Launching an Apple Search Ads campaign in Mainland China is only the first step. Once an app and advertiser have received the necessary approval, the more difficult task is identifying the searches that matter, understanding local competition and converting paid visibility into valuable users.
Apple now refers to the service as Apple Ads, but Apple Search Ads and the abbreviation ASA remain widely used by app marketers. It focuses on search results campaigns, where advertisers bid on keywords to reach people actively searching for apps in the Chinese App Store.

Apple Search Ads are often effective because the advertisement appears while a user is actively looking for an app. However, this high level of intent does not make every keyword valuable.
According to Apple’s explanation of search results ads, the auction considers both keyword bids and the relevance of the app to the user’s search. If Apple does not consider an app relevant, it will not enter the auction regardless of the amount the advertiser is prepared to pay.
Success therefore depends on four connected factors:
Optimising only one part of this system can produce misleading results. A campaign can achieve a low cost per install while attracting users who never complete onboarding or pay. Another campaign may have a higher CPA (cost-per activation) but generate better retention and substantially more revenue.
Campaign optimisation begins before the first bid is placed. The app should already have a credible Chinese App Store presence, a working measurement system and a clear commercial objective.
The product page should be properly localised into Simplified Chinese. Localisation should cover:
It should communicate the functions available in the Chinese version rather than promising features that have been removed or changed for regulatory, infrastructure or commercial reasons.
The product itself should also be tested under real conditions in Mainland China. The following need to function reliably:
Paying to acquire users before these issues are resolved can increase negative reviews and waste the campaign budget.
Finally, the advertiser should define the event that represents success. Depending on the product, this could be a completed registration, trial start, paid subscription, purchase, booking, qualified lead or retained user. Cost per install can remain a useful operational metric, but the bid strategy should ultimately reflect the value created after installation.
If the app has not yet been approved for advertising, or its Chinese licences and content requirements remain unclear, follow our technical Apple Search Ads China guide before beginning performance optimisation.
Competitor research should begin with the Chinese App Store rather than a global list of well-known brands. The strongest competitor overseas may have limited visibility in China, while a local app with an unfamiliar English name may dominate the relevant searches.
The competitor set should include three types of app:
For each app, review the following information:
| Area | What Should Be Examined? | Why Does It Matter? |
| Organic visibility | Category rank and ranking for priority keywords | Shows where the competitor already has relevance and discoverability |
| Metadata | Chinese name, subtitle, keyword themes and description | Reveals how the product is positioned for Chinese searches |
| Creative | Icon, screenshots, video, claims and visual sequence | Identifies the benefits and use cases competitors prioritise |
| Paid visibility | Paid Share of Voice, sponsored position and keyword coverage | Shows where competitors are investing and where auctions may be less crowded |
| Product model | Free tier, trial, subscription, in-app purchases and price | Helps compare traffic value and likely bid tolerance |
| Reputation | Ratings, reviews and repeated complaints | Reveals conversion barriers and unmet customer needs |
| Localisation | Chinese features, login, payments and customer support | Distinguishes genuinely localised competitors from translated listings |
Organic and paid visibility should be analysed together. A high organic position with no paid Share of Voice can indicate an opportunity to reinforce a keyword on which the app is already relevant. High paid visibility but weak organic ranking may suggest that a competitor is using advertising to compensate for limited organic reach.
The absence of detected competitor spending does not automatically mean a keyword is inexpensive or commercially useful. It may represent an opportunity, but it could also indicate low demand, restricted eligibility or poor conversion. Search popularity, estimated cost, relevance and downstream user value must be assessed together.
The keyword list should be created from Chinese user intent rather than translated from the global campaign. A literal translation may be linguistically correct but have little search demand, while users may prefer a shorter colloquial term, an English brand name, an abbreviation or a phrase based on the intended audience.
Useful keyword sources include:
A balanced keyword set normally contains several types of intent:
| Keyword Group | Examples of Intent | Typical Purpose |
| Brand | App name, company name, Chinese brand name and common variations | Protect and measure existing demand |
| Category | Broad terms describing the app category | Reach larger groups of relevant users |
| Feature | A specific tool, function or capability | Capture users with a clearer requirement |
| Problem or benefit | The result the user wants to achieve | Reach demand that is not expressed as a product category |
| Audience or use case | Age group, profession, situation or activity | Identify high-intent niches and long-tail demand |
| Competitor | Names of relevant alternative apps | Reach users already comparing solutions |
| Seasonal or event-based | Holidays, school periods, travel seasons or campaigns | Respond to temporary changes in demand |
High-volume category keywords can provide scale, but they are often broader and more competitive. Long-tail terms may have fewer impressions but stronger intent and better conversion. The most effective account normally combines both rather than treating search popularity as the only measure of quality.
Apple recommends separating search results campaigns into four themes: brand, category, competitor and discovery. This structure is particularly useful in China because it prevents established brand demand from hiding the performance of unfamiliar category terms.
| Campaign | Keyword Approach | Main Objective |
| Brand | Exact match for the app, company and brand variations | Capture and protect existing demand efficiently |
| Category | Exact match for relevant non-branded terms | Acquire users actively searching for the product category |
| Competitor | Exact match for selected competing apps | Test demand from users comparing alternatives |
| Discovery | Broad match and Search Match | Find relevant queries that are not yet in the controlled keyword list |
This follows Apple’s campaign-structure recommendations. Apple advises using broad-match versions of the established keywords in one discovery ad group and Search Match without manually added keywords in another.
Apple also recommends creating separate campaigns for each country. This means China should have dedicated campaigns and budgets rather than being grouped with other markets. A dedicated structure makes it easier to evaluate Chinese keywords, local creative, seasonality, user value and regulatory constraints without mixing them with global results.
Campaign names should also make the market, app, traffic type, match type and objective easy to identify. Consistent naming becomes increasingly important when several apps, custom product pages or automated bidding rules are managed in the same account.
Exact match provides the greatest control and should be used for proven brand, category and competitor terms. Apple may still match close variants, including reordered words, spelling variations and translations, but the pool is narrower than broad match.
Broad match can identify related searches and variations that the initial keyword research missed. Search Match goes further by using information from the App Store product page, similar apps and other signals to match the advertisement with relevant searches automatically.
Apple recommends considering both broad and exact match because they serve different purposes. Its guidance on keyword match types explains that broad match improves coverage and discovery, while exact match provides tighter control where user intent is clearer.
The discovery process should operate as a repeated cycle:
Negative keywords should also block searches for functions, audiences or use cases the app does not support. According to Apple’s negative-keyword guidance, this can improve efficiency by preventing ads from appearing for irrelevant searches or terms that repeatedly fail to convert.
A bid should be based on what a user is worth, not simply on the amount required to win more impressions. The account first needs a target acquisition cost or return on advertising spend derived from the product’s conversion and revenue model.
For a simplified install-based estimate:
Maximum CPT = Target CPA × Expected Tap-to-Install Conversion Rate
If the acceptable CPA is US$2 and half of the users who tap install the app, the corresponding maximum CPT would be approximately US$1. This is only a starting point. The calculation should eventually use the probability that an install becomes a paying or otherwise valuable user.
Budgets should be distributed according to intent and demonstrated performance. Brand terms commonly convert efficiently but may have limited scale. Category terms can provide growth, discovery campaigns require controlled experimentation, and competitor terms should earn additional budget through measurable results.
Impression Share and sponsored position can help identify constrained keywords. If a highly relevant term converts well but repeatedly receives limited impression share or appears in a lower sponsored position, a measured bid increase may reveal additional scale. The change should be tested on the individual keyword rather than applied across the entire account.
Conversely, a high bid should be reduced when a keyword consumes budget without producing valuable downstream actions. Automation can make these adjustments more quickly, but the rules still need to reflect the correct commercial objective.
Apple creates the default search results advertisement from the app’s product page. This means the App Store listing is part of the campaign, not a separate ASO project.
Custom product pages allow advertisers to create different combinations of screenshots, app previews, promotional text and deep links. These pages can then be assigned to ad variations so that the creative matches a particular keyword theme or audience. Apple permits up to 70 custom product pages and allows them to be managed without submitting a new app version, according to its ad-variation guidance.
A broad productivity app, for example, might use different pages for document scanning, electronic signatures and PDF editing. A game might separate creative by game mode, player type or visual theme. The message in the advertisement should continue naturally on the landing product page.
For China, the page should use Simplified Chinese and locally meaningful examples. Screenshots should accurately represent the Chinese app version, including available login, payment and content functions. Creative should also be reviewed for restricted claims and absolute terms such as “best” or “most effective,” which can cause approval or advertising-compliance problems.
Creative testing should assess more than visual preference. Different pages may attract users with different levels of intent and value. The best-performing page is therefore the one that contributes to the campaign objective, not necessarily the one with the highest tap-through rate.
Apple’s dashboard provides impressions, taps, tap-through rate, installs, conversion rate, average CPT, average CPA and spend at campaign, ad-group, advertisement and keyword level. Its reporting guidance also allows advertisers to analyse search terms, match sources, new downloads, redownloads and view-through activity.

These metrics explain the acquisition process, but they do not show whether the acquired users are commercially valuable. Apple Ads data should therefore be connected with App Store Connect, a mobile measurement partner or the company’s internal analytics and business-intelligence systems.
| Funnel Stage | Useful Metrics | Question Answered |
| Visibility | Impressions, Impression Share and sponsored position | Is the app entering and winning the relevant auctions? |
| Engagement | Taps and tap-through rate | Does the listing appear relevant and attractive? |
| Acquisition | Installs, conversion rate, CPT and CPA | Can the campaign acquire users efficiently? |
| Activation | Registration, onboarding completion or trial start | Are the users genuinely interested in the product? |
| Monetisation | Trial-to-paid conversion, purchases and revenue per install | Does the traffic generate revenue? |
| Retention | Active paid users, renewals, churn and cohort retention | Does the acquired value continue after installation? |
| Profitability | Lifetime value, payback period and ROAS | Should the campaign be scaled? |
Definitions must remain consistent. Revenue per install divides revenue by acquired installs, while ARPU normally uses active users as the denominator. Revenue per payer uses only paying users. Comparing these different metrics as though they were identical can produce incorrect conclusions about campaign quality.
No single tool replaces strategy, Chinese-language research and product knowledge. A practical technology stack can nevertheless reduce manual work and reveal opportunities that are difficult to see in the native interface.

Apple’s own tools provide campaign results, Search Popularity, Search Match, search-term reporting, recommendations, Impression Share insights and custom product pages. These should remain the source of truth for account delivery and spending.

SplitMetrics Acquire is designed specifically for Apple Ads management and optimisation. Its capabilities include:
These functions are particularly useful when a business manages many campaigns, apps or keywords and needs to connect market intelligence with account actions.

App Radar focuses more heavily on organic App Store optimisation. It can be used to research keywords, monitor rankings, compare competitor metadata and creative, manage localised listings and measure the effect of ASO changes. Combining organic and paid analysis helps identify keywords where advertising can reinforce existing relevance or compensate for weak visibility.

China-focused platforms such as Qimai Data (七麦数据) can supplement the research with Chinese App Store rankings, keyword information, category movement, metadata changes and competitor monitoring. Figures produced by third-party platforms are estimates and should be validated against Apple Ads and first-party performance data before budget decisions are made.
A mobile measurement partner or internal attribution system should connect each campaign and keyword with post-install events. The most useful setup combines Apple Ads delivery data with registration, trial, subscription, revenue, retention and refund information. Without this connection, optimisation can favour inexpensive installs rather than profitable users.
The most common strategic mistake is copying the global keyword list and creative into China with only a literal translation. This ignores local competitors, Chinese product terminology and different patterns of search demand.
Other recurring problems include combining brand, category and discovery traffic in the same campaign; allowing broad match to consume budget without reviewing search terms; failing to use negative keywords; bidding on large category terms without measuring downstream quality; and changing bids across the account instead of testing individual constrained keywords.
Campaigns also underperform when the Chinese App Store page is treated as a static compliance requirement. Unconvincing screenshots, unavailable features, weak ratings or a poorly localised value proposition can reduce conversion even when the keyword and bid are appropriate.
Finally, advertisers should not judge success using CPA alone. A low acquisition cost can hide weak activation, rapid subscription churn or low revenue. The objective is not to purchase the greatest number of installs; it is to acquire users whose value justifies the advertising cost.
As part of AppInChina’s Apple Search Ads services, we can help by:
Contact us to book a free consultation and discuss how to improve your Apple Search Ads performance in China.
