Branded Keyword Separator

Branded searches are navigation, not discovery — someone typing your app’s name already decided to install it, so those impressions and conversions measure brand strength, not ASO effectiveness. Blend them into your keyword performance and everything inflates: conversion rates look stellar, keyword coverage looks broad, and a listing that wins zero new users from generic search can still look healthy on paper. Splitting the buckets is what makes “how is our ASO actually doing?” answerable — the generic bucket is the honest answer.

The same logic runs through paid UA. In Apple Search Ads, own-brand terms convert at defensive-play rates while generic terms measure real acquisition efficiency; averaging them produces a blended CPA that flatters the account. And competitor-branded terms in your search reports are strategy signals — evidence of users comparison-shopping — not keywords to stuff into metadata, where Apple’s guidelines prohibit other companies’ trademarks.

Brand detection is fuzzier than a simple text match. Users type brands with spacing variants, missing punctuation, and habitual misspellings — and most branded queries are not the bare name but brand-plus-modifier phrases like “spotify offline” or “duolingo for kids”. The separator normalizes case and spacing and flags any keyword containing a listed brand token, so embedded and combined forms land in the branded bucket instead of slipping through as false generics.

The quality of the split follows the quality of your brand list, so build it generously: every competitor you know, product names distinct from company names, abbreviations users actually type, and the misspellings your search-term reports reveal. One caution for the other direction — brands that collide with dictionary words (“Notion”, “Bear”) will pull some genuinely generic phrases into the branded bucket, so give ambiguous matches a quick manual pass before acting on the totals.