the destination, the timeline, or both — for calibration on what competence actually takes, see the honest ranges in our learn AI from scratch roadmap.

Step 3: Verify the instructor exists outside the course

Take the instructor’s name and search it off-platform. Fifteen minutes, three questions:

1. Did they do the thing they teach? Verifiable employers, shipped projects, publications, talks, a track record in the field — not just a bio paragraph on their own sales page. LinkedIn, GitHub (for technical fields), and plain search results usually settle this fast.
2. Is their only credential selling courses about the thing? The “guru loop” — people whose expertise in X consists of selling courses about X, whose income proof comes from course sales — is the defining structure of the junk-course economy. It’s most rampant in money-adjacent niches (trading, dropshipping, “AI side hustles”), and it’s exactly the pattern we flag in the prompt-guru corner of the prompt-engineering course market.
3. Does anyone credible reference them? Practitioners in real fields leave footprints in communities, citations, and colleagues’ mentions. A ghost who exists only in their own funnel is a finding, not an absence of data.

To be fair to newcomers: plenty of excellent teachers aren’t famous, and teaching skill and practitioner fame are different things. The check isn’t “are they a celebrity” — it’s “can any part of their claimed background be independently confirmed.” Unverifiable is the red flag.

Step 4: Cross-check the reviews — off the sales page, middle stars first

Testimonials on a sales page are marketing copy. Sellers choose them, and in the worst cases manufacture them. Real review diligence happens elsewhere:

Where to look. Search "[course name]" review reddit and the same in the subject’s real communities — subject-matter forums and communities are where unfiltered experiences live, and where “did anyone actually get value from this?” gets answered by people with no commission. Also search the instructor’s name the same way; serial junk-course sellers leave trails.

How to read on-platform ratings. Where marketplace ratings exist, skip the five-star wall and read the 2–4 star reviews first. Fake positives cluster at five stars and sound generic (“Great course! Learned so much!”); genuine middling reviews cite specifics — outdated sections, thin modules, unanswered questions — and those specifics are the actual product information you came for. A large volume of detailed, mixed reviews is a healthier signal than a small wall of ecstatic ones.

Manipulation tells. Review bursts in a short window; recycled phrasing across reviewers; reviews describing life outcomes rather than course content; and on self-hosted platforms, the total absence of any negative review anywhere — real products at scale always have some. We cover the general skill in more depth in the fake-review context of comparison shopping, but the short version: distrust any review population the seller fully controls.

One more cross-check: search "[course name]" refund and "[course name]" scam. Ugly results aren’t automatically damning — unhappy customers exist for good products — but patterns of refund-fighting stories are disqualifying, which is also step 7’s subject.

Step 5: See through discount theater and fake urgency

Urgency exists for one reason: to make you buy before doing the checks on this list. The mechanics are worth naming so they lose their power:

• The perpetual discount. A course that is always “$499 $49, ends tonight” costs $49. The anchor price is fiction, manufactured to make a mediocre price feel like a steal. Note that deep-discount pricing is the normal model on marketplaces like Udemy — which is why, there, the discount tells you nothing and the syllabus, instructor, and review checks tell you everything. Never pay list price on a marketplace, and never credit the “savings.”
• Countdown timers that reset. Open the page in a private browser window tomorrow: same timer, same “last chance.” An evergreen deadline is a small, checkable lie — and a seller who lies about small checkable things is telling you how they handle big uncheckable ones.
• Scarcity that can’t exist. “Only 3 spots left” for a self-paced video course with no cohort is arithmetic-proof nonsense. Recorded video has no seat limit.
• The webinar funnel. The free “masterclass” that’s 50 minutes of story and 10 minutes of pitch, ending in a one-time offer that expires when the call does. The format is engineered to bypass deliberation; the price at the end is usually in the hundreds or thousands. Legitimate education survives you sleeping on the decision. Junk can’t afford to.

The universal counter-move is boring and total: never buy inside an artificial deadline. If the course is good today, it will be good — and almost certainly the same real price — next week.

Step 6: Price the certificate at its true value

For many buyers the certificate is half the purchase, so price it honestly: a certificate is worth exactly the reputation of its issuer. Not its border, its Latin, or its LinkedIn shareability.

The checks: search whether employers actually recognize the credential (job postings that name it are the strongest evidence — the way postings genuinely do reference, say, well-known cloud or data analytics certifications). Be actively suspicious of impressive-sounding accreditation from bodies you’ve never heard of — “certified by the International Board of X” often traces back to an entity that exists to certify the certifier. And watch for the certificate-mill inversion: when the sales page emphasizes how the credential looks — frame-ready, badge-ready — over what the coursework contains, the credential is the product, and credentials-as-product are precisely what employers have learned to discount to zero.

Calibration matters here, honestly applied even to things we recommend: name-brand certificates from Google or big platforms are recognizable but are still starting-point signals, not job guarantees — we make exactly that argument, at length, about the Google Data Analytics Certificate. If a famous certificate is only a foot in the door, an unknown one is not even that. And if you’re buying certificates in volume, check whether a subscription changes the math — that’s Coursera Plus’s actual use case — rather than paying mill prices for mill paper.

Step 7: Read the refund policy — it’s the seller’s confession

Last check, and quietly the most revealing: the refund policy tells you what the seller believes about their own product.

Good signals: a clear, unconditional window (“full refund within X days, no questions”), stated plainly and findable before checkout. Platforms with audit-free-then-pay models are effectively offering infinite trials, which is the confidence ceiling.

Bad signals: consumption caps (“refund only if you’ve watched under 10%”) — which punish you for discovering the course is bad by checking whether the course is bad; absurd windows (48 hours on a 40-hour course); “case-by-case” discretion, which means no; refunds that require “proving you did the work,” a hoop designed for abandonment; and no findable policy at all, which is the same as “no.”

Protect yourself procedurally: screenshot the refund policy as it existed when you bought; pay by credit card, which gives you a dispute path if a seller stonewalls a policy-compliant refund; and on any platform purchase, know the platform’s own refund rules, which sometimes override the instructor’s. If a refund request meets hostility, note that chargeback processes and consumer-protection complaints exist — sellers who fight fair refunds are counting on you not knowing that.

The 30-minute pre-purchase checklist, assembled

Run it top to bottom before any course purchase over trivial money:

1. Marketing: sells skills, not income or identity? No manufactured enemies?
2. Syllabus: complete, public, specific, honest about prerequisites and timeline?
3. Instructor: verifiable off-platform, with expertise beyond selling courses?
4. Reviews: cross-checked off the sales page, middle-stars read, no manipulation tells?
5. Urgency: none — or ignored, with the decision slept on?
6. Certificate: issuer reputation checked against real job postings, mill tells absent?
7. Refunds: clear window, no hoops, policy screenshotted, paid by credit card?

A course that passes all seven is rarely a bad purchase — the worst case is usually “decent but not for me,” which the refund window covers. A course that fails two or more is a pattern, and you should trust the pattern over the testimonials every time.

Our honest take

The uncomfortable summary is that the junk-course economy works because its buyers want the promise to be true — a purchasable shortcut past the slow, unglamorous work that skills actually require. The sellers aren’t confused about this; the checklist above is, at bottom, seven ways of asking “is this product built for my learning or my hope?”

The good news is the honest end of this market has never been better: transparent syllabi, free audit tracks, real instructors, and courses that cost less than a textbook. That’s the end we try to live on — every recommendation on this site has been through the checklist you just read, and our beginner AI course picks and free course roundup are reasonable places to see it applied. Everything else we’ve published lives at the AI & Data Skills hub.

And if this page just talked you out of a $497 purchase with a countdown timer on it — good. That was the job.