Fruit Days Ads: A New Cozy Game Goes Narrow and Deep

Fruit Days went live on September 28 and spent its first six weeks buying nineteen markets — four of them hard. Nearly every plan pointing at Thailand points at nothing else, and the creative behind those plans is rebuilt per market. The plan book is narrow. The per-market commitment is not.

The skeleton is one plan, two placements

Start with what did not change. Fruit Days reports 94.47% ad-plan coverage on Facebook and 94.47% on Instagram — identical to two decimal places. Cozy Florist reports 86.84% and 86.84%. Identical coverage across two bidding surfaces is not a coincidence of taste; it is the signature of bulk plan creation with both placements switched on by default. A buyer who tests placements separately produces different numbers. A buyer who clones a plan template produces the same number twice.

The rest of the stack fits that reading. Fruit Days reaches 81.59% coverage on Audience Network and 38.29% on Threads; Cozy Florist sits at 30.48% and 45.96%. Both keep AdMob near 5% and everything else — AppLovin, TikTok, Messenger, YouTube, WhatsApp — in the low single digits. In other words: Meta's own surfaces do the work, the open market gets a token placement, and no third network is given a real test budget.

Geography is where the two products diverge

Now the split. Cozy Florist's top ten markets read like a European media plan with an American anchor: the United Kingdom at 773 plans, the United States at 588, Germany 483, France 480, Spain 470, Austria 440, Italy 431, Norway 431, Romania 425, Switzerland 402. Sixty-five markets in total, each getting a thin horizontal slice.

Fruit Days reads like four countries and a tail. The Philippines carries 342 plans, Indonesia 312, Thailand 306, Malaysia 305. Then the volume falls off a cliff: Taiwan 59, Hong Kong 31, Singapore 27, Brunei 4, Australia 3, the United States 3. Nineteen markets, four of them holding 51.5% of all plan coverage.

The exclusive-plan counts make the intent clearer than the coverage counts. Thailand has 306 plans touching it and 303 that touch nothing else. The Philippines has 342 and 274. Those are not regional buys that happen to include Thailand — they are Thailand-only buys, which in practice means a locally built plan, a locally built creative set, and a language-specific ad copy batch. Indonesia (312 plans, 7 exclusive) and Malaysia (305, 2) are the opposite: broad buys that sweep them in alongside neighbors.

What a local batch looks like in the creative layer

Single-market plans only pay off if the creative is rebuilt for that market, and the public creative ledger supports the claim that it was. The global set runs on one repeatable idea — a fruit-collection fantasy sold with a short imperative line about getting real fruit — and the same idea is cloned across plans with minor edits rather than re-shot. That is the cheap half of the strategy: one concept, many variants, no market-specific production.

Thailand is the exception, and it is a big one. A market carrying 303 exclusive plans out of 306 cannot be served by a global English line. The Thai batch behaves like a separate production run — local-language headline and body copy, its own hook order, its own creative cadence — wrapped inside the same Meta buying template. The studio did not localize the strategy; it localized the assets and left the buying structure untouched.

That division of labor is the interesting part. Localizing strategy is expensive: new markets mean new benchmarks, new bid levels, new creative briefs, new failure modes. Localizing only the assets keeps a single benchmark set and a single optimization loop. The cost lands in production, not in planning — which is exactly the trade a small studio should want.

Reinvestment, read properly

One number needs care before it becomes a claim. A surface-level reuse counter can read zero while the account is plainly recycling winners, because it measures something narrower than "did this studio re-run proven creative". The window-level view is the honest one: 570 of Fruit Days' 1,084 creatives carry a reinvestment flag, and 898 of Cozy Florist's 1,819 do.

That is a 53% and 49% share — the two products are being run the same way. Roughly half of each active library is a master that has already been approved and pushed back into rotation, and the other half is fresh intake. For a two-title studio this is the sanity-preserving ratio: you cannot fund a test budget large enough to replace the whole library every month, and you cannot survive on masters alone.

What is missing is scale. Cozy Florist converts 4,954 plans and 1,819 creatives into an estimated 130.2 million in lifetime creative exposure. Fruit Days converts 2,460 plans and 1,084 creatives into 27.9 million — roughly a fifth of the exposure on half the plan volume. The gap is not efficiency in the buying layer; it is market depth. Four concentrated Southeast Asian markets simply cannot absorb what sixty-five European and North American markets can, no matter how well the copy is localized.

The uncomfortable question

There is a version of this strategy that is not efficiency at all, and it is worth naming. If one studio can run two cozy-simulation titles off one template, so can the next ten. The template is not a moat; it is a floor. What separates RiftSky from a clone house is that the market books differ — one product is allowed to be a European coverage play and the other is built around four deep Southeast Asian markets, each with its own buying logic.

A clone house would run both titles on the same markets, because that is the cheapest possible version of the template. RiftSky did not, which suggests the geography decisions are being made by someone with a thesis about where each product can win, not by a spreadsheet that maximizes coverage. That is a weaker moat than proprietary creative, but it is a real one, and it is measurable: it shows up as exclusive-plan share in markets that a generic play would never bother to isolate.

What to take from this

The transferable move is not "copy your own creative". Most studios do that. It is the separation of concerns: hold the buying structure steady, spend the marginal budget on local asset production rather than local planning, and let the market list carry the difference. RiftSky runs one benchmark set, one optimization loop, one plan template — and two very different market books.

The test for whether it worked is not plan count. It is whether the concentrated markets hold their exclusive-plan share over the next two quarters. If Thailand's 99% exclusive share drifts down toward the regional average, the studio has quietly given up on local production and reverted to broad buys. If it holds while creative volume keeps growing, the template is doing what it was built to do.

What the broad book costs

Cozy Florist's European book is the mirror image of Fruit Days' concentration, and it is worth pricing honestly. Sixty-five markets at roughly 400 plans each is a coverage play: every market gets enough plans to learn something, and none gets enough to dominate. The advantage is optionality — a title that is already a year old can afford to let twenty small markets each return a modest signal, because the downside of missing one is small and the upside of finding an unexpected winner is real.

The disadvantage is that broad coverage forces generic creative. Germany at 483 plans and 4% exclusive share is not a German campaign; it is a European campaign that happens to include Germany. Only the United States gets single-market treatment at scale — 505 of 588 plans target the US and nothing else — and that corresponds to the market where the title already has the deepest install base. The pattern is consistent: exclusivity follows proven depth, not ambition.

For a new title the same logic inverts. Fruit Days cannot buy sixty-five shallow markets and learn anything useful in six weeks; it does not have the plan budget or the benchmark history. So it buys four deep ones, accepts the smaller total exposure, and uses the exclusive-plan structure to force local creative into existence. The smaller book is a constraint turned into a method.

The creative idea is deliberately thin

The global concept is thin on purpose. A fruit-collection fantasy, a short imperative about getting real fruit, and a clone tree six to ten variants wide is not creative ambition — it is a template that survives translation. Thin concepts travel: they need no cultural setup, no local casting, no market-specific humor. The cost is that they also travel badly when they meet a market that expects its own idiom, which is precisely why Thailand needed a separate batch.

This is the most copyable part of the whole operation, and also the most misread. Studios see a clone tree and conclude that volume is the strategy. It is not. The clone tree exists so that the buying layer has something to rotate when a plan fatigues; it is inventory management, not creative strategy. The strategic choice is the one above it — which markets get built, and which get swept in.

Where the template breaks

Two failure modes are visible from the outside. The first is benchmark drift: a studio that runs one optimization loop across nineteen markets will eventually optimize for the average of those markets, which is none of them. Thailand-only plans partially guard against this by isolating the signal, but Indonesia and Malaysia — 312 and 305 plans with 7 and 2 exclusive — are being run as regional sweep, and a Thai winner will not tell you anything about either.

The second is production debt. A separate Thai batch means a separate production calendar, separate approval, separate localization QA. If the team grows faster than that calendar, the local batch is the first thing to thin out — and when it does, the 99% exclusive share becomes expensive plans running generic assets. That is the number to watch next quarter: not plan count, not creative count, but whether Thailand's exclusivity survives its own production cost.

FAQ

Is Fruit Days a new release?

Yes. Its monitoring record carries a Sep 28, 2026 release date, which puts the whole Oct 10 – Nov 9 window inside its first six weeks.

How is this different from the Cozy Florist breakdown?

The September article read one product's Meta template. This one reads the studio: the same template applied to two products, and the decision about where each product's markets come from.

Why does Thailand matter so much here?

Because 303 of the 306 plans covering Thailand target Thailand and nothing else. That level of single-market buying only makes sense with market-specific creative, which makes Thailand the clearest evidence that the studio does rebuild assets per market.

Does a 53% reinvestment share mean the library is stale?

No. It means about half the active library is proven masters re-approved for rotation and half is new intake. Both products sit near that ratio, which reads as a deliberate balance rather than a shortage of new material.

Can plan count be used to compare the two products' scale?

Only within each market book. Fruit Days runs 2,460 plans across 19 markets and Cozy Florist 4,954 across 65; the per-market density is broadly similar, but total exposure differs by roughly five times because the market pools differ.

Analysis by Maya Chen · CreativeScope · Data window Oct 10 – Nov 9, 2026 · Figures are delivery-monitoring estimates and describe scale and structure, not campaign performance.