PolyBuzz runs one German ad, in Germany, and it has kept it live for 97 days. That is the whole strategy in one line. First seen on 5 April 2026, "Chatten mit KI Chatbot" was still running when we pulled the data on 11 July. For a rising app in a crowded category, that restraint is the interesting part.
We read this from Apple's public Ad Repository, which discloses the ads an advertiser runs in the EU. Data window ending 11 July 2026, EU storefronts. PolyBuzz is a challenger in AI companion apps, not an incumbent, and its first ad in this window only appears in April. So its choices read as a young app deciding where to point a small budget. Here is what the public data shows.
One market, one message, 97 days
PolyBuzz's footprint is 26 ad records that collapse into 2 distinct creatives with the same copy, all in App Store search results placement, the slot that shows after someone searches. One country: Germany. One language: German. One headline it has held since April.
Longevity is a proxy here, not a measured conversion rate. We cannot see PolyBuzz's installs or its cost per download. What we can see is that it found a line it trusts in one market and kept paying for it for over three months rather than swapping in something new every few weeks. When a young advertiser holds a single creative that long, it usually means the ad is earning its slot.
What the incumbents do differently
Put PolyBuzz next to the two established names in the category and the contrast is clean.
Character AI runs its top creative, "Make Your Own Chatbot on c.ai", across 10 EU countries at once, from Bulgaria to Sweden, in one untranslated English line, and has held it the full 365-day window. It is still live. That is an incumbent covering the whole map with a single proven message.
Replika sits in the middle. It keeps a stable of five creatives past 200 days, and it localizes, running German and Spanish variants of "Always here to listen and talk" alongside English ones. It recently opened a fresh angle, "Someone you'll love to talk to", live 19 days, while the old winners keep running underneath.
Two mature approaches, one thin and wide, one deep and localized. PolyBuzz does neither yet. It picked a single market and stayed there.
Why the narrow footprint is the smart opening move
The pull for a rising app is to look like the incumbents on day one: many markets, many creatives, spend spread everywhere. PolyBuzz's data argues for the opposite opening. Prove one message in one market before you widen.
Germany is a large, high-value App Store market, and running German copy there removes the language question entirely. By holding one creative for 97 days, PolyBuzz gets a clean read on whether that ad pays before it copies the motion into a second country. Character AI earned the right to run 10 markets on one line by holding that line for a year first. Replika earned its localized stable the same way, over months. Neither started scattered.
What a solo founder should take from this
If you are one person shipping an iOS app, PolyBuzz is the closest model in this category to your actual constraints. You do not have the budget to test 20 creatives across five countries. You have enough for one good ad in one market, held long enough to trust the number.
Three things carry over. First, pick one market and commit to it. PolyBuzz chose Germany and stopped there, which keeps the signal clean and the spend legible. Second, hold your message. Ninety-seven days on one creative is how you learn whether it works, and rotating too early throws that away. Third, read longevity before you spend. A rival's 97-day ad, or a 365-day one, tells you which creative it trusts, and that is the fastest read on what converts in your category without running the test yourself.
See any advertiser's ads yourself
You can read PolyBuzz's live EU creatives, or any advertiser's, the same way we just did. Run a free scan at recao.app. No signup. You get the advertiser's live App Store ads, ranked by how long each has run, so the longest-lived winners sit at the top.
Method note: every figure here comes from Apple's public Ad Repository (DSA Article 39), for the data window ending 11 July 2026, covering the EU storefronts. We count distinct App Store creatives and their placements as Apple discloses them, and we use each ad's first-seen date to measure how long it has run. Longevity is a proxy for what converts, not a measured conversion rate: advertisers tend to keep paying only for ads that work. We invented no numbers; if a figure is not in Apple's disclosure, it is not in this piece.