Meta Ad Library · United States · Active ads only

What the competition is actually running on Meta.

A 360° competitive teardown of paid social in the ecommerce-services category, built from live Meta Ad Library data pulled through the Meta Marketing API. Every ad below is running right now, and every link opens the real creative on Meta's own servers rather than a screenshot.


Ads catalogued

Distinct advertisers

11

Keyword sweeps

6,600+

Total active matches

25 mo

Longest continuous runner

18

Advertisers running a guarantee

3

Ads naming a price

0

Servizio ads live

Captured . Meta publishes no spend or conversion data for commercial ads, so nothing here is a performance figure.

01 — Executive summary

Six findings that change the plan.

Read nothing else and read these. Each is drawn from ads live at the moment of capture, not from best-practice commentary.

01

The guarantee is no longer a differentiator

Eighteen advertisers here run some version of "or you don't pay." Two escalate to work-till-done. A guarantee now buys parity, not attention.

What still cuts through is a falsifiable number. RevCent footnotes a sample size and a date range on its recovery claim. Nobody else does.

02

The oldest ads are the quietest ones

The two longest-running creatives — 25 and 24 months — are both static, both low-promise, and neither carries a guarantee or a revenue claim.

Everything loud is under four months old. The longest-sustained free-teardown campaign in the sweep runs fourteen near-identical creatives with zero guarantee, zero scarcity and zero revenue figure.

03

Anti-agency positioning is the dominant angle — run by agencies

OmniFunnel Marketing alone runs roughly forty-five concurrent variants built on it: "Ask your agency about these 4 dates," "Your Agency Shipped 1 Ad, Not 30," "Your 4X ROAS is 1.8 POAS."

Our own "three separate companies, the money leaks in the gaps" line sits in the most contested territory in the category.

04

The winning hook is arithmetic, not aspiration

Top performers open on a broken number the reader can verify in five seconds. "You shipped 50 products to influencers. 7 posted. You paid for all 50."

The diagnosis doubles as the qualifier. Our current opener names a situation, not a number.

05

Revenue claims are passing review

$200,000,000. $130,000,000/year. $3M/month. 14x ROAS. $1.39M to $1.95M. All live, all approved, all in market today.

Specific client results survive review. Implied personal-earnings promises are what carry disclaimer risk, and only the aspiring-owner advertisers carry those disclaimers.

06

Published pricing is nearly unheard of

Across roughly ninety-five ads examined in detail, exactly three name a price: $299, $499.99/mo and $799. Everyone else hides behind a call.

Our $999 base with a revenue slider is, by itself, an angle nobody in the category is currently using.

02 — Market map

Three markets, one keyword set.

The most important thing to understand before spending a dollar: the keywords we would naturally bid on return three completely different businesses. Only one of them is the competition.

Cluster A

Growth services for real brands

This is the competitive set. Agencies selling management, creator programmes and paid media to brands that already have revenue. Chase Chappell, OmniFunnel, Ascend Commerce, Konsort Social, Lure Agency, GNO Partners, Dahaus, TVG, Sail Away, IntentWave.

Sophisticated copy, application gates, revenue-floor qualifiers, named-client proof. This is where we compete and where the expensive impressions are.

Cluster B

Ecommerce automation and passive income

Noise, but expensive noise. Sells done-for-you stores and capital-deployment pitches to people who want to own a shop, not to brand owners. Ecom Accelerator, Ecom Redefined, Marketplace Xpertz, Ecomsharksss, Elle Liana, SayPro.

Larger by ad count, older creatives, WhatsApp and Messenger CTAs. They bid on the same words and drive up the auction without ever being a real rival.

Cluster C

SaaS and tooling sold to brands

Adjacent, occasionally instructive. Software vendors: Levanta, Social Snowball, Cohley, Helium 10, Redo, Tapcart, Instant Commerce, ConvertCart, Kong, Dashboardly.

Worth studying for offer construction — free pilots, gift-card demos, instant scorecards — but they solve a different buying problem on a different timeline.

The practical consequence

Broad keyword or interest targeting in this category puts our budget against Cluster B's volume. The revenue-floor qualifier in the ad copy is not just a positioning device, it is the cheapest available defence against paying for the wrong audience.

03 — Live creative library

The ads themselves.

Pulled from the Meta Ad Library through the Marketing API at capture time. Each card links to the live snapshot on Meta, where the actual image or video, the full body copy, the landing page and the platform breakdown are rendered by Meta itself.

Meta blocks framing of Ad Library pages, so creatives cannot render inline on a third-party domain. Every link is pinned to the US filter, because Meta otherwise defaults the country to the viewer's own location and returns nothing.

04 — Longevity analysis

The only performance proxy available.

Meta does not publish spend, impressions or conversions for commercial ads. Continuous run time is the one honest signal: nobody keeps paying for a creative that loses money.

What the long runners have in common

Static format. A lead magnet rather than a pitch. No guarantee, no revenue claim, no countdown. The two survivors past 24 months are a CRO checklist download and a plain TikTok Shop service description. Neither would win a copywriting award, and both have outlasted every aggressive competitor in the set.

What this implies for a first campaign

Testing the loudest possible creative first is the opposite of what the durable advertisers do. A static, specific, low-promise lead magnet is cheaper to produce, cheaper to test, and has the better empirical record in this exact category.

05 — Hook taxonomy

Four shapes, repeated endlessly.

Every high-spend hook in the set reduces to one of four structures. Verbatim examples, all currently live.

Shape 01

The arithmetic diagnosis

Names a number the reader can check immediately. Strongest and most common in Cluster A.

You shipped 50 products to influencers. 7 posted. You paid for all 50.UseCreable
Your 4X ROAS is 1.8 POASOmniFunnel Marketing
300 posts a month from "top creators" won't fix a 0.3x ROI.Lure Agency
Shape 02

The status qualifier

States a revenue floor in the first line. Filters the audience and flatters the reader simultaneously.

For eCom Brands Doing $50-500K/MonthSocial Scales
Food & beverage brands doing $50k+/mo on AmazonOnline Biz Momentum
We do not work with dropshippers. We do not work with AI-only brands.RDC Marketing
Shape 03

The incumbent attack

Turns the reader against their current supplier. Note that these are agencies attacking agencies.

Ask your agency about these 4 datesOmniFunnel Marketing
Your agency is optimizing the wrong numberPatrick O'Driscoll
Your Amazon agency gets paid based on how much ad spend they manage.PCOStudio
Shape 04

The anti-guru negation

Defines the offer by what it is not. Dominant in Cluster B, occasionally borrowed by Cluster A.

Every "free TikTok Shop group" leads to a $997 course. This one doesn't.OMNIsquad
This FREE Training at 8 PM EST Might Piss Off Some "Gurus"Brook Hiddink
This isn't about going viral. It's not about building a personal brand.Moe Alamawi
06 — Offer architecture

The free audit, deconstructed.

This is the offer pattern closest to our 30-minute teardown. Six rules the market follows, with the evidence for each.

  1. Name it narrowly, never generically

    Nobody sells "a free consultation." It is a free Google Ads audit, a free EU Compliance Audit, a free CRO audit, a free tracking audit, a free decline teardown, a free Amazon Account Audit. Narrowing the scope makes the free thing feel like a deliverable rather than a sales call, and pre-frames the specialty they want to be hired for.

  2. Promise to find the leak, never to give a plan

    The verb set is consistent: uncover, reveal, find, show you exactly where. "I'll Name 3 Things Broken In Your Account." "Find Out What's Buried in Your Ad Account." Diagnosis is cheap to deliver, creates loss aversion, and the fix is the paid product.

  3. Say "you keep it either way" out loud

    Only a handful do. Echelonn: "No strings attached. Keep the audit or discuss having us manage your account. Your choice." TVG: "You keep the audit either way." Welcome Lane: "You keep the notes either way." Our Guarantee 01 already says almost exactly this.

  4. Promise to disqualify them

    The strongest audit ads offer to reject you. RevCent: "if RevCent is not a fit, we tell you on the call." TVG: "If we can't beat your current numbers, we tell you before we take you on." This reframes the audit as mutual due diligence rather than a pitch, which is what makes the give-away credible rather than gimmicky.

  5. Match friction to ticket size

    Three tiers in use. Zero-friction self-serve scorecards for tool-style offers — INNELS runs an EU compliance check with no account required, results in under two minutes. Booked calls for mid-ticket. Application gates for high-ticket.

  6. The upgrade is proof-of-work, not proof-of-analysis

    The newest versions give away labour rather than a report. A complete landing page built in under 48 hours, free. A digital employee run on your live account for 30 days. $10K in TikTok ad spend. A 100k-follower influencer gifted at signup.

07 — Closest analogue · full teardown

OmniFunnel is running our playbook, one tier up.

Of every advertiser in this sweep, one runs a near-identical model to our teardown offer, at roughly forty-five concurrent variants. It is the most developed version of the thing we are about to build, so it is worth reading line by line.

They put a price on the free thing

Every variant assigns the audit a dollar value before giving it away. "A ranked fix list worth $5,000, yours whether we work together or not." "Get a FREE audit (normally $2,500)."

Our teardown is currently just "free." Naming a value converts a giveaway into a quantified concession.

A named human builds it live

"Michael Tate builds your audit live on the call." "You'll talk to us, not a sales rep." "No PDF, no recorded video, no sales rep." The founder's name and his start year appear in nearly every variant.

Ours is delivered by an unnamed "we". Attaching Hashir or Shaheer by name is free and raises credibility immediately.

The best no-show reducer in the sweep

A working session worth $5,000 whether we work together or not. If you don't show up, it doesn't get built.OmniFunnel Marketing

The audit does not exist until the call happens. That clause removes the "I'll get the report anyway" escape hatch that plagues every free-audit funnel.

Disqualification stated as confidence

And if you're already past your wall without us, we'll tell you.OmniFunnel Marketing

Also: "If the math doesn't work for you, we won't waste your time or money." The same move as our "we'll say so and end the call early."

Capacity scarcity, phrased as a standard

We take 2 to 3 new clients a month. That's a ceiling we enforce, not a number we're trying to grow past.OmniFunnel Marketing

Almost word-for-word our six-accounts-per-strategist line. Two agencies converging independently is the strongest evidence in this report that the framing works.

The revenue floor is in every single variant

"$25K+/month e-commerce spenders." "Ecommerce $5M+, spending $50K+ a month." "Fifty thousand a month minimum."

It appears in the body copy, the headline and the link description. Not once is it left to a form field. This is the gap in our copy, demonstrated forty-five times over.

The one thing we should not copy

Their hooks lean on platform-news urgency — "January 12th Meta removed 7 day view attribution," "Mark August 17 on your calendar," "Since January, Shopify quietly started grading your pixels." Effective and specific, but it dates fast and demands constant rewriting to stay true. With three client-facing people, that is a maintenance burden we cannot carry. The evergreen sampling-arithmetic hook is the better bet.

08 — Guarantee landscape

Everyone guarantees. Nobody proves.

A census of risk reversal in the category, and the one construction that still stands out.

AdvertiserGuaranteeStructure
Ampifi100 micro-influencers in 90 days, or we keep going until we doWork-till-done
Carson MatthewsIf it doesn't work we will work for free until it doesWork-till-done
KongGuaranteed 20% ROAS uplift or you don't payMetric threshold
GNO PartnersIf you don't see a significant lift, you don't pay us a dollarVague threshold
Lijo DecorRevenue holds or grows in 90 days or you don't payMetric threshold
Dahaus Digital30% attributed email revenue in 60 days, or adds $50k–$450k, or you don't payMetric threshold
Vex Media GroupWe beat your email flows or you don't payComparative
Social ScoutIf my flows don't out-perform your current ones, you pay absolutely nothingComparative
Kicksend AIFull refund if it doesn't outperform your current emailsComparative
Utilise SocialScale to £1M/month — or you don't payOutcome
Influencer AdvantageIf it doesn't work, you don't payOutcome
Ascend CommerceYour first TikTok Shop sales in 30 days, guaranteedOutcome
DispoTagEliminate delivery losses in 30 days on your own baseline, or your tags are freeBaseline-anchored
RDC MarketingCommission-based, no contracts. If we don't perform, you don't payStructural, no fee
UseCreableThey post. The feed talks. You pay only for resultsStructural, no fee
Ecom RangersWe only get paid if your store makes moneyStructural, no fee
Creative EcommerceMoney back guaranteeGeneric
Cohley90-day free pilotFree trial

The one that stands out

RevCent publishes "40.4% of declined sales recovered" with a stated sample size of 64,065 and a defined twelve-month window, sourced. It is the only falsifiable claim in the set. Where every advertiser promises the same thing, the specificity is the differentiator, not the promise.

Where we sit

Three written guarantees puts us at parity with the top of the market, not ahead of it. The 48-hour dispatch guarantee is the most unusual of the three because it is operational and verifiable rather than an outcome promise. That is the one to lead with.

09 — Claims and review risk

What Meta is actually approving.

An earlier working assumption was that our aggregate revenue figures would attract scrutiny. The live evidence says otherwise.

Live and approved right now

$200,000,000 in influencer revenue. A $130,000,000/year funnel. $3K to $3M per month. 14.04× ROAS. $1,002,924 across 50,006 units. 119,000 videos for a single named client.

Specific client results, named or categorised, are clearly passing review at scale.

Where the disclaimers appear

Only Cluster B advertisers — those selling a business opportunity to individuals rather than a service to companies — carry explicit income disclaimers. Brook Hiddink's ads state "Results shown are not typical" with a link to full disclosure.

The dividing line is not the size of the number. It is whether the ad implies personal earnings.

Our residual risk

Two things worth pre-empting. The TSP and CAP badge claims should be substantiable on request, and the case-study figures come from team members' prior engagements rather than signed Servizio clients.

Neither is a review problem. Both are a due-diligence problem the moment a prospect asks on the teardown call.

10 — Internal · Servizio vs the market

Where we match, where we don't.

Assessed against the current template landing page and the live site. Three of these are hard blockers: spending money before they are fixed wastes it.

ElementWhat the market doesWhat we haveVerdict
Free audit offerNarrowly named, finds the leak, keep it either way, offers to disqualify30-minute teardown, screen shared, "the number is yours to keep", "we'll say so and end the call early"Best in class
Capacity scarcityFramed as capacity, not countdown6 accounts per strategist, 3 taken, stated as written internal policy that does not resetBest in class
Anti-agency angleThe most contested position in the category"Three separate companies. The money leaks in the gaps between them."Match
GuaranteesTable stakes. 18 advertisers countedThree written: quantified leak, 48hr dispatch, 30 videos liveParity
Published pricingAlmost nobody. 3 prices in ~95 ads$999 base, +$699 Shopify, +$399 Amazon, revenue sliderDifferentiator
Named-client proofName the brand, or give one client's full arc with cost and order countSix referenceable clients named on the template; case studies anonymised under NDAPartial
Hook structureOpens on arithmetic the reader can verify in five secondsOpens on "Your next growth is on a channel you don't have the team to run" — a situation, not a numberGap
Qualifier in copyRevenue floor stated in line oneRevenue dropdown exists on the site form; no qualifier in the page copyGap
Creative libraryStatic dominates the long runners; video dominates the pitch ads35 videos in Drive, all TNC assets. Zero Servizio ad creative existsGap
Conversion trackingEvery ad routes to a measured eventNo Meta Pixel, no CAPI, no GA4 on the template page. GA4 only on the main siteBlocker
Working CTAEvery ad routes to a booked call, application or downloadEvery "Book the teardown" button is an anchor with no href. No Calendly, no form, no mailtoBlocker
Ad accountRequiredNo Business Manager, no ad account, no Facebook Page. Instagram @servizioprimoo existsBlocker
Meta results to showCompetitors publish Meta ROAS freelyTemplate states "Meta is the one we have not put on this page yet"Risk
11 — What follows from this

The sequence.

Ordered by dependency, not by ambition. Nothing below step three is worth doing before step three is done.

  1. Fix the three blockers

    Business Manager, Page and ad account. Pixel plus Conversions API on the template page with a defined lead event. Wire every CTA to Calendly or the existing Web3Forms endpoint. Create the DNS record so the page stops living on a .pages.dev URL. None of this is optional and none of it costs media budget.

  2. Rewrite the hook as arithmetic

    The current opener describes a situation. Every durable competitor opens on a number. Our sampling story already contains the sharpest available version — the gap between samples sent and videos published is exactly the UseCreable structure, and it is a number a prospect can check against their own shop in seconds.

  3. Put the revenue floor in the copy

    Not just in the form dropdown. Given Cluster B's volume in this keyword space, the qualifier is the cheapest defence available against paying for unqualified reach.

  4. Test static lead magnets before video pitches

    The empirical record in this category favours the quiet, specific, static offer. It is also cheaper to produce, which matters given there is currently no Servizio ad creative of any kind.

  5. Lead with the 48-hour guarantee, not the revenue figures

    Outcome guarantees are commodity. An operational promise that can be verified — a sample leaves Houston within 48 hours or the fee is waived — is unusual in this set and harder to copy.

  6. Borrow four things from OmniFunnel immediately

    Put a dollar value on the teardown before giving it away. Name the person who delivers it. Add "if you don't show up, it doesn't get built." Move the revenue floor out of the form and into the copy. All four are free to implement and all four are proven across forty-five live variants.

  7. Decide the Meta-results line before launch

    Running a Meta ad whose landing page admits to having no Meta results is a weak position. Either remove the line or reframe it deliberately as candour. Do not leave it as an oversight.

12 — Method and limits

How this was built.

Sources

Two independent passes over the Meta Ad Library. First, eleven keyword sweeps read through a rendering browser session to capture full body copy, offers, claims and landing domains. Second, a structured pull through the Meta Marketing API's Ad Library endpoint to capture verified ad IDs, page IDs, delivery start timestamps and canonical snapshot URLs.

Scope in both cases: United States, active ads only, all media types.

Limits worth stating

Meta publishes no spend, impression or conversion data for commercial ads, so nothing here is a performance figure and run time is the only available proxy. The Ad Library paginates by infinite scroll and sorts by impressions, so this is the high-spend head of each result set rather than the complete tail. Every claim attributed to an advertiser is that advertiser's own assertion, independently unverified. Delivery start times refer to the creative, not the campaign.