New App Store: What Are the Changes from Search Results to Conversion Rate?
How to Optimize App Metadata Mostly to Gain More Downloads?
App Store optimization consists of a series of techniques and methods to help developers to promote apps and acquire downloads. Among all the elements, app title, text description and keywords are the three main ones. App optimization is very important as more than 65% iOS apps are found by users’ searching on App Store. Therefore, if developers do not optimize apps on store, they may lose a precious opportunity to promote their apps for users.
In some certain situations, developers may select and combine words together to form new keywords. Because the keywords are limited by App Store, which cannot satisfy our usual demands, so we need to combine some words to show more keywords within the limited characters in order to display our apps from more different aspects and add the probability of impressions.
It looks like easy to combine words. However, in many occasions, it may be related to some basic knowledge of search engine. Therefore, we will firstly introduce some basic knowledge of word segmentation and searching.
The modern searching engine is designed based on the basis unit of “word”. And “word” is the smallest meaning language component. Word segmentation means segment a long word and then to combine the segmented parts together to form a new one. It can greatly help developers save characters and select more suitable keywords.
Here is an example of word segmentation. For example, there is a news app named “A news” and other named “B news”, and the two words may be segmented as “A | news” and “B | news”.
Basis of Search: Word of Bag
The word of bag is a simple hypothesis of search engine. It does not consider the order of “word” and regard texts as the “combination” of a group of words. That is to say, like the two phrases “A news” and “B news” are totally the same for search engine. This hypothesis is also the basis why we combine words.
Due to that the “word of bag” model is a “combination” model; we may ignore the repeated elements. Therefore, the words “news” of the above “bag” can be regarded as one single word. In other words, for search engine, “A news” plus “B news” equals “AB news” totally.
Then, we also can use the above model to spaniefly describe the searching process of users. When users have typed some words into the search box, search engine will firstly segment the typed words and then orderly look for the segmented words from the bag of keywords. Just as the mentioned example, if a user has searched “A news”, the search engine will firstly segment the phrase into “A” and “news”, and then look for the keywords bag which consists of “A”, “B” and “news”.
As the search system searches on the basis of the basic unit of “word”, so it should segment words at first and then will encounter a combining process.
To combine words means the process of classifying. By the means of “online clustering”, it actually combines the words with same character strings and then cycles constantly until there are no repeated words. During the combining process, we have taken Apple’s official rules into consideration which say a word should not be too long; otherwise, it may be regarded as keywords stuffing. Therefore, we have limited the maximum number of combination as 5.
Actually the maximum combination of words has combined all the probable words. But there are still many developers who will ask whether the comma can be deleted. By this way, they will have more space to add more keywords that may value a lot.
However, according to the above “word bag” model, even though we delete all the commas, there are no changes that will happen to the word bag. So it has less influence on searching results. Also, we should attach attention to the following two points.
1. Limited by Apple’s rules. Before 2015, almost all the keywords had no commas. And Apple now have more strict checking standards, so the old model may be regarded as keywords stuffing.
2. Influenced by weight. According to the word bag model, when users search “A news”, if your keywords are “A news” or “AB news”, your app will be found successfully. Both of them are ok. However, the keyword “A news” is completely same as what users search, in such a situation, search engine will generally more or less weighs much.
To conclude, it’s better not to add too many combined keywords, and developers should control them within 5. For the apps that have less downloads, don’t attach too much attention to combine words because your apps have more space to rank higher in search results.