Native Ads Compliance Checklist: FTC + Every Network's Rules

Native advertising compliance operates on three layers: FTC rules (and their international equivalents) that require clear disclosure and substantiated claims; ad network policies from Taboola, Outbrain, MGID and the rest that decide whether your campaigns run at all; and vertical-specific rules for health, finance and sweepstakes offers, where enforcement runs hottest. A compliant native funnel passes all three layers at all three stages — the ad unit, the pre-lander, and the landing page. This checklist works through each layer with the specific items to verify before launch, because the cost of failure ranges from rejected creatives to permanently banned accounts to regulatory action.

Why native advertising gets extra scrutiny#

Native advertising is designed to resemble the editorial content around it — that is the format's strength and its legal weakness. Because the format blends in, regulators place a heavier disclosure burden on it than on obviously-an-ad formats like banners. The FTC's position, laid out in its Native Advertising: A Guide for Businesses, is that an ad's format must not mislead consumers about its commercial nature, and that any needed disclosure must be clear and conspicuous. Sponsored content labels exist because of that principle, not as decoration.

This is not hypothetical law. The FTC has brought actions over undisclosed native placements and deceptive advertorial formats, and networks have tightened review in response — which means the practical enforcement you will meet first is usually the network's, not the government's. Both run on the same underlying question: would a reasonable reader know this is advertising, and are the claims true?

Layer 1: The FTC baseline#

Five principles cover most of what the FTC expects, and every item on the checklists below traces back to one of them:

• Disclosure must be clear and conspicuous. Proximity to the content it qualifies, prominence a reader cannot miss, and language a normal person understands. "Advertisement" works; a gray "partner content" footnote in 9-pixel type does not.
• Disclosure must come before engagement. A reader should know a page is advertising before consuming it — not discover it at the checkout.
• The net impression must be truthful. Literal truth is not a defense if the overall impression misleads. A technically accurate headline over a fake news layout still deceives.
• Substantiation must exist before the claim runs. Health and performance claims require competent and reliable evidence in hand — not sourced retroactively when a regulator asks.
• Endorsements follow the Endorsement Guides. Material connections disclosed, testimonials representative of typical results or clearly qualified, and experts actually possessing the expertise they claim.

"Clear and conspicuous" deserves one practical translation, because it is where good-faith advertisers still fail: judge disclosure on a phone, not a desktop. A label that sits above the fold on a 27-inch monitor can be two screens down on a mobile advertorial. Contrast matters — light gray on white is a choice regulators have specifically called out. And placement matters more than wording: a correct label positioned after the content it qualifies protects nobody, including you.

Outside the US: the same rules with different letterheads#

If you run international geos — and most native buyers do — the FTC is only one regulator among several enforcing the same core principle. The UK's ASA requires ads to be obviously identifiable as ads under the CAP Code; EU member states enforce the Unfair Commercial Practices Directive against disguised advertising; Australia and Canada have their own equivalents. The practical rule: build the funnel to the strictest standard among your target geos, because maintaining per-country disclosure variants is more fragile than simply being clearly labeled everywhere. Deceptive-format rules travel; loopholes don't.

The ad-unit checklist (headline + thumbnail)#

The ad itself is small, but it sets the net impression everything else is judged against:

• The headline's promise is something the landing path actually delivers — curiosity is fine, bait-and-switch is not.
• No fabricated celebrity endorsements, and no celebrity or public-figure imagery without documented permission.
• No imagery implying news coverage, government status or official warnings.
• Before/after imagery: most networks restrict or prohibit it in health and beauty — check the current policy for each network you run.
• Your display name and branding do not impersonate a publication, an agency or another brand.
• Urgency is real: dates, discounts and stock levels you can verify.
• Thumbnails stay inside network taste policies — shock imagery, medical gore and lookalike UI elements (fake play buttons, fake close buttons) are routine rejection triggers.

A note on the gray zone, because native runs on curiosity: a curiosity gap is compliant, a lie is not, and the test that separates them is simple. Ask what the reader believes at the moment they click, and whether the landing path honors that belief. "The Surprising Reason Your Windows Streak" promising an explanation the page then delivers is a compliant tease. A headline implying a celebrity died, a ban is imminent, or the reader has already qualified for money — when none of that is true — fails the net-impression test no matter how much fine print follows. Curiosity about a real thing: fine. Manufactured false beliefs: that is the line.

The pre-lander and advertorial checklist#

The page between click and offer is where most native compliance failures live:

• An advertising label ("Advertisement," "Sponsored Content" or equivalent) sits visibly at the top of the page. The full standard is covered in our guide to FTC disclosure rules for advertorials.
• The page does not imitate a specific real publication — no borrowed mastheads, no lookalike logos. The advertorial format is legal; impersonating a newsroom is not.
• Every narrator, expert and testimonial is a real, documented person who consented to appear. Invented personas are the most common takedown trigger in advertorial funnels.
• No fabricated comment sections. Fake comments are fake endorsements, and the FTC treats them exactly that way.
• Countdown timers and stock counters are real or absent.
• Review-style and comparison pages disclose the affiliate relationship — the material-connection rule applies squarely here.
• No exit-intent traps, back-button hijacks or forced redirects; most networks ban them outright.

The landing page and offer checklist#

• The claim chain is consistent: ad, pre-lander and offer page make compatible claims. A modest ad feeding a miracle-cure offer page fails on net impression.
• All costs are disclosed before payment: subscription terms, auto-ship enrollment, trial-to-billing conversions. Negative-option billing is a stated FTC enforcement priority, and networks mirror that in policy.
• Contact information, privacy policy and terms are present and reachable.
• Regulated products are geo-fenced to where they may legally be sold.
• No cloaking, ever. Serving reviewers a clean page and users a different one is the fastest route to a permanent ban, and it converts a policy problem into deliberate fraud. We document how it works — and how it gets caught — in ad cloaking detection; the affiliate-side variant is covered in the cloaking glossary entry.

Layer 2: Network policies — what the platforms themselves enforce#

Each network publishes its own advertiser policies, and the details shift, but the broad expectations are stable enough to checklist:

Where networks differ is enforcement posture rather than principle. Networks built on premium publisher relationships review creatives and landing pages more strictly, because their publishers demand it. Mid-tier networks accept more aggressive angles, but still act against deceptive health claims, impersonation and cloaking — and all of them run periodic enforcement waves that retroactively sweep categories that used to slip through. Two practical rules follow: read the current policy documentation for each network before entering a new vertical, and never build a funnel whose economics depend on a policy not being enforced.

It also helps to understand the review lifecycle you are operating inside. Creatives are reviewed at submission, but approval is not permanent: landing pages get re-checked after edits and on periodic sweeps, and accounts accumulate standing — a history of clean funnels buys faster reviews, while a history of rejections buys slower ones and closer scrutiny. Three behaviors damage standing fastest: resubmitting a rejected creative unchanged hoping for a different reviewer, swapping the landing page after approval (which reads as intentional evasion even when it's sloppiness), and racking up rejections in bulk by submitting untested claim language across dozens of ads at once. When a rejection seems wrong, use the appeal path with a short factual note — reviewers do reverse mistakes, and a documented appeal beats a stealth workaround every time.

There is one more enforcement layer below the network: publishers. Premium publishers maintain their own blocklists of advertisers and categories, and a funnel that irritates enough publishers loses access to exactly the inventory that converts best. Nothing about that appears in any policy document — it shows up as your best placements quietly disappearing.

Layer 3: Vertical-specific traps#

• Health and nutra. Disease claims are the brightest line in the entire space: the moment copy claims to cure, treat or prevent a condition, it has crossed from supplement marketing into drug claims. Structure-function language, qualified and substantiated, is the lane. The full reality of running this vertical is covered in nutra on native ads.
• Finance. Earnings guarantees, "risk-free" investment framing, misstated loan terms and fake government-program framing all draw both regulator and network attention. Disclose terms, qualify projections, and never dress an advertiser as a public agency. Crypto and trading offers sit at the sharpest end of this: the category is saturated with fake-endorsement scams, so legitimate advertisers inherit extra scrutiny — expect longer reviews and document your claims accordingly.
• Sweepstakes. A sweepstakes offer needs the classic trio — no purchase necessary, odds disclosure, official rules — plus a clearly identified sponsor. Sweeps also draw elevated network scrutiny because the category has been abused so heavily.
• Brand impersonation. Scam advertisers clone real brands' pages to ride their trust — we've documented the pattern in copycat landing pages. Beyond the obvious "don't do this," it cuts the other way: if your brand is the one being impersonated in native feeds, report it and preserve the evidence rather than just complaining into the void.

Compliance is a monitoring practice, not a launch gate#

The checklist above is a launch gate, but funnels drift: copy gets edited, offers get swapped, a new geo goes live with the old disclosure. Three habits keep a compliant funnel compliant:

• Re-audit after every funnel edit, not on a calendar. The version of the page that got approved is not automatically the version running today.
• Watch your vertical's enforcement line empirically. Which claims survive for weeks, and which vanish days after appearing? An ad library makes this observable: OpenAdLibrary's index of 725,000+ live native creatives with captured landing pages (June 2026) shows how long each angle in your vertical actually survives — a practical read on where networks are drawing the line right now, via the ad intelligence platform. Disappearance alone doesn't prove enforcement — ads also die of unprofitability — but a whole claim-pattern vanishing across advertisers usually does.
• Keep a substantiation file per claim. If a claim runs in your ads, the evidence for it should exist in a folder before the first impression, organized well enough to hand over on request.

Keep approval records too. When a network approves a creative and its landing page, archive the approved versions with dates. If a later sweep flags a funnel that ran approved for months, that archive turns a standing-damaging dispute into a short factual conversation — and internally, it tells you exactly which edit introduced the problem.

The condensed checklist#

FTC layer — disclosure clear, conspicuous, above the fold · disclosure before engagement · net impression truthful · substantiation on file · endorsements real, representative, disclosed

Ad unit — deliverable promise · no fake endorsements · no news/government imagery · before/after checked per network · honest urgency · no impersonation

Pre-lander — ad label at top · no borrowed mastheads · real people, documented consent · no fake comments · real timers · affiliate disclosure on review formats

Landing page — consistent claim chain · full cost and billing disclosure · contact/privacy/terms present · geo-fencing correct · zero cloaking

Verticals — health: no disease claims · finance: no guarantees, full terms · sweeps: no-purchase, odds, rules · no brand impersonation

Ongoing — re-audit on every edit · watch the vertical's live enforcement line · substantiation files maintained

None of this is exotic. The advertisers who survive in native for years are rarely the ones with the cleverest loophole — they are the ones whose funnels never depend on a reviewer, a regulator or a competitor failing to look closely.