Guidelines For Search - App Store Training - BaseLine
Guidelines For Search - App Store Training - BaseLine
https://baseline.apple.com/training/evaluations/3/guidelines 1/29
2/23/25, 10:45 AM Guidelines for Search - App Store Training — BaseLine
SS1.png •628 KB
Updated 7/18/2024
Added more examples for Complex queries covering
Updated 6/24/2024
Added more examples for Complex queries covering
Updated 6/20/2024
Added more examples for Complex queries covering
Updated 6/6
Added more examples for Similar App queries under Complex Queries section
Updated 5/16/2024
Added section on Complex Queries
https://baseline.apple.com/training/evaluations/3/guidelines 2/29
2/23/25, 10:45 AM Guidelines for Search - App Store Training — BaseLine
Updated 2/12/2024
Added Section on Vision App Store Tasks
Updated 8/2023
Added Very Ambiguous Queries section under Common Points of Confusion for queries like [..], or [,].
Updated 7/18/2023
Added additional examples of story results for App Navigational queries: [angry birds], [warzone]
Updated 6/26/2023
Added examples under App Navigational for [netflix] query.
Introduction
In this document, we explain relevance rating guidelines for App Store content. You will use the BaseLine tool to make these
ratings.
Mandatory Comment
Each rating must be explained in the comment box [even if “optional” is indicated - rating comments are always mandatory].
The comment should be concise and must only explain why the rating chosen is the correct one in application of the guidelines.
See example below:
SS2.png •65.8 KB
Table of Contents
Introduction
Query Intent: App Navigational
Query Intent: Functional
Query Intent: Dev Navigational
Common Points of Confusion
https://baseline.apple.com/training/evaluations/3/guidelines 3/29
2/23/25, 10:45 AM Guidelines for Search - App Store Training — BaseLine
Examples: [facebook], [twitter], [instagram], [pinterest], [chalk draw free!], [audible], [gmail], [angry birds] Note: this query currently
points to the Angry Birds 2 game, since AB Classic was removed from the store, [candy crush], [minecraft free], [imovie free].
SS3.png •637 KB
A query with both app and developer intent is classified as App Navigational when that developer has one canonical app associated
with them, i.e.Facebook, Amazon, Bank of America.
https://baseline.apple.com/training/evaluations/3/guidelines 4/29
2/23/25, 10:45 AM Guidelines for Search - App Store Training — BaseLine
franchise.)
[gwent] → Card Battle Games → Good (The story doesn't have the intended game but has many high quality card battle games
similar to the intent.)
[brawl stars] → What we’re playing now → Acceptable (The intended app itself is not included in the story. However, The story
contains some games that are partially similar.)
[warzone] → Get Ready for Call of Duty: Warzone Mobile → Excellent (Although the app is in pre-order, the story still contains the
intended app, and is specifically about the intended app.)
Add Ons → Excellent if the add-on is for the intended app, Good if it's for another similar functionality app that is good quality.
Acceptable if the add-on is for an app that is low to poor quality, or if the add-on belongs to an app that satisfies a secondary intent
like a complimentary app or an app with limited functionality.
Bundles → A bundle can get Perfect. Example: [cut the rope] → Cut the Rope for iPhone - Value Pack -- Perfect; contains the
intended original app along with others apps in the series. If the bundle contains the Navigational result, along with other unrelated
apps created by the same developer, it should be rated Excellent. If the bundle contains only apps with similar functionalities that
might satisfy the intent it shall be rated according its relevancy as Good or Acceptable.
Other apps by the same developer → Excellent to Acceptable (Example: for the query [gmail] → Google Translate is Acceptable,
since Google is considered a canonical developer (canonical here defined as widely recognized and authoritative.) App by the
same developer and in the same realm -> Excellent; App by the same developer but in different realm from the intended app ->
Good or Acceptable; App by the same developer but hasn’t been updated for over 3 years and has poor or no reviews ->
Acceptable.
Apps by other developer(s) with similar functionalities that are good quality → Eligible for Excellent
Apps by other developer(s) with similar functionalities that are good quality but have no reviews and ratings → Eligible for
Excellent.
Apps by other developer(s) with similar functionalities but are of lower quality → These apps can be rated Good or
Acceptable. You can determine quality by looking at the developer screenshots within the App Store Preview.
Apps by other developer(s) but have not been updated for 3 or more years and have poor or no reviews → Acceptable. For
App Navigational queries, this rule does not apply to the following scenarios:
1. The result is the intended app. The intended app should receive a rating of Perfect regardless of updates.
2. The result is a Story/Collection. Stories should not be demoted to Acceptable if the app or apps within it have had no updates for
3 or more years and have poor or no reviews.
You can check the last update date by looking at 'Version History' on the App Store Preview Page.
If the app has had no updates for 3 or more years and has positive reviews, do not consider the lack of updates when rating. Rate
based on the relevance and quality of the app.
Apps by other developer(s) but not all the functionalities, and/or lower quality → Good or Acceptable
Apps that keyword match the query, but may not satisfy the user intent → Acceptable
Unacceptable: Off Topic: A piece of content that is unrelated to the query, and does not satisfy a primary nor a secondary intent of
the query.
Unacceptable: Spam: A piece of content that contains obvious keyword spam and/or misleading description or screenshots.
Unacceptable: Other: Any other unacceptable content that is not included above.
Example ratings:
Query: [gmail]
Gmail - Email by Google (App Navigational) → Perfect
App of the day story - featuring Gmail → Excellent
Google LLC (Developer Card) → Excellent
“Clean Out Your Inbox (for Real This Time)" (Story/Collection) → Good. Contains the navigational app Gmail along with other
relevant email apps.
myMail box: email client app (App by another developer, with partially similar functionality) → Good. This is email client app that
enables users to work with different mail accounts. This app satisfies a secondary intent.
Yahoo Mail - Stay Organized (Email app by another developer with similar functionalities) → Excellent
Google Translate (another unrelated app by Google) → Acceptable
https://baseline.apple.com/training/evaluations/3/guidelines 5/29
2/23/25, 10:45 AM Guidelines for Search - App Store Training — BaseLine
Flashlight ϟ → Acceptable. This app has had no updates for more than 3 years has mostly poor reviews.
Query: [netflix]
Netflix → Perfect
Hulu: Watch TV shows & movies → Excellent
Krispee Street → Good. This is a game by the same developer but would satisfy a secondary intent. This game and the intended
app are not similar.
Query: [whatsapp]
WhatsApp Messenger (App Navigational) → Perfect
Keep in Touch (Story/Collection) → Good. Contains the navigational app WhatsApp along with other relevant messaging apps.
Lock WhatsApp web dual chat (add-on) → Acceptable. This is an add-on, but for an app that is complimentary to the intended app
(parent app Messenger Duo for WhatsApp satisfies a secondary intent).
Query: [amazon]
Amazon - Shopping made easy (App Navigational) → Perfect.This is the canonical app associated with Amazon.
AMZN Mobile LLC (developer card) → Excellent
Amazon Music→ Excellent. The query [amazon] is considered an App Navigational query, but it also has strong developer intent.
This is another app by the intended developer and eligible for Excellent.
Amazon Prime Now→ Excellent. Another app by the intended developer and eligible for Excellent.
Kindle→ Excellent. Another app by the intended developer and eligible for Excellent.
"TRY SOMETHING NEW - 3 Must-Try Tips for Amazon Prime Video" (Story/Collection) → Good. This is a Story about Amazon
Prime Video, providing users with tips and tricks on how to best use the app.
eBay→ Good. Satisfies a secondary intent of the query.
"Amazon FreeTime Unlimited Subscription" (Free trial offer, add-on) → Good. This is a subscription for another Amazon app.
NOOK - Read Books, Magazines, Newspapers & Comics→ Unacceptable: Off Topic.
Query Type:
Query: [drawApp Navigationalepic]
a stickman
If Query
the user does
Type: not specifically ask for a free or paid version of an app, both should be rated Perfect.
Functional
Apple Arcade
Query Type:
Common Points Functional
of Confusion
Definition: Queries
Disagreement in this
with the category
Query Type express intent for apps with a particular function or theme.
Examples: [free apps], [top free games], [girls games], [rpg], [action games], [news], [cloud storage], [wallet], [fitness], [batman],
Researching Query Intent:
[spongebob games], [scanner], [harry potter], [libros] (query in Spanish, translates to "books") etc.
Primary and
Functional Secondary
queries can intent
also be incomplete queries that do not have a clear intent: [my], [the], [in], [zo] etc.
ThisSpell
alsoCorrection
includes:
-Native Apple apps and functions that cannot be removed and re-downloaded from the App Store such as: [app store], [airplay],
Queries with multiple likely intents
[apple pay], [carplay], [siri], [game center], [icloud] etc.
Adult Content
-Queries expressing intent for apple devices such as: [homepod mini], [iphone], [appletv remote] etc.
Problem: Other
Foreign Language
Unacceptable: Spam
Common Misunderstandings
SS4.png •727 KB
Popular and well-known apps that clearly satisfy the primary intent of the query → Perfect
Editorial Item (Story/Collection) →A story containing multiple apps, and all of those apps satisfy the intent can get Perfect.
App/Game of the Day Stories, as well as other stories about a single app that satisfies the intent can be rated Excellent. Stories
containing apps that satisfy a secondary intent or partially satisfy the intent can be rated Good. Stories containing apps that poorly
satisfy the intent or only slightly satisfy the intent can get Acceptable.
Apps that are less popular/known but still clearly satisfy the intent of the query *→ Excellent
*Bundles→ Rate the same as the most relevant app in the bundle
Apps that satisfy the intent but have no reviews and ratings → Eligible for Excellent. These apps can be rated Good or
Acceptable if the app appears to be lower quality. You can determine quality by looking at the developer screenshots within the App
Store Preview.
Apps that satisfy the intent but are of low or poor quality → These can be rated Good or Acceptable. You can determine quality
by looking at the developer screenshots within the App Store Preview.
Apps that satisfy a primary or secondary intent but have not been updated for 3 or more years and have poor or no
reviews → Acceptable. For Functional queries, this rule does not apply to the following scenario:
https://baseline.apple.com/training/evaluations/3/guidelines 7/29
2/23/25, 10:45 AM Guidelines for Search - App Store Training — BaseLine
-The result is a Story/Collection. Stories should not be demoted to Acceptable if the app or apps within it have had no updates for
3 or more years and have poor or no reviews.
You can check the last update date by looking at 'Version History' on the App Store Preview Page.
If the app has had no updates for 3 or more years and has positive reviews, do not consider the lack of updates when rating. Rate
based on the relevance and quality of the app.
Apps that partially satisfy the intent, or lower quality → Good, or Acceptable
Add Ons → Excellent if the add on is for an app that satisfies the functional intent and is of high quality. Good if the add on is for an
app that is less popular. Acceptable if the add on is for an app that is poor quality, or satisfies a secondary intent.
Developer Card → Acceptable, if the developer contains apps that satisfy the query intent or if the developer name matches with
the query. (Example: [free games without wifi] → Ketchapp, [playg] → Playgendary Limited)
Unacceptable: Off Topic: A piece of content that is unrelated to the query, and does not satisfy a primary nor a secondary intent of
the query.
Unacceptable: Spam: A piece of content that contains obvious keyword spam and/or misleading description or screenshots.
Unacceptable: Other: Any other unacceptable content that is not included above.
Example ratings:
Query: [podcast]
iCatcher! Podcast Player→ Perfect
Podcast Paradise, The Daily List (Editorial Item) → Perfect. This is a curated collection of several popular, high quality podcast
apps.
Podcast Player → Excellent, not that well-known or popular.
TuneIn Radio→ Good. Contains podcasts, but doesn't seem to have the all the functionalities of an app dedicated to podcasts only;
partially satisfies the intent.
Soundcloud→ Good. Contains podcasts, but doesn't seem to have the all the functionalities of an app dedicated to podcasts only;
partially satisfies the intent.
NPR News→ Acceptable Limited to NPR podcasts only, poorly satisfies a secondary intent of the query.
The Podcast App Premium Subscription (remove all adverts from the app add-on) → Excellent. This is an add on for an app that is
popular and clearly satisfies the intent of the query.
Cloud Music Player for Clouds → Unacceptable: Off Topic. No podcast functionality.
Query: [radio]
RadioApp - A Simple Radio → Perfect
Satellite Radio Channel Guide for Sirius XM → Acceptable. App has not been updated for over 3 years and has mostly poor
reviews.
Query: [siri] - The intent for this query is not for a single app, but for apps that support Siri.
Things 3 → Perfect
Just Ask Siri (Editorial Item) → Perfect. This is a story containing several apps that are supported by Siri.
https://baseline.apple.com/training/evaluations/3/guidelines 8/29
2/23/25, 10:45 AM Guidelines for Search - App Store Training — BaseLine
Minion Rush -> Good. Game is for older kids (9+) and adults.
PUBG MOBILE -> Unacceptable: Off Topic. This game is for 12+ age group.
Query: [my]
MyFitnessPal: Calorie Counter -> Perfect. Intent is unclear. This app is well-known, good quality, and matches with the query in the
first word of the app title. It could be a possible primary intent.
My Talking Angela -> Perfect. Intent is unclear. This app is well-known, good quality, and matches with the query in the first word of
the app title. It could be a possible primary intent.
My Singing Monsters -> Perfect. Intent is unclear. This app is well-known, good quality, and matches with the query in the first
word of the app title. It could be a possible primary intent.
My Town: Sticker Book -> Excellent. Intent is unclear. This app is less popular but is good quality, and matches with the query in
the first word of the app title. It could be a possible primary intent.
Query: [game]
Pokemon Go -> Perfect. Popular app that satisfies broad range of users.
Shoe Race → Good. App targets a specific group of users
Game of the Day - Candy Crush -> Excellent. This is a story about a single game that satisfies the intent of the query.
The Wheels Man On Happy Bike → Acceptable. This game has not been updated for over 3 years and has mostly poor reviews.
Definition: A query that refers to a specific developer which the user queries for by name.
Examples: [rovio], [dumadu], [disney], [apple], [google], [lightricks], [gameloft], [ketchapp], etc.
ss5.png •660 KB
Note: [top free games] does not qualify as a developer query even though there exists a developer named top free games. The
reason is that the vast majority of users entering this query on the App Store are looking for popular free games, not for a developer
called Top Free Games. Similarly, queries like [photo and video editor], [emoji icons free] are functional, not developer queries,
despite the fact that there are developers called Photo & Video Editor and Emoji Icons Free, and Emoji Icons Free and Color
Texting Messages Pro Studio.
https://baseline.apple.com/training/evaluations/3/guidelines 9/29
2/23/25, 10:45 AM Guidelines for Search - App Store Training — BaseLine
*Note:
Relevant apps that have had no updates for 3 or more years (36 months or more) and have poor or no reviews should receive a
rating of Acceptable. For Developer Navigational queries, this rule does not apply to the following scenarios:
1. The result is an app by the intended developer. Apps by the intended developer that have had no updates for 3 or more years
can get a rating of Excellent.
2. The result is a Story/Collection. Stories should not be demoted to Acceptable if the app or apps within it have had no updates for
3 or more years and have poor or no reviews.
You can check the last update date by looking at 'Version History' on the App Store Preview Page.
If the app has had no updates for 3 or more years and has positive reviews, do not consider the lack of updates when rating. Rate
based on the relevance and quality of the app.
Example ratings:
Query: [savage interactive]
Savage Interactive Pty Ltd (Developer Card) → Perfect
Procreate Pocket (app by the developer) → Perfect
Meet The Developer - A Creation Story → Perfect.This is a story about Savage Interactive Pty Ltd, and the road to developing the
Procreate app.
Drawing Desk: Draw & Paint Art→ Excellent. An app by another developer, offering similar functionality as the intended
developer's app.
Query: [sega]
Sega (Developer Card) → Perfect
Made by SEGA → Excellent. Contains a couple of popular games by Sega.
Retro Revival: Sonic Dash 2 → Excellent. Contains the game ‘Sonic Dash 2: Sonic Boom’.
“GAME OF THE DAY - Mascot Mashup" → Excellent. Contains the "SEGA Heroes: Match-3 RPG Quest" game.
Sonic Dash→ Perfect
Super Mario Run→ Good. A popular side-scrolling, auto-runner game slightly different from Sonic the Hedgehog games, satisfies a
secondary intent.
Taster Bundle (Add-on) for Sonic Dash 2 → Excellent. This is an in-app purchase option for a game by the intended developer.
Complex Queries
https://baseline.apple.com/training/evaluations/3/guidelines 10/29
2/23/25, 10:45 AM Guidelines for Search - App Store Training — BaseLine
When working on App Store Search eval, you may encounter complex queries that are longer than the standard search and
structured like a sentence. Because of their structure, understanding what is being asked for will be key in judging results for these
queries.
Complex queries could ask for multiple functions or features. If a result only partially satisfies what is being asked for, then it should
not be given a rating that is higher than Good.
Example:
This query is asking for abc alphabet learning apps with tracing and writing letter function. A game like LogicLike: Kids Learning
Games does have a ABC learning function, but does not include a function for writing or tracing letters. Although this result is good
quality, it only partially satisfies what is being asked for. Therefore, a rating of Good is adequate. ABC Kids - Tracing & Phonics
however, has all that is being asked for. It is an alphabet learning app with a letter tracing function that teaches kids how to write
letters. This is also a popular app and can receive a rating of Perfect.
Some complex queries will ask for apps containing specific tv or movie content, like the query [apps to stream series mr. bean]. The
easiest way to check the availability of the video content on a given platform is to search for the result on a reliable search engine
like Google to verify. For example, if Hulu: Stream TV shows & movies was the result for the mentioned query, you can do a search
on Google for 'where to watch mr bean series' or 'mr bean series hulu' to confirm if the intended tv or movie content is available on
Hulu. Streaming guides like justwatch.com can also confirm where you can watch the intended content.
Examples:
Results:
Hulu: Stream TV shows & movies → Perfect (Hulu has the Mr. Bean tv series. Perfectly satisfies the user intent.)
Xfinity Stream → Perfect (This platform has the TV series Mr. Bean. Satisfies the primary intent.)
Max: Stream HBO, TV, & Movies → Good (This platform has only Mr. Bean: The Animated Series, but the animated series is not as
popular as the TV series and would partially satisfy the user intent.)
Peacock TV: Stream TV & Movies → Off-Topic (Peacock TV does not have the intended content.)
Results:
Crunchyroll → Perfect (This platform is for streaming anime shows and movies. Has the intended content and fully satisfies the
intended Dragon Ball anime.)
Tubi: Movies & Live TV → Good (This video platform has a live action movie called Dragon Ball: Comienza La Magia that is not
well known but may satisfy a secondary intent.)
Peacock TV: Stream TV & Movies → Off-Topic (Peacock TV does not have the intended content.)
DRAGON BALL LEGENDS → Off-Topic (This is a Dragon Ball game but is not releated to watching Dragon Ball series.)
Some complex queries will ask for apps or games released on a specific year. You can search on a reliable search engine like
Google to verify if an app result was released on the mentioned year. For example, if Pico Tanks was the result for the query [battle
games launched in 2020], you can do a search on Google for 'pico tanks release date' to confirm if the game was launched in 2020
as specified in the query.
https://baseline.apple.com/training/evaluations/3/guidelines 11/29
2/23/25, 10:45 AM Guidelines for Search - App Store Training — BaseLine
Results:
Pico Tanks → Perfect (This is a popular and good quality battle arena game, launched in 2020.)
Towers Defense 3D Battle Games → Excellent (This is a good quality battle game launched in 2020, but is less popular.)
Clash Royale → Good (This is a battle game, but was launched in 2016. Partially satisfies the intent.)
Game of War - Fire Age → Acceptable (This is a battle game launched in 2020, but has poor reviews.)
Play Golf 2020 → Off-Topic (This is a game launched in 2020, but is not a battle game and would not satisfy the main genre of
game that is being queried for.)
Queries for similar apps are complex queries asking for apps that are like or similar to another app. For example, the query [apps
that are similar to doordash] is asking for apps similar to DoorDash - Food Delivery, like Uber Eats: Food Delivery or Grubhub:
Food Delivery.
Please Note: If the result for a Similar App query is the mentioned app (If DoorDash - Food Delivery was the result for [apps that
are similar to doordash]), the mentioned app should receive a rating of Acceptable.
Examples:
Results:
Adobe Fill & Sign-Form Filler → Perfect (Adobe Fill & Sign is a high quality and well-known app that shares the same core
functionality as Docusign.)
eSign App - Sign PDF Documents → Excellent (eSign App shares the same core functionality as Docusign, but is less popular.)
CamScanner - PDF Scanner App → Good (CamScanner does have a document signing function, but is missing some capabilities
that Docusign has like allowing others to sign the documents, sending reminders to sign, and getting notifications on who and when
the documents are signed)
Docusign - Upload & Sign Docs → Acceptable (Although Docusign is the app mentioned in the query, the user is asking for a
similar app, not the app itself. For this query, Docusign is only eligible for a rating of Acceptable.)
Results:
Apple Music → Perfect (Apple Music is a high quality and well-known app that shares the same core functionality as Spotify.)
Youtify - Music & Playlists → Excellent (Youtify is a high quality and well-known app that shares the same core functionality as
Spotify, but is less popular.)
SoundCloud: Discover New Music → Good (SoundCloud is also a music streaming app, but primarily used for discovering new
music and artists and is more limited in its catalog. This app would satisfy a secondary intent.)
Essential Music Apps (Story) → Good (Although this story contains some apps that are similar to Spotify, many of the other apps
are for music creation and learning which are not similar. If the story had apps that were all similar, it would be eligible for Perfect.)
Cloud Music Player - Listener → Acceptable (Cloud Music Plater is an offline music player and downloader. Although you can play
music offline which is a similar function included in Spotify, it is barely similar as it lacks other important functionalities.)
Some queries will ask for the top, popular, or best apps in a particular country or region. For a query like [multiplayer games popular
in pakistan], what you can do is click on the result and go to the App Store Preview. In the App Store Preview, go to the link above
and change the country 2-letter code in the link to the mentioned country’s code. 2-letter country codes (US for United States, JP
for Japan, etc.) can be found by searching on a reliable browser like Google. In this example, you would switch ‘us’ for ‘pk’ in the
link:
Once you refresh the link, the App Store Preview will show you the ratings, reviews, and ranking of that app in the specific market:
https://baseline.apple.com/training/evaluations/3/guidelines 12/29
2/23/25, 10:45 AM Guidelines for Search - App Store Training — BaseLine
Results:
Roblox → Perfect (This is highly popular multiplayer game, available in Pakistan.)
Brawl Stars → Perfect (This is highly popular multiplayer game, available in Pakistan.)
AFK Football: RPG Soccer Games → Good (This is a multiplayer game available in Pakistan but is not popular. Only 2 reviews in
Pakistan market.)
Minion Rush: Running game → Off-Topic (This is a runner game that is not multiplayer.)
Please note that certain apps may not be available in the country specified in the query. If the App Store Preview does not load
once you change the 2-letter country code to the specified country or region and refresh the link, it likely means the app is not
available in that country. In the screenshot below, the country code for Hulu: Stream TV shows & movies was changed from US
to GB (GB being the country code for the United Kingdom). The App Store Preview did not load, and screen remained in the
'Connecting...' screen:
https://baseline.apple.com/training/evaluations/3/guidelines 13/29
2/23/25, 10:45 AM Guidelines for Search - App Store Training — BaseLine
You can search on a reliable web browser and verify whether the app is available in the specified country, such as 'is Hulu available
in the UK' on Google. If the app is not available in the specified country, then the rating should be Unacceptable: Off Topic. For
example, for the query [best tv streaming apps in uk], the app Hulu: Stream TV shows & movies would receive a rating of
Unacceptable: Off Topic as the app is not available in the UK.
Results:
Netflix → Perfect (This is a popular and High quality tv and movie streaming app in the UK, and is available in the market.)
YuppTV - Live TV & Movies → Good (This is a tv and movie streaming app available in the market, but is less popular and lower
quality according to the ratings. There are better options for tv and movie streaming in the market that can be considered 'best'.
YuppTV can satisfy a secondary intent.)
Hulu: Stream TV shows & movies → Off-Topic (This app is not available in the UK market.)
Some complex queries will ask for apps with a specific cost, like [photo app below $5] or [video editing apps under $10]. For these
cases, If the app result is within the price range that the user is asking for (clearly marked under the stars in the App Store
Preview), and does not require purchasing a membership or a ‘PRO' version to actually use the app, it can be eligible for a rating of
Perfect. The app would also have to satisfy the functional intent that the query is asking for (‘photo app’ or 'video editing apps' in
the mentioned queries).
If the app does fulfill the functional intent, but is above the price range, then the app can be given a rating of Acceptable.
Results:
Sketchbook® → Perfect (This is a popular sketching app that is free and satisfies the functional intent.)
Procreate Pocket → Acceptable (This is a sketching app, but is $5.99 according to the App Store Preview.)
**(This is another location specific complex query but is not asking for top apps. Please note that this is query is only asking for
apps for local travel in England, so it does not require a change in the country code like the example above.)
Results:
Booking.com: Hotels & Travel → Excellent (This is not an app specifically for England, but is useful for local travel in the country.
Popular and high quality.)
British Airways → Excellent (This is not an app specifically for England, but is useful for local travel in the country. Popular and
high quality.)
TrainPal: UK& EU train tickets → Excellent (This is an app for purchasing railway tickets in UK and EU. Relevant and high quality
but less popular.)
London Travel Guide → Good (This is a travel guide but only for London and thus can only partially satisfy the intent. Good quality.)
Business Traveller (UK) → Acceptable (This is a magazine for the corporate traveller in UK. Very specific and can only satisfy a
secondary intent.)
French Travel Phrases & Words → Off-Topic (This is an app relevant for travel to France not UK.)
Some complex queries will require you to check specific sources on the web to figure out what is intended. For example, the query
[apps that won apple design award for innovation in 2024] will require you to search for the app that won the Apple Design Award
for Innovation in 2024. By searching 'apple design awards 2024' on a reliable search engine like Google, you can find the winners
of those awards for that particular year. Procreate Dreams was the Innovation winner for 2024 and would be the correct result.
Query: [apps that won apple design award for innovation in 2024]
Results:
https://baseline.apple.com/training/evaluations/3/guidelines 14/29
2/23/25, 10:45 AM Guidelines for Search - App Store Training — BaseLine
Procreate Dreams → Perfect (This is the app that won the Apple Design Award for Innovation in 2024 and would be the intended
app.)
Lost in Play → Good (This app did not win the mentioned award, but was a finalist for that award in the same year. Satisfies a
secondary intent.)
ToonSquid → Off-Topic (This app did not win the mentioned award. Would not satisfy the intent.)
Canva: Design, Art & AI Editor → Off-Topic (This app did not win the mentioned award. Would not satisfy the query intent.)
Complex queries asking for 'highest rated', 'most reviewed,' 'most', etc.
Some complex queries will ask for superlative results and will contain words like 'highest rated', 'most reviewed,' 'most positive
reviews', 'most', 'best', etc.
The query [highest-rated free music streaming app] for example is asking for a free music streaming app with the highest rating. In
this case, you can refer to the ranking on the App Store Preview page of a result to know if it is a highly rated app, and in this case,
the highest rated free music streaming app. Spotify - Music and Podcasts is currently the highest-rated free music streaming app in
the App Store:
Spotify
Results:
Spotify - Music and Podcasts → Perfect (This is a free music streaming app that is rated #1 in Music.)
Musi - Simple Music Streaming → Excellent (Although this is a highly popular music streaming app with a free service, it is rated
#5 in Music.)
Amazon Music: Songs & Podcasts → Excellent (Although this is a highly popular music streaming app with a free service, it is
rated #4 in Music.)
YouTube Music → Excellent (Although this is a highly popular music streaming app with a free service, it is rated #2 in Music.)
YouTube: Watch, Listen, Stream → Acceptable (This app can be used to stream music, but is is primarily used for streaming
videos.)
Groovepad - Music & Beat Maker → Off-Topic (This is a music maker app. Not relevant to the query.)
**(‘greatest’ would be considered superlative. In this example, it is unclear as to what app has the greatest number of brewing
methods.)
Results:
The Great Coffee App → Excellent (Result is an app with coffee brewing guide, but may not have the greatest number of brewing
methods.)
The App Barista → Excellent (Result is an app with coffee brewing guide, but may not have the greatest number of brewing
methods.)
Acaia Coffee → Good (Result is an app with coffee brewing guide, but can only be used with the Acaia coffee scale machine.)
Nespresso → Acceptable (This app does not have a brewing guide, but does have a recipes function that may satisfy a secondary
intent.)
ChatGPT → Off-Topic (Not relevant to the query.)
https://baseline.apple.com/training/evaluations/3/guidelines 15/29
2/23/25, 10:45 AM Guidelines for Search - App Store Training — BaseLine
Some queries may ask for apps without ads. For queries like [hidden object games with no ads], only popular and high quality
games with no ads can get a rating or Perfect. Please note that the app must specifically say in the title or description that it is ads
free in order to be considered for a rating of Perfect.
On the other hand, a popular and high quality hidden object game result that does have ads but offers an in-app purchase to
remove the ads, can satisfy a secondary intent and get Good.
You can check if the app has an in-app purchase by going to the result's App Store Preview. In the Information section at the
bottom, there will be an In-App Purchases section listing the in-app purchases for that app. In the screenshot below, you can see
the in-app purchases for the app Hidden Objects - Find It Out:
image-7.png •190 KB
Query: [hidden object games with no ads]
Results:
Sherlock・Hidden Object・Match 3 → Good (The app is popular and of high quality and has hidden object aspects, but it is also a
match 3 game so it only partially fulfils the user's intent.)
Hidden Objects - Find It Out → Good (The app a popular and high quality hidden object game, but according to the reviews it does
have ads. There is an add-on to remove ads however, so it can satisfy a secondar intent.)
Hidden my game by mom → Off-Topic (The app is an escape room game, not a hidden object game.)
Some queries may ask for apps with a specific number of stars or ratings. For example, for the query [calculators with 4 stars or
more], only high quality and popular calculator apps with 4 or more stars can get a rating of Perfect. Any high quality calculator
apps that have less than 4 stars can receive a rating of Good, as these would only partially satisfy the query intent.
Result:
Calculator₊ → Perfect (This is a popular calculator with 4.7 ratings which satisfies the intent.)
Calculator → Good (A good quality calculator app but with below 4 stars (3.1 stars). Only partially satisfying the intent.)
Gauth: AI Study Companion → Off-Topic (This not a calculator app. Does not satisfy the intent.)
https://baseline.apple.com/training/evaluations/3/guidelines 16/29
2/23/25, 10:45 AM Guidelines for Search - App Store Training — BaseLine
Query: [books apps developed by kobo]
Results:
Kobo Books → Perfect (This is a book app by the intended developer Rakuten Kobo Inc.)
Barnes & Noble NOOK → Good (This is a popular book app, but not by the intended developer. Partially satisfies the intent.)
Rakuten Kobo Inc. (Developer Card) → Good (User's primary intent is for apps by the developer. Developer card would satisfy a
secondary intent.)
Books Reading & Library BookVa → Acceptable (This is a less popular book app, but not by the intended developer.)
https://baseline.apple.com/training/evaluations/3/guidelines 17/29
2/23/25, 10:45 AM Guidelines for Search - App Store Training — BaseLine
of working professionals who do shift work, so it should be rated as Good.)
My Study Life - School Planner → Acceptable (The app is a calendar and planner app but for students, so it poorly fulfills the
user's intent.)
MinimaList: To Do List &Widget → Off-Topic (The app is a to do list with no calendar functionality, so it should be rated as off-
topic.)
Vision Pro Native apps will have screenshots showing the Vision Pro UI
Because of Vision Pro's recent release (February 2nd, 2024), many Vision Pro Native apps will have little to no ratings or reviews.
In order to rate these apps more fairly, we do not want to rate these apps based on popularity. For App Navigational, Developer
Navigational, and Functional queries, We want to judge Vision Pro Native app results strictly on the relevance the query has to the
app.
When rating Functional query results, Vision Pro Native apps that satisfy the query intent will be eligible for a rating of Perfect
https://baseline.apple.com/training/evaluations/3/guidelines 18/29
2/23/25, 10:45 AM Guidelines for Search - App Store Training — BaseLine
regardless of popularity. For example, if the query is [weather], the app CARROT Weather: Alerts & Radar would satisfy the query
intent and could receive a rating of Perfect. The BaseLine screenshots confirm that this result is a Vision Pro Native weather app:
Query: [games]
Super Fruit Ninja → Perfect. This is a Native Vision Pro game in Apple Arcade. For Vision Pro native apps, we are only rating
based on relevance to the query. This app satisfies a primary intent.
What The Golf? → Perfect. This is another Native Vision Pro game. This app satisfies a primary intent.
All other apps that are not Native to Vision Pro can be rated using the guidance under the Query Type sections above. The only
change is the addition of Vision Pro Native apps and how we rate them, detailed in this section.
Apple Arcade
Apple Arcade is a game subscription service launched in September 2019, featuring carefully curated games with no ads and no in-
app purchases.
If the query intent is for one of these Apple Arcade games, all returned Arcade games are at least of Acceptable relevance.
Rating example:
Apple Arcade games have a banner at the top of the App Store Preview page like seen in the screenshot below:
https://baseline.apple.com/training/evaluations/3/guidelines 19/29
2/23/25, 10:45 AM Guidelines for Search - App Store Training — BaseLine
SS6.png •453 KB
Note:
Queries in all languages are valid, and results should be rated based on relevance. Please refer to query examples [ATH Móvil],
[ücretsiz oyunlar], and [libros] later in this document.
https://baseline.apple.com/training/evaluations/3/guidelines 20/29
2/23/25, 10:45 AM Guidelines for Search - App Store Training — BaseLine
There is no clear connection between the query and the app. There could be a partial match somewhere, but not readily apparent.
If you run into cases like this, please make sure to rate the result as Unacceptable: Off Topic.
The less likely intent of a query such as [amazon]: the primary intent here is the Amazon marketplace app, along with other apps
by AMZN Mobile LLC, and a secondary intent would be apps about the Amazon rainforest.
An app with a more restricted functionality than the primary intent. For example, for queries like [news] and [weather], the primary
intent is apps with nationwide news and weather content, while regional news and weather apps would be considered secondary
intent.
Intents that are complimentary to the primary intent. For queries requesting a specific app such as [razzle], [minecraft],
[facebook], etc., a common secondary intent are apps for cheats, wallpaper, emoticons, etc.
Spell Correction
Some queries are misspelled by our users. When this is the case, assume the query intent is for the correctly spelled query. If the
query is [mincraft], the user likely means [minecraft]. Use your best judgement.
https://baseline.apple.com/training/evaluations/3/guidelines 21/29
2/23/25, 10:45 AM Guidelines for Search - App Store Training — BaseLine
SS7.png •557 KB
SS8.png •267 KB
https://baseline.apple.com/training/evaluations/3/guidelines 22/29
2/23/25, 10:45 AM Guidelines for Search - App Store Training — BaseLine
Adult Content
All adult or pornographic content is Unacceptable: Off-Topic for non-adult and non-porn queries. For example, any adult or porn
content is Unacceptable: Off-Topic for the query [free], or [games] because the user has not expressly indicated that they want
content of this type. On the other hand, if the query is [erotica] or [porn], any adult or porn content is relevant.
Problem: Other
A problem or technical issue with the task in BaseLine that makes it impossible to judge relevance. Please notify your Support team
immediately of such cases.
ss9.png •20.8 KB
Problem: Other rating should not be used if the content is unavailable in App Store but it is still possible to determine relevance
between input and output based on the information provided in BaseLine, and by completing a side search on Google and/or on
other platforms such as App Annie.
Foreign Language
If the query is in foreign language for the market the results recalled shall be in the language of the query, language of the market,
or in English. If the app recalled is an exact it shall be rated regardless of its language.
Functional queries
Query: [libros]
Consider the above query example in the US App Store. According to Google Translate the query is in Spanish and the intent is for
"books". This is another functional query for which we should be returning apps related to books in either the language of the query
(Spanish), or the market (English). Any other book app that is not in Spanish or English should be rated Unacceptable: Off Topic.
Please reference the functional query type section above for further relevance rating guidance.
Example Results:
Amazon Kindle → Perfect. This app is in the language of the market. It is popular and clearly satisfies the intent.
El Libro Total → Excellent. This app is in the language of the query and satisfies the intent but is less popular.
Gleeph - gestion bibliothèque → Unacceptable: Off Topic. This app is in French.
https://baseline.apple.com/training/evaluations/3/guidelines 23/29
2/23/25, 10:45 AM Guidelines for Search - App Store Training — BaseLine
The query is in English, and the intended app BeautyPlus by Meitu Technology, Inc. will be rated Perfect.
An app titled Beauty Plus + Beauty Plus (not available any more) by WORKING EVOLUTION SRL would receive Acceptable for
keyword matching the query, even if the app itself is in Hebrew Italian. This app is unlikely to satisfy user intent, but it is a slightly
relevant keyword match.
Unacceptable: Spam
Unacceptable: Spam: A piece of content that contains obvious keyword spam and/or misleading description or screenshots. This
also includes a huge chunk of keywords in the Developer Name (Example below) and where the screenshot and description do not
match the app’s title.
Please see below example of obvious keyword spamming in the app title.
https://baseline.apple.com/training/evaluations/3/guidelines 24/29
2/23/25, 10:45 AM Guidelines for Search - App Store Training — BaseLine
ss10.png •989 KB
Example 2: a long list of keyword in the developer’s name
https://baseline.apple.com/training/evaluations/3/guidelines 25/29
2/23/25, 10:45 AM Guidelines for Search - App Store Training — BaseLine
ss11.png •1.31 MB
Example 3: The screenshot and description do not match the app’s title
https://baseline.apple.com/training/evaluations/3/guidelines 26/29
2/23/25, 10:45 AM Guidelines for Search - App Store Training — BaseLine
ss12.png •1.17 MB
Example 4:
Common Misunderstandings
If a result contains a keyword matching to the query, it does not always mean it is Unacceptable: Spam. Please see below
example of common misunderstanding. For this case, the result contain keyword matching in the description but the screenshots
and description are relevant to the app’s title. The app’s quality is also valid and may satisfy secondary intent of the user. Therefore,
these are not considered as spam.
https://baseline.apple.com/training/evaluations/3/guidelines 27/29
2/23/25, 10:45 AM Guidelines for Search - App Store Training — BaseLine
ss13.png •1.67 MB
https://baseline.apple.com/training/evaluations/3/guidelines 28/29
2/23/25, 10:45 AM Guidelines for Search - App Store Training — BaseLine
ss14.png •947 KB
https://baseline.apple.com/training/evaluations/3/guidelines 29/29