Navigate the Guide
- The basics of Apple Search Ads keywords
- Discovery campaigns in Search Ads
- Search Ads keyword expansion: keywords pickling
- Search Ads keyword expansion: keywords analysis
- Keyword pickling and analysis: summary
Keyword expansion is a technique used to add new keywords to your current Apple Search Ads list of keywords. If done right, it helps drive traffic and find new users. Read on to learn the ins and outs of keyword expansion and how to perform it properly.
Look at the media buying funnel below. The top of it is fueled by ads impressions, while the revenue generated from those ads is at the bottom.
The goal of any business is to maximize its revenue. As far as the media buying funnel is concerned, marketers can influence 2 things:
1. Conversion rate (CRO). Here, optimization helps increase the number of users converted from impressions to taps and further down the funnel, so that they eventually bring you profit. The bigger the conversion rate, the less Advertising cost of sales (ACoS) you incur.
A/B testing is one of the methods to boost conversion. Such experiments help improve the attractiveness of your product page elements and by this means make users convert. Here you can learn how to run valid A/B tests on the App Store. CRO is a topic which is yet to be covered in another article.
2. Impression volume. The bigger number of users view your ad, the more of them can potentially get to the bottom of the funnel (with a certain conversion rate and profitability maintained). This article is exactly about how to make your impression volume grow.
Let’s figure out how to maintain a balance between impression volume, relevance, popularity and competition in order to boost traffic but stay within your Target ROAS.
This article will cover:
- The basics of Apple Search Ads keywords;
- Discovery campaigns workflow and search terms report optimization;
- Approaches to searching new keywords and analyzing them.
Below is the list of things that also contribute to Search Ads account profitability, but are not covered in this article:
- Bids management;
- Negative keywords;
- Conversion rate optimization (CRO);
- Approaches to account structuring, that is different ways to break keywords into groups and manage them.
The basics of Apple Search Ads keywords
The characteristics of App Store search queries
Firstly, search queries on the App Store are less variable compared to Google Search queries. Besides, we don’t use the App Store to search, for example, the model of a product or the address of a restaurant, at least until Apple allows in-app indexing.
There are 2 key types of search queries:
- Utility-centric (branded): an app name or a developer.
Try to recall what your latest App Store search was. You probably entered the name of an app you had learned about from friends or the media.
Users often know how to solve their problems and look for specific apps. Thus, the App Store attracts high-intent users who are further down the funnel in the first place.That is why Search Ads boasts such a high average conversion rate — 50%. In comparison, Facebook ads have to kindle user interest at the very top of the funnel.
- Discovery-centric (generic): general-type queries.
They are used when users have no specific goals. Rather, they look for something interesting or inspiring, for example, “relaxing games” or “shopping apps”.
According to AppsFlyer’s initial findings, 66% of all searches include the name of an app or a competitor. Only 34% of the queries are of a general type (discovery-centric).
Here’re more findings on App Store search queries:
- The average length of App Store queries is 2.4 words;
- 64% of queries include only 1 or 2 words;
- 36% of queries are considered long-tail keywords (3+ words).
Therefore, a typical query on the App Store is a 2-word branded keyword.
Keyword popularity and long-tail keywords in Apple Search Ads
The Apple Search Ads creators have powered the platform with a Search Popularity indicator. Ranged from 5 to 100, it shows a share of users who have entered a search query in a particular storefront. The value isn’t absolute and depends on the population of a particular storefront: 50% in the US and the UK storefronts will differ in absolute terms.
The bigger search popularity of a keyword, the more search traffic it gets. Such keywords are limited in number, hence competition for them is very high.
Keywords with search popularity of 5 can be ignored, as they have little to contribute to the number of impressions. Obviously, competition for top 10% of keywords is very high. In order to manage this imbalance, Apple Search Ads has another important feature – technical relevance score (find more below).
Additionally, according to AppTweak, top 20 keywords in the US (Search Popularity > 80) are all brands, with the exception of “games”. This speaks to the fact that branded queries bring about the biggest increase in traffic.
Unfortunately, the App Store doesn’t provide any data on the keywords search volume, so it’s difficult (yet possible) to make assumptions about the search volume of long- and short-tail words:
- Based on the statistics above, long-tail keywords are relevant for Search Ads but with some limitations;
- On Google, long-tail keywords account for 70-80% of the search volume. On Search Ads, their traffic volume is far less than on Google, given the shift towards short-tail branded queries. Based on the data above, they make 30-40%.
Google search demand curve (Source: Ahrefs)
- According to our data, there were 5 billion search queries in the US storefront in 2017, or 400 million per month. 30% of them equal to 120 million impressions for long-tail keywords only in the US storefront.
- Usually, copying the structure and keywords from Google Ads doesn’t work due to the huge difference between the essence of Google and Search Ads queries.
Why use long-tail keywords:
- Traffic of a higher quality and bigger conversion rates. Long-tail keywords are more specific. Compare “games” vs. “education games for kids”. The second query is more relevant to the searcher, so it can attract traffic of a higher quality and has a greater conversion potential.
- Long-tail keywords are less competitive, and therefore cheaper. People often ignore such keywords, as they imply more work and advertising expertise.
- More sustainable results. If you target only popular keywords, the chances are that a big player will come and outbid you, which will cause your impressions and revenue immediately drop. With long-tail keywords, all traffic is distributed between multiple keywords, which makes the problem less pressing.
Keyword technical relevance in Search Ads
Apple Search Ads functions on an auction basis. For your bid to make it into the auction, your app should be considered “relevant” to the keyword that matches to a particular user query. This is called “technical keyword relevance”. The Apple’s algorithm determines this relevance based on your app category and metadata — title, subtitle and keywords.
So, if you get no impressions for a keyword regardless of a high bid, it may be that your app’s score is below the relevance threshold. In this situation, try to add the keyword to your metadata to improve relevance and be eligible for the auction.
Here’s how the second price auction works. App 1 and App 2 have higher CPT bids, but Apple Search Ads ignores them due to low technical relevance. App 4 wins the auction because of a higher bid:
To sum up, when choosing a keyword, consider the following criteria:
- Impressions volume, which can be assessed by search popularity;
- Competition, which can be assessed by the number of search results;
- Technical relevance, which is the app’s position in search results (preferably <100);
- Logical relevance: the bigger TTR and download rate for the app, the better;
- Default CTP Bid: the average bid for a category or a given app to reach target KPIs.
According to Neil Patel,
80% of SEOs and marketers do keyword research wrong. They plug in a main keyword to Google’s Keyword Planner, download the result, and then start sorting through them in a spreadsheet. But here’s the thing: literally thousands of other marketers have already searched the exact same keyword.
Hence, it’s vital to discover uncommon keywords which are still likely to bring value.
Now, let’s explore 2 approaches to finding relevant keywords.
Discovery campaigns in Search Ads
The key purpose of discovery campaigns is to find relevant terms, long-tail keywords, synonyms, misspellings, etc., using the Search Ads match types with the widest reach.
Broad match keywords
As you can see in the picture, if you create a group with several broad match keywords and set a small budget, you can generate partial words, long variants, synonyms or related search terms. Just what we need! And most importantly, these keywords will have a high relevance score.
After your campaign has run for a while, you can dive into the search terms report and start generating great keyword ideas.
Below you can see a search terms report for a broad match keyword “games”. On purpose, we’ve selected the search terms without the word “game”. In other words, these are related (new root) keywords.
Search Match automatically matches your ad to relevant search terms without you having to identify and actively bid on keywords. It relies on multiple resources, including your App Store page metadata, the information on similar apps in the same genre, and other search data.
The basic idea of Search Match is similar to that of the broad match technique, but it doesn’t require to specify the list of keywords. The method helps discover keywords that could be missed, for example, search queries for competitors’ apps or even localized keywords.
Search Match can be wildcard, as you never know what exactly it will come up with.
Thus, the most effective way to increase your traffic volume is to gradually add new relevant keywords to your ad account. Below you can find a 6-step keyword expansion algorithm.
Keyword mining workflow and search terms report optimization
In order to mine keywords, you have to create a separate discovery campaign for each storefront and set a comfortable budget to spend on searching new keywords.
Obviously, the bigger the budget, the faster you will find more high-performing keywords. On the other hand, there’s no point in cutting the budget severely, as you may miss the potential profit that the discovered keywords can bring. A good idea is to apply the 20/80 rule, i.e. allocate 20% of your budget for the discovery campaign.
Since the main purpose of a discovery campaign is to generate as many search terms as possible, there is no much point in breaking it into targeting groups. This will only produce low volume terms and make management more complicated.
Step 1. Сreate a proper account structure
- Create a separate Discovery campaign for each storefront and set a comfortable budget.
- Create a Search Match ad group without keywords and with Search Match enabled, and set your bid at the ad group level.
- Within the same campaign, create a separate Broad Match ad group with broad match keywords and with Search Match disabled.
- If you already have a list of exact match keywords, add them to the broad match group as broad match keywords. This will speed up the process.
- To avoid competition for impressions, all keywords from your broad match campaign should be added as negative exact match keywords to the discovery campaign. To avoid competition for impressions, all keywords in exact match groups should be added to the discovery campaign as negative exact match keywords.
Step 2. Analyze search terms performance
Give your campaign a couple of weeks to run. This will guarantee that you collect sufficient data to analyze new search terms that Apple Search Ads matches to your ads.
Go to the “SEARCH TERMS” tab and set a 14-day timeframe. The best way is to analyze search terms every week or fortnight.
Arrange the search terms (STs) into 3 groups — Good search terms, Bad search terms and Insufficient data — and work with them as shown in the chart below.
Step 3. Identify Good STs
Good STs are relevant high-performing search terms which make the traffic grow without hurting its quality. All the steps above lead to discovering these queries. Good STs should be moved to the exact match keywords group.
The best way to evaluate the quality of search terms is to look at their Cost per Acquisition (CPA). For example, if your target CPA is $5.0, then queries with Downloads > 0 and CPA < $5.0 can be considered winning. Below are the filters you can use to discover winning queries:
After you’ve found your winning query, take the following steps (see in the chart above):
1) Add the winning search term to the exact match group. Depending on your account structure, there can be several exact groups where you need to move this keyword.
2) Add the winning query as an exact negative keyword to your discovery campaign, just the way you did in Step 1. This keyword is included in the exact match group, so you have to make sure that Search Ads doesn’t match ads to this search term in your discovery campaign.
Negative exact match excludes your ad for searches of the exact keyword phrase, but it may still show for searches that include related and long/short keyword versions. That’s why it’s reasonable to add close variants (e.g. “game” and “games”) and misspellings to negatives as well.
Step 4. Identify Bad STs
Bad STs are irrelevant or non-performing search terms. They don’t bring any installs and eat up a sufficient amount of money, usually 3 x Target CPA. Here are the examples of filters that can help find such queries:
Once you’ve discovered such search terms, add them to your discovery campaign as negative exact keywords.
Step 5. Identify keywords with insufficient data
These are search terms with insufficient data to be moved to any of the categories. There are 2 cases:
- There are no conversions and your spend threshold hasn’t been reached yet. For example, the search terms have spent less than 3 x Target CPA and haven’t brought any downloads yet (e.g. Download = 0 and Spend < $15).
- Low volume term. The data you request may be below Apple’s privacy threshold. For example, search terms must reach at least 10 impressions, otherwise the “low volume” value appears in your search terms report. Age, location or gender reports require minimum 100 impressions before Search Ads can show any values.
If you report has returned the “low volume” value, give the data some time to accumulate or extend the date range up to 30/60/90 days to get more data.
Step 6. Keep mining new keywords
Take steps 2-5 every week or fortnight and use the data from the latest review up to the present moment. By following these recommendations step-by-step, you can gradually increase the number of relevant keywords and, as a result, boost your traffic volume.
Naturally, the traffic volume in discovery ad groups will start dropping in a while. It has to do with exact negative keywords which are regularly being added to discovery ad groups. To keep mining new relevant keywords, you can gradually increase the CPT bid in discovery ad groups in order to make it into the auction with more expensive keywords.
Search Ads Discovery campaign: summary
- Discovery campaigns are a relatively simple method of expanding the keywords list gradually;
- There’re hardly any preparatory activities or research needed to get started;
- Discovery campaigns allow to immediately find relevant keywords (technical relevance) and evaluate their performance (CPT, TTR, etc.), which is impossible using Keyword Planner;
- Discovery campaigns also help find negative keywords.
- The discovery process requires expertise and constant attention;
- It takes time to obtain good results;
- You have to allocate part of your budget for the discovery;
- Apple Search Ads has a limitation on the number of keywords in a group; this can be a problem for large accounts, but in practice everything usually works well.
Search Ads keyword expansion: keywords pickling
Let’s explore another way to quickly obtain new relevant keywords with no budget waste. The technique starts with keywords pickling, that is with drawing up the list of as many keywords as possible. The second step is keywords analysis, when you have to analyze all the keywords and select those with the best relevance/volume/competition balance.
1. Own Search Ads organic report
The simplest strategy to search relevant keywords is to use those which your app is already ranked for in a particular storefront.
In order to make the searching proсess easier, we’ve powered SearchAdsHQ with the Keyword Planner:
Start with selecting an app and setting a storefront for which you want to make a report. In the example above, it’s Angry Birds 2 and the US storefront. Alternatively, you can specify a keyword for which you want to get insights.
The Keyword Planner can provide relevant statistics for 500 keywords, which can be downloaded as an CSV report. You can also make a separate keywords list using the right panel and then add the keywords to your account or download them for further analysis.
Let’s explore the key report columns:
- Keyword: the keyword you are interested in.
- Search Ads Popularity: the popularity of a keyword in a given storefront. This indicator can help estimate the keyword traffic. The bigger popularity, the better. Words with popularity <= 5 can be simply ignored.
- Search Ranking: this is an organic search position of your app for a keyword. The higher the position, the more relevant the keyword and the more efficient it is. Also, the higher the position, the bigger the outflow of organic traffic due to so-called “keyword cannibalism”. In practice, this percentage can hardly be calculated, only at the level of trends.
- Top by Rank: top 5 apps ranked for a keyword in a given storefront.
2. Competitors research
Apple enables to match ads to competitors’ brand names and keywords, which is a great opportunity to get more impressions. Like in the previous method, using the Keyword Planner you can obtain a list of keywords which your competitors rank for.
For example, in case of Angry Birds 2, you can research keywords for Clash Royale:
The Keyword Planner shows top 5 apps in search results for each keyword. As shown in the screenshot above, you can get insights by navigating through the apps right in the Keyword Planner. As a rule, we recommend to research up to 10 main competitors. This is enough to get hundreds of keyword ideas.
3. Related keywords in Apple Search Ads
Suppose you already have an initial keywords list or a list of keywords that have performed well in your account. It’s logical to assume that keywords which are similar in meaning or more specific variants of them may perform the same way. Apple Search Ads has a built-in recommendation tool that helps search related keywords:
Source: Apple Search Ads + SAHQ Extension
4. App Store suggestions
Like in method 3, in order to obtain more specific variants of your top keywords, you can refer to App Store search hints. The new keywords are likely to have lower popularity, but competition for them will be lower as well.
Suppose you have been running a campaign for the Angry Birds 2 app for a while, and the keyword [fun games] has proved to be well-performing. To get recommendations, go to the needed storefront and start entering the keyword in the search box.
The App Store will start offering hints at the end of the phrase. These are the keyword variants that have already been searched. Use them to get additional traffic.
However, this is an old-school approach. Alternatively, you can use special tools, such as our Keyword Planner, to aggregate all possible hints for a given storefront automatically and quickly. AppTweak has built a great free solution for this:
Let’s look at the basic settings needed to work with Keyword ideas:
1. Targeting settings: you have to specify the location and language that you do research for. In the “Search networks” box, you can select from “Google” and “Google and search partners”. Taking into account the specific nature of the App Store, it’s better to be picky about keyword variants and choose “Google” as a search network.
2. Related keywords: a keyword or keywords which you generate ideas for.
3. Filters for keyword ideas: here you set a filter that enables to show only closely related keyword ideas, thus narrowing down variations and excluding irrelevant ideas:
4. Keyword (by relevance): the list of keyword ideas sorted by relevance. All keywords and data can be easily exported to a CSV report.
7. Keywords multiplicator and synonyms
Another way to expand the list of top keywords is by adding synonyms, and there are many online tools that can help you find them.
For example, Power Thesaurus generates hundreds of synonyms and allows to sort out frequently used ones. Suppose you’ve made a casual game, and the keyword “relaxing games” works well for you. By using the tool, you can find synonyms for “relaxing”:
As you can see above, we have found 454 words with a similar meaning. The list can be narrowed down by using filters, but some good options are already generated: “calming”, “easy” and so on.
You can go even further. Expand the keyword with “for girls”, “for boys”, “for adults”, etc., and then use any keyword multiplier tool to get long-tail phrases:
This approach is widely used in e-commerce to get all possible product descriptions that people may search on Google. To obtain such combinations, Google Ads Keyword Planner has a built-in feature.
Naturally, this approach is not perfect for Apple Search Ads, as searching on the App Store has a specific nature. But in some situations, it can help get a dozen new keywords.
Search Ads keyword expansion: keywords analysis
Now that we’ve mined multiple keywords, let’s exсlude non-performing and irrelevant ones and divide the remainder into the groups — top, middle and low performers.
1. Get keywords performance metrics
To begin with, collect the minimum data for each keyword:
- Search Ads popularity: a popularity indicator from 5 to 100. The higher the indicator, the more search traffic can derive from such a keyword.
- Search ranking: the app’s ranking for a keyword in a particular storefront. The higher the ranking, the more relevant the keyword is technically and logically.
- Total search results: the number of search results for a keyword. It can be assumed that the more search results, the more the keyword is popular with publishers, and as a result, more publishers will compete for this keyword.
Almost any Search Ads research tool can help collect these data.
2. Remove low volume terms and irrelevant keywords
Regardless of the method you use to make the list of keywords, it’s necessary to make sure that users really search these keywords on the App Store. Otherwise, you’ll see zero impressions for such keywords in the report. To do this, remove all the keywords with popularity less than or equal to 5.
You can get rid of the keywords which your app doesn’t rank for. However, keep in mind that Search Ads can match your app to the search terms which you don’t rank for. We’ve taken this into account when calculating the Keyword Efficiency Index (KEI) so as not to lose high-quality impressions. KEI was calculated in such a way that keywords which we don’t rank for have a lower index.
3. The Keyword Efficiency Index (KEI)
The Keyword Efficiency Index (KEI) is an important indicator originally used in SEO to identify efficient keywords. This is a mathematical relation between keyword popularity (traffic volume) and competition for the traffic.
The idea behind the index is to find a balance between traffic volume and competition, that is identify keywords with sufficient volume but low competition to avoid highly competitive queries.
Here’s the formula to calculate KEI:
Let’s consider the parameters. We’ve already covered Popularity and Competitors. Relevance is a range from 1 to 100, and the bigger the number of a search position, the less relevant the keyword (inverse value).
Here are some things that matter for KEI:
- If your app doesn’t rank for a keyword, then the keyword relevance is low and doesn’t influence KEI. Such a keyword has relevance equal to 1 and is a degenerate multiplier that doesn’t affect the formula above.
- The keyword relevance equals to 5 if the app’s search position is 200+; if its position is 100-200, then relevance is 10.
- If the app ranks among the first 100 search results for the keyword, then relevance is calculated as the inverse value of the app’s position, i.e. Relevance = 100 – Ranking. Thus, the influence on KEI is bigger when the app ranks in top 10 than in top 50.
The bigger the relevance and popularity of a keyword, the more important it is. The influence grows exponentially, while the competition increases linearly.
Now let’s calculate KEI as shown in the screenshot. The bigger KEI, the more efficient the keyword.
Keywords with KEI < 1 can be ignored. For an app that doesn’t rank for a keyword, keyword relevance equals to 1. If popularity = 10 and competition is 100+, then KEI = 1*1 * 10*10 / 100 = 1. Such keywords are inefficient, so you can remove them and concentrate on what will definitely work.
This step guarantees that we will have no low volume or irrelevant keywords.
4. Keywords grouping
Now let’s divide keywords into groups – Brand, Generic, Competitors. The idea is to group keywords with similar metrics (conversion rate, cost), which will simplify bids management and account structure creation.
There are 3 main groups:
- Brand: keywords that include the app and company name enable you to capture high- intent users.
- Generic/Non-brand: keywords based on themes and descriptive terms for your app also help reach new users and raise your brand awareness.
- Competitors: using keywords relevant to major competitors is a great way to get new users.
5. Bids tailing
Brand keywords have the highest conversion rate and bring more engaged users, as they use brand keywords to search your application. It is very important to protect yourself from competitors who may bid on your brand keywords. Therefore, regardless of KEI, bid high on such keywords. For example, in our case it could be $10.
Let’s say our base bid is $4. Now we divide the remaining keywords into top, middle and low performers and set bid adjustments:
At this stage, we’ve done the keywords research. Now we just have to enter keywords and bids into our Apple Search Ads account.
Keyword pickling and analysis: summary
- This is a relatively quick and cheap way to obtain new keywords;
- The method helps find a bigger number of relevant keywords in comparison to discovery campaigns.
- KEI doesn’t help assess the potential profitability of a keyword or price per click, therefore bids will not always be set in the best possible way. However, this method is better than playing a guessing game.
The process of Search Ads keyword expansion can be performed efficiently if backed by SearchAdsHQ. The major purpose of the tool is to provide transparency of the whole optimization process so that publishers make data-backed decisions.
We have successfully applied this simple keyword expansion method to our customers. As a result, the traffic volume has increased by 300% with no harm to its quality.