Traffic Cannibalization in ASO and Apple Search Ads

Causes of Traffic Cannibalization: An exploration of the factors that can lead to keyword cannibalization.

There are several reasons to pursue traffic cannibalization:

Protecting your own brand

Application owners employ this method to maximize the number of views and downloads for their branded term while also preventing competitors from getting downloads through the same term. However, the question remains whether this is feasible, given that users who visit the store are likely to know that they will download your application for a specific branded term well ahead of time. Therefore, many different factors need to be taken into account, which we will delve into below.

Protecting the general search term

For example, let’s suppose you occupy the first place for the term "yoga." Since this is a quite popular search term, getting as much traffic as possible is advisable. With this in mind, you deliberately launch ASA for this term to get the maximum number of views and downloads, thereby depriving competitors of the opportunity to do so.

It's worth noting that this is only possible if you have a good conversion rate from impressions to installs for this search term (at least 15%). If the conversion rate is less than 15%, it is most likely not a very important search term in the context of cannibalization. Therefore, you can run ads, and be in the first organic place, but not have any downloads.

Pros and cons of traffic cannibalization: Examining the potential outcomes of keyword cannibalization

• A positive outcome of traffic cannibalization for a particular application is the "monopolization" of the top search results for a specific term. There is a possibility that the user will not see any competitors on the page until they scroll down, as the entire initial screen is occupied by a single app. And considering that the first three positions in search are the most conversion-friendly (especially when it comes to applications, not games), your application has a chance to garner the highest number of views and downloads..
• A negative outcome of traffic cannibalization is the unnecessary expenditure on advertising, as users end up downloading the application through ads instead of picking it up from organic search results. In other words, roughly speaking, you end up paying money for what you could have gotten for free.

In the graph below, we can see what happens to downloads as soon as a brand launches ASA for a term where the app is already at the top:

We clearly see a significant decrease in organic downloads (Search) due to the appearance of paid downloads (Search Ads). In essence, what we used to get for free, now incurs costs after launching ASA.

Here are several examples of what happens to downloads as soon as a brand decides to no longer pursue cannibalization and disable ASA for the targeted term:

As we can see, during the cannibalization period, paid traffic increases, and organic traffic, on the contrary, decreases. However, after disabling ASA, organic downloads start to increase once again.

How to determine whether mobile applications would benefit from traffic cannibalization?

Before starting the process of traffic cannibalization, you should divide search terms into brand and non-brand search terms.

1. Brand search terms.
   These are terms that contain the name of the application. When users enter such terms in the AppStore search, they most likely already know that they will download your application well in advance. However, you should note that not all applications have an equally well-established brand, leading users to proactively seek them out in the app store. You should also keep in mind that more popular competitors can rank for brand terms, or in some cases, the brand name itself becomes generic and is thus used to search for similar apps.
2. General search terms.
   These are terms that require more protection, as the potential user's choice is not focused on a specific brand.

Therefore, before applying traffic cannibalization, brand terms should be separated from general ones and each of them should be evaluated separately, thus analyzing the search results and competitors in the process and assessing how your application stands out among others in the search results for a specific term.

Go through a short checklist and answer a few questions.

1. Are there other applications running ads for your brand? This will answer the question of whether there is a need to protect the brand, which may not be the case if no competitors are targeting it as part of their advertising campaigns.
2. Do any competitors possess visual similarities to your application? In some niches, there are many visually similar applications. Considering that users mostly rely on visual cues, they may mistakenly download a competitor's branded app just because it is visually similar to yours. By monopolizing the search results, you reduce the likelihood of this happening.

3. Calculate the profitability and approximate the costs of cannibalization. It is difficult to ascertain with 100% confidence that traffic cannibalization will be profitable in each individual case. But by following the strategy, you can evaluate the results after a while.

To do this, we will use the following calculations:

‍Organic Search Units = First-time downloads - New Downloads

‍First-time downloads – the first downloads of the application by users that haven’t downloaded it before. We take the data from Apple Store Connect, selecting the required filter at App Store Search.

New Downloads – first-time downloads of an application through ASA. We take the data from the Apple Search Ads dashboard where downloads are categorized as first-time downloads. Although we are able to segregate them by App Store Search traffic, ad-driven downloads are still included. This is why we need to differentiate them using New Downloads as a separate metric.

Example 1

In 30 days when you were in the top spot of the search results, you had 3200 first-time downloads, of which 835 were paid (New Downloads).

Organic Search Units = 3 200 - 835 = 2 365

Now, after analyzing the previous 30 days when there was no advertising, we can conclude whether cannibalization was beneficial for us. For example, if you had 2,000 organic downloads (first-time downloads) in 30 days without advertising, and with advertising, there were 2,365 organic downloads, then we have a clear uplift, and we can safely conclude that, in this case, this method is justified.

Example 2

Another scenario could be where the total number of downloads remained unchanged after launching ASA, but the number of organic downloads decreased. This means that advertising led to the cannibalization of downloads. In other words, we did not get additional downloads, and, in addition, paid for a certain portion of downloads that we could have received for free without advertising.

In 30 days when you were in the top spot of the search results, you had 3,200 first-time downloads, of which 835 were paid (New Downloads).

Organic Search Units = 3200 - 835 = 2 365,

but during a similar period without advertising, the app amassed 3,100 organic downloads. Consequently, in this situation, you end up paying for a substantial number of downloads that could have been obtained at no cost. This raises doubts about the rationale behind such a decision.

Avoiding traffic cannibalization: Evaluating optimal approaches to avoid search term cannibalization

To avoid cannibalization (if this aligns with your goals), you need to manually track search results for the targeted search terms and make adjustments by either pausing or relaunching them in advertising. But considering that most applications have a large semantic core, this will take a lot of time.

For RadASO clients, this process is automated using our in-house tool RadASO Tech Boosted Solution. It can be programmed to automatically stop an ASA campaign based on specific criteria, such as if the mobile application has already occupied the desired position for it. This can be the first, third, or fourth position - we can stop the campaign while considering the unique nuances of the specific application or its promotional objectives.