Rebuilding Two Store Listings Around How Each Store Actually Ranks
Rewrote the iOS and Google Play listings against each store's own ranking model: title-weighted metadata with a non-duplicating keyword field on Apple, a semantically complete long description on Play. This one is the method and the calls behind it, not a measured listing delta.
- Product
- B2B team-management SaaS for small and mid-size companies
- Model
- $10 per user per month, self-serve, with a companion mobile app
- Timeframe
- June 2026
The situation
Both store listings were written for people who already knew the brand, which left the app invisible to anyone searching for the problem it solves.
The approach
- Wrote the iOS metadata against Apple's field weighting, title over subtitle over keyword field
- A character in the title is worth more than a character in the keyword field, so the non-brand half of the title went to the highest-volume category term instead of to a word like "Mobile" that adds no keyword value
- Apple ignores terms repeated across the three fields, so the keyword field was written to add terms the title and subtitle did not already carry
- Wrote the Play long description for semantic retrieval rather than keyword density
- Play's Gemini-powered guided search reads the whole long description for semantic relevance, so a thin description cannot rank on any term regardless of how good the title is
- Covered the use cases, the feature names, and the buyer's own vocabulary in prose a person would still want to read
- Made the metadata serve the warm audience and cold search at once
- The title carries the category language someone searching the problem would type
- The subtitle carries the proof that this is the official product, which is what makes a stranger trust the result
- Cut keyword terms for capabilities the mobile app did not have
- Two of the most tempting terms named features that existed in the desktop product but not on the phone
- Ranking for a feature the app does not have buys an install that uninstalls, and a review that says so
- Briefed the screenshots as the conversion lever, and specified portrait for Play
- Portrait shows three screenshots in Play search results where landscape shows one
- The first screenshot has to answer "why download this" before anyone taps through to the description
- Queued Apple Custom Product Page keyword linking as the step after the base listing
- Keyword-field terms can be assigned to Custom Product Pages, so a specific search intent lands on a page written for that intent
- It multiplies a listing that already converts and does nothing for one that does not, which is why it was sequenced second
The two stores do not rank the same thing, which is why one set of copy pasted into both is the common way to lose here. Apple ranks a short, weighted set of fields and penalizes repetition across them, so the work is allocation: decide what each of the three fields is for and never spend a character twice. Play reads the long description semantically, so the work is coverage: say the whole thing, in the buyer's words, at length. The judgment call that mattered most was subtractive. The strongest keyword candidates named capabilities the mobile app did not ship, and taking them out costs ranking on paper while protecting the install that would have churned on open. This study has no after. The listings were written and handed to the client's marketing team, and no before-and-after on impressions, store conversion rate or installs was ever measured against them. Read it as the method and the reasoning behind each call, not as a result.
We help consumer apps and products reach their next phase of growth.
More case studies
A CRO Program That Consistently Beats Benchmarks
Building B2B Product-Led Growth From Zero
Nearly Tripling Paid LTV:CAC in Four Months
Scaling Ad Spend 3x and Reaching Profitability in Seven Months
Growth notes, in your inbox
New writing on subscription growth and product-led strategy when I have something worth saying. No spam.