What advertisers actually put in the ad

The sample

Subscription apps, a payments company, a team-software company and a hardware subscription, all advertising in the US in September 2026. Two of them we have taken apart in full: Runna and Headspace.

What each advertiser leads with

Advertiser	Category	The line they lead on	Offer in the ad?
Runna	Running plans	"First week free. No credit card required."	Yes, fully qualified
Headspace	Meditation	"Start your free trial today" · "Feel less stressed in 10 days"	Yes, plus a time-bound claim
Sonar Health	Health data	"4.8/5 rating… Download Sonar and try it out for free."	Yes, after social proof
DuoSwim	Swim training	"Access to over 1,000 swim workouts" · Install Now	No – volume as the argument
A running app	Running	"Join over 1 million runners competing for territory."	No – scale as the argument
Oura	Hardware + subscription	"You asked. We listened. Oura Ring 5 is 40% smaller."	No – product news
Wise	Payments	"Say Good-Bye to Hidden Fees"	No – objection removal
Notion	Team software	"Trusted by OpenAI, Figma… Enterprise-ready with SSO and SCIM."	No – borrowed judgement

Source: Live copy from Google Ads Transparency Center and Meta Ad Library, US, checked 11 September 2026. One advertiser left unnamed, for reasons given below.

Three strategies, and what each one assumes

Remove the risk

Runna, Headspace and Sonar all put the free trial in the creative. Runna goes furthest by qualifying it – "no credit card required" – which removes the last objection before anyone taps.

The assumption underneath is that the product needs time to prove itself. You cannot demonstrate a training plan or a meditation habit in ninety seconds, so buying a longer look is worth the conversion rate you give up by not asking for card details.

Assert the scale

DuoSwim leads on "over 1,000 swim workouts"; the unnamed running app on "over 1 million runners". Neither mentions price or trial. The argument is that the thing is already big enough to be safe.

This works when the number genuinely answers the reader's doubt. "1,000 workouts" answers "will I run out of material". A user count answers "will anyone else be there", which matters for a product with a social mechanic and much less for one without.

Remove the objection instead of the risk

Wise and Notion never mention an offer. Wise names the thing people hate about the category – hidden fees. Notion names companies the reader respects and the two acronyms that decide whether IT will approve it.

Both assume the reader already wants the category and is choosing inside it. At that point a trial is irrelevant; what matters is the specific reason to pick you rather than the obvious alternative.

Where the offer decision actually gets made

It is not a copywriting choice. It falls out of two things about the product.

Time to value. If someone can feel the benefit in the first session, hold the trial back – you would rather filter for intent than buy volume. If it takes a week, put the trial in the ad, because you need them still around on day seven.

Who has to agree. One person deciding for themselves responds to risk removal. Several people deciding together respond to evidence that the decision is defensible. That is the whole reason Notion's ads look nothing like Headspace's.

The pattern worth avoiding

One advertiser in this sample has run the same paragraph – word for word, not paraphrased – across roughly ten ad units with start dates spread from late April to mid August 2026. Nearly four months, no line rewritten.

We have left them unnamed because the point is the pattern rather than the company, and because we would be guessing about their results. There are innocent explanations: the copy may have decisively won a test, or the team may be holding message constant while testing video.

But the same evidence fits the most common failure in app advertising. Creative gets treated as a launch asset rather than a consumable. Performance decays gradually, no report says "your assets are stale", and by the time anyone investigates the conclusion is usually that the channel stopped working.

Questions people actually ask

Should the free trial go in the ad or wait for the paywall?: Putting it in the ad raises clicks and trial starts, and lowers trial-to-paid because the trial arrives before any commitment. It suits products that need days rather than minutes to prove themselves. Hold it back when your value is obvious on the first screen and you would rather filter for intent.
Is "no credit card required" always better?: No, and it is the most consequential offer decision most subscription apps make. It reliably multiplies trial volume and reliably cuts conversion. Which side wins depends on your onboarding, and it is testable in about three weeks at reasonable volume.
Why do B2B ads skip offers entirely?: Because the obstacle is not personal risk, it is organisational agreement. A free trial does nothing for a buyer who has to convince four colleagues and a security reviewer. Evidence that comparable companies already decided does.
Do star ratings in ad copy work?: They are cheap and common enough now to be close to wallpaper. They do more when paired with a count – "4.9 from 100,000 runners" carries weight that five emoji stars do not, because the count is the part that is hard to fake.
What is the most common mistake in this sample?: Running one block of copy unchanged for months. One advertiser here has had the identical paragraph live across roughly ten ad units since late April. Even if it won a test, four months is long past the point where creative fatigue outweighs whatever the test showed.

Or have us do it on your account.

The free audit reads your advertising the same way, with the performance data visible underneath rather than inferred from what is public.

Request the audit All teardowns