The popularity of keywords has dropped. What is Apple trying to do?

Since mid-July, many CPs and ASO ers have reported that the popularity of many keywords in the Appstore has been declining, and there has been no sign of recovery so far. Since search popularity is the most critical factor in ASO, large-scale adjustments in popularity are bound to have an impact on all aspects of ASO, so everyone hopes to have an explanation for this phenomenon. We have conducted a detailed analysis of the relevant data collected by appbk, hoping to provide some data clues for everyone.

Let’s take a typical popular word “Taobao” as an example. The popularity trend chart in the past six months is as follows:

Figure 1: Search popularity trend of the keyword "Taobao" in the past six months

There are a large number of highly popular keywords, and the popularity trend chart is similar to that of "Taobao". In the more than half a year before July 2016, there was no drastic change in popularity. However, starting around July 16, it suddenly dropped from 9,700 to around 8,800 within two weeks. According to our previous search volume estimates, the search volume it represents has at least dropped by half.

Because there were more than 40,000 words with a popularity greater than 4605 in the AppStore before (according to experience, the average daily search volume corresponding to a popularity of 4605 is 1, and only words with a popularity greater than or equal to 4605 are valuable), and there are actually dozens of examples of feedback from everyone, so is the decline in search popularity just an isolated phenomenon?

To answer this question, we first counted the number of words with a popularity greater than 4605. From July 1 to August 1, 2016, the number of these words changed as shown below:

Figure 2 Changes in the number of keywords with a popularity greater than 4605 from July 1 to August 1, 2016

It can be seen that the number of words with a popularity greater than 4605 has rapidly dropped from about 40,000 to about 23,000, which indirectly shows that the decline in search word popularity is an overall trend. On July 1 and August 1, the number of keywords was counted by popularity range. The specific comparison is shown in the following table:

Table 1. Keyword quantity statistics by region on July 1 and August 1

From Table 1, we can see that the number of words with a popularity of 4605 did not change much in early August compared with a month ago, but the number of words between 4605 and 10,000 was almost halved, and only a very small number of words above 10,000 increased. Since the words above 4605 have less than half of the words in each popularity range, and popularity reflects the search volume, we get an interesting conclusion, that is, the average daily search volume of AppStore has dropped by half. If it is true that Apple cleaned up the fake search volume, then before. . . . . , the AppStore is searched by half people and half machines.

Furthermore, we selected all the keywords above 4605 on July 1st, and compared their popularity on August 1st one by one. We found that the popularity of about 96% of the words decreased, and only about 4% increased or remained the same. The following table shows the words with the most obvious drop in popularity:

Table 2 Keywords with the most obvious drop in popularity in the past month

Based on the above analysis, we can determine that the decline in the popularity of AppStore keywords is a common phenomenon rather than an isolated case. Then the second question arises, is this change in popularity a natural decline, which means that the number of user searches has indeed decreased? Or did Apple clear out some invalid search volume? This question can also be roughly guessed based on statistical data.

If it is a natural decline, then not only will the search volume decrease, but the corresponding App download volume will definitely decrease as well. Apple does not give the number of app downloads, but because the number of app reviews and downloads is in direct proportion, by counting the number of reviews per day in the past month, we can roughly get a picture of the overall download trend of the AppStore China region, as shown below:

Figure 3: Comment volume trend chart in China from July 1 to August 1, 2016, which roughly reflects the changes in download volume

It can be seen that in the past month, the overall number of reviews on the AppStore has not decreased significantly. Its changing trend is very consistent with the "summer effect". In early July, as students went on vacation, the overall magnitude increased significantly, and then by the end of July, it began to gradually decline.

Furthermore, we compared the data from some regions in the United States and did not find any obvious downward trend in popularity in China.

Therefore, the conclusion is that the large-scale adjustment of the search term popularity on the AppStore is likely to have cleared some invalid machine search volumes from the popularity calculation. However, we also found that the popularity of words such as " One Yuan Treasure Hunt " and "Golden Flower" briefly dropped on July 16, and then quickly rebounded to over 10,500. Could it be the legendary "The higher the virtue, the higher the evil"?

Figure 4 “One Yuan Treasure Hunt” popularity trend chart

For the ASO industry, the reduction in effective keywords does not seem to be a good thing, but with more realistic search popularity, we can better develop ASO plans and evaluate the effectiveness of ASO. By the same token, real search popularity is also crucial to search advertising plans. Therefore, this adjustment in popularity may also be Apple’s preparation for search advertising.