Show One App Feature Without Inventing the Outcome

An app feature showcase carousel should help a specific reader understand one job the app can perform. Use current interface evidence to show the starting state, the action, and the result visible inside the product. Then name what the feature does not establish.

Do not convert a capability into an unsupported outcome. A scheduling feature can show how a user creates a reminder; it cannot prove that every user saves hours or never misses a deadline. If the team has measured an outcome, preserve the population, period, method, and limitations with the claim.

Instagram supports feed posts containing multiple photos or videos as a carousel. The current publisher controls the available media count and applies one orientation to the set, so confirm the live interface before exporting rather than treating an old specification as permanent.

The ledger is the approval boundary for the creative. Product, support, and marketing should be able to inspect it without opening the design file. If a field is unknown, mark it unknown; do not let the draft fill the gap with plausible-sounding copy.

Use one ledger row per feature and freeze the product version used for the screenshots. Recheck the row after a release that changes the interface, availability, price, permissions, or result.

This fictional InboxArc example demonstrates the method; it is not a claim about a real app. Assume its product owner has verified only this fact: a user can apply a saved rule that moves matching messages into a chosen folder. The team has no evidence about time saved, inbox stress, or productivity.

The seven-frame sequence is intentionally compact. Use fewer frames when the action is simple, and confirm the current publishing limit before building a longer sequence.

1. 1

   Frame 1 — recognize the situation

   Headline: “Receipts are mixed into the inbox you use for client work.” Visual: an anonymized, deliberately staged inbox state. This names the situation without claiming how common it is.

2. 2

   Frame 2 — state the available action

   Headline: “Create a rule for messages that contain your billing address.” Show the current rule-entry screen and annotate only the field being discussed.

3. 3

   Frame 3 — choose the result

   Show the folder selector and the selected “Receipts” destination. The copy describes the configured behavior rather than promising a business outcome.

4. 4

   Frame 4 — verify the setup

   Show the review state before activation: condition, destination, and status. Include any product-specific permission or plan requirement that affects availability.

5. 5

   Frame 5 — show the observable result

   Use a staged message that matches the rule and show it in the chosen folder. Label staged data as demonstration data if a viewer could mistake it for a real account.

6. 6

   Frame 6 — state the boundary

   Copy: “This rule files future matching messages; it does not verify the contents of a receipt.” Replace the example boundary with the actual limitation documented for the feature.

7. 7

   Frame 7 — continue the same promise

   Invite the reader to inspect the current feature page or matching store product page. Do not change the promise from “file matching messages” to an unsupported result such as “reclaim your week.”

Capture the current product state rather than rebuilding a convenient approximation in a design tool. Stage accounts with fictional data, remove personal information, and keep enough surrounding interface visible for the highlighted action to make sense.

Annotations should identify where to look; they should not imply functionality that is absent from the screen. If a tap, transition, or background process cannot be shown in one still, use adjacent frames and describe the gap plainly.

Preview the exported sequence on a phone. Meta's current help notes that the chosen square, portrait, or landscape orientation affects every item in the carousel. Decide the crop at the sequence level, then verify that labels, system text, and required disclosures remain readable.

A benefit sentence is useful when it translates a verified capability into a relevant user action without skipping to a guaranteed result. “Route matching receipts into one folder” stays close to the observed behavior. “Never lose another expense” introduces an outcome the rule alone cannot guarantee.

Quantified examples need a traceable source. Keep the numerator, denominator, observation period, product version, and relevant cohort with the claim. A screenshot containing demonstration values is not evidence that customers achieved those values.

Comparisons need the same discipline. Do not write “faster,” “only,” or “better than” unless the comparison has a defined alternative and a current, repeatable method.

The next page should continue the exact story in the carousel. A generic homepage can force the reader to rediscover the feature. A focused product page, help article, or app-store product page can preserve the audience, capability, and visual language they just saw.

Apple documents custom product pages that can use distinct screenshots, app previews, and promotional text and be shared with a unique URL after review. When that capability fits the campaign, align the carousel with the approved custom page rather than promising a feature that the destination does not show.

Test the link on the device and in the region used by the intended audience. Confirm that fallback behavior, consent screens, store availability, and deep-link destinations do not break the journey.

For a professional Instagram account, use the metrics currently available in Insights for the published post and record their definitions from the interface. For an iOS destination, App Store Connect campaign links can associate eligible store activity with a campaign token once Apple's reporting conditions are met.

Keep platform events separate. An Instagram view, a profile action, a store product-page view, and a download are different observations collected by different systems. A change between them can suggest a place to investigate; it does not by itself prove the carousel caused the download.

Compare versions only when the audience, destination, publication window, and measurement definitions are sufficiently similar. Document distribution changes or tracking gaps before deciding that a hook or frame order won.

Once the ledger and approved screenshots exist, AttentionClaw can draft a branded carousel sequence from that context. This can reduce blank-page work and make alternate hooks or layouts easier to review.

AttentionClaw does not verify that a screen is current, substantiate an outcome, approve a store page, publish through Instagram, or attribute a download. Product and marketing owners still review every frame against the ledger and use the current platform interfaces for publishing and measurement.