App Store Promotional Text: The 170-Character Evergreen Field

Every other word on your App Store listing is locked behind a review cycle. Promotional text isn’t: 170 characters, rendered above your description, editable any afternoon and live in minutes. It is the closest thing iOS gives you to a landing-page headline you can rewrite at will — and the field indies most often leave blank, stale, or filled with the exact copy Apple rejects.

This guide covers the field’s actual mechanics, how it divides labor with the subtitle, a writing formula with five worked examples, the guideline 2.3.1 traps, and how to test variants now that custom product pages carry their own promotional text.

It pairs with the keywords field deep-dive (the ranking half of metadata) and the ASO playbook for the full system.

What promotional text actually is

A 170-character field in App Store Connect, displayed at the top of your description on the product page. Four properties define how to use it:

• No review. Edits go live in minutes without a binary or a reviewer — unique among listing fields.
• No ranking weight.Apple doesn’t index it for search. It can’t win you impressions; it can only convert the ones you already get.
• Above the fold.It renders in full before the “more” tap that hides the rest of your description — so it’s read by nearly everyone who scrolls past the screenshots.
• Per-locale and per-CPP. Every localization has its own 170 characters, and every custom product page can carry different promotional text from the default page.

Put together: this is the highest-read, lowest-cost, fastest-iterating sentence on your listing. The leverage-to- effort ratio is unmatched anywhere else in App Store Connect.

Promotional text vs subtitle — the difference

The division of labor: the subtitle spends its 30 characters on your single best keyword wrapped in a benefit; promotional text spends 170 expanding that benefit into a reason to install today. When the two say the same thing twice, the listing wastes its most valuable real estate saying one thing — the subtitle helper covers the other half of the pairing.

The 170-character formula

1. Positioning sentence. Who the app is for and what it does differently — written so it stays true in any month, after any release.
2. Concrete proof. A specific capability, number, or scenario that makes sentence one believable. Skip it if the first sentence needed the room.
3. Optional soft CTA.An imperative without dates or prices — “Start with the free plan” works; “50% off this week” gets rejected.

Two disciplined sentences beat three cramped ones. The wasted patterns to avoid: restating the subtitle, generic superlatives (“the best,” “amazing”), and feature lists that belong in the description. Check the length as you draft with the character counter, which covers every metadata field’s limit.

7 worked examples across categories

1. Habit tracker (productivity)

Weak:“Build better habits with the best habit tracker. Download today!”

Strong:“A habit tracker for people who quit habit trackers. Three habits max, one tap to log, and a weekly review that takes ninety seconds. No streak shaming.”

2. Expense splitter (finance)

Weak:“Split expenses easily with friends. Try it now!”

Strong:“Made for flatmates and trips: split unevenly, track who paid, settle in one transfer at the end. Works without anyone else installing the app.”

3. Sleep app (health)

Weak:“Sleep better tonight with relaxing sounds. New content weekly!”

Strong:“Wind-down audio engineered to end — every session fades out fully so nothing plays all night. Built for light sleepers who wake when loops repeat.”

4. Recipe manager (food)

Weak:“All your recipes in one place. Cooking made easy!”

Strong:“Paste any recipe link and get just the recipe — no life story, no ads. Scales servings, builds the grocery list, and works offline at the stove.”

5. Word game (casual)

Weak:“The most addictive word game of 2026! Play free now!”

Strong:“One hand-built word puzzle every morning, hard enough to need coffee. Finish in five minutes or carry it around in your head all day. No energy systems.”

6. Interval timer (fitness)

Weak:“The ultimate workout timer for all your training needs!”

Strong:“An interval timer you can read from across the room mid-burpee. Huge digits, voice cues, and presets for HIIT, EMOM, and Tabata. Set up a circuit in under a minute.”

7. Language flashcards (education)

Weak:“Learn languages fast with fun flashcards. Start your journey today!”

Strong:“Flashcards built from the 2,000 words that cover most real conversations. Spaced repetition decides what you review; ten minutes a day does the rest. Works offline.”

The shared pattern: name the person, state the difference in concrete terms, and let specificity do the persuading. Superlatives tell users you have nothing specific to say.

Localizing promotional text

Every storefront language gets its own 170 characters, and because the field is conversion copy, weak localization shows immediately — a literal translation of an English idiom reads like a vending-machine label. Three rules carry most of the value:

• Rewrite, don’t translate.The positioning survives; the phrasing shouldn’t. Sentence rhythm, formality level, and what counts as a credible proof point all differ by market.
• Counter-cycle the seasons. A January fitness push is mid-summer messaging in Australia and Brazil; tax-season copy lands in April in the US but differently almost everywhere else. Per-locale promotional text is how one app stays seasonally right in thirty markets.
• Keep every locale in the consistency sweep. A feature claim you soften in English but forget in German is still a 2.3.1 mismatch. Push My App’s 14-language localization regenerates the field per market in one pass, which keeps the sweep tractable.

The 4 rejection traps under guideline 2.3.1

Trap 1 — Time-bound offers

“Sale ends Friday,” “free for the first 1,000 users.” Metadata must stay accurate indefinitely; dated claims can’t. Campaigns belong in-app or in push — the rejection index shows how these letters read.

Trap 2 — Claims you can’t defend

Rank claims you don’t currently hold, invented user counts, “featured by Apple” without the feature. Cite only what’s verifiable right now, in the storefront where it’s shown.

Trap 3 — URLs and contact details

Links, emails, and phone numbers don’t belong in promotional copy — Apple provides the Support and Marketing URL fields for exactly that, and reviewers move anything else back there via rejection.

Trap 4 — Promises the listing doesn’t back up

Reviewers read promotional text, description, and screenshots as one claim set. A feature that appears only in promotional text reads as misleading metadata — when you update one field, sweep the other two for consistency.

Testing it: CPPs and timed comparisons

Apple’s native A/B tool won’t help here — product page optimization varies only icons, screenshots, and previews. Promotional text gets tested two other ways:

• Custom product pages as parallel variants. Each CPP (up to 70 per app) carries its own promotional text. Point different traffic sources — an Apple Ads ad group, a newsletter link, an assigned search keyword — at pages whose copy matches that intent, and compare conversion per page in App Analytics.
• Timed comparisons on the default page. Run copy A for two weeks, copy B for two, holding screenshots and pricing still, and compare page-view-to- install conversion. Cruder than a true split test, but the field’s zero cost of change makes the cadence sustainable.

Either way, the rule is one moving part at a time — a promotional text test during an icon test measures neither.

When to actually update promotional text

1. Seasonal emphasis. The same app leads with different value in January (fitness), April (taxes), and September (school). Rotate the lead, keep the positioning.
2. Release-day callouts. The binary ships, and promotional text acknowledges the headline feature the same hour — no review lag between the feature existing and the listing saying so.
3. Scheduled copy tests.The two-week comparison cadence above, logged somewhere you’ll actually check.

The anti-pattern is reflexive churn: weekly rewrites with no measurement window make every change unreadable. Two weeks minimum per variant; the ASO playbook slots this into the broader iteration loop.

And add one recurring check: stale promotional text is worse than plain promotional text. Copy that still celebrates a “new” feature from three versions ago, references a season that ended, or names a price that changed quietly tells visitors the listing is unattended. A thirty-second read-through belongs on every release checklist — it’s line one of the metadata pass in the 9-minute pre-flight checklist.

Frequently asked questions