5 Little-known Data Measurement and Reporting Mistakes in App Store Optimization (ASO)
And the solutions to prevent them
Digital marketing is often synonymous to data-driven marketing. This applies to mobile apps as well. Measurement, analytics and reporting operations, therefore, are crucial in-app marketing.
However, while it’s already efficient to measure and report on advertising campaigns, doing so for ad-free, organic projects remains a challenge. The reason lies at its core: App store optimization (ASO) is the centre of organic app marketing or growth, and the strategies currently employed to track, measure and report on its performance are flawed in multiple ways.
The consequences for such flaws could be devastating. You could be measuring the wrong metrics and key performance indicators (KPIs) and get uninformed in your operations. You could be analyzing data from corrupted sources or processed in outdated systems that show you a false reality of progress. You could be making many other ASO reporting mistakes that may cause expensive negative effects on the overall marketing efforts. That’s why it’s important to identify and develop solutions to prevent them.
In order to help, here’s a list of top five most simple yet significant data-related mistakes in ASO reporting — and the steps you can take to minimize the risk of making them.
1. Prioritizing on data with low relevance
ASO is meant for increasing app installs in general and organic traffic as well as installs in particular. What’s more, those installs are meant to expand the user base for the app, who are ultimately expected to become paying customers. Only then can the app generate revenue and profit to sustain and grow the business that it supports. This means the top-priority data to report on should be app installs, the revenue they generate, the traffic that leads to them, and the conversion rates (CVRs) from one to another.
Yet, ASO reports still very often focus on lower-value metrics and KPIs. For some, the only thing worth reporting on is keyword performance. However, even achieving high ranking positions for the best keywords isn’t enough to increase installs. For others, the entire purpose of ASO is to improve the average user ratings in the app store. Unfortunately, even having a five-star rating average could mean nothing if the app doesn’t generate more installs. Either way, the same kind of mistake is made: Data with lower relevance to the actual performance of ASO tends to have higher priority, and vice versa.
In order to avoid making the same mistake, make sure your reports rely on the priority levels of the factors representing how well an ASO strategy is doing. They consist of three levels, with a descending order of importance as follows:
- Business-related data: This is the top level of metrics to report on. They include the revenue, profit and other economic implications of your ASO strategy. In order to make this kind of report accurate, you need to calculate a reliable customer lifetime value (CLV or LTV). Once you know the value your business can expect to receive from acquiring new users (organically), you’ll gain insights into the core business value of ASO.
- Installs-related data: Installs are the direct outcome of ASO, therefore their data is the most obvious indicator of your strategy’s performance. Specifically, they include the sheer number of installs, the growth in installs over time, the rate of such growth, and the CVRs from traffic (especially from organic channels) into installs. Looking into them is the best way to learn how well ASO does its day-to-day job: Generate installs.
- Opportunity-related data: They include any and all factors indirectly affecting ASO that appear in the form of opportunities for your app to convert into installs. The most important type of opportunity is the visibility of your app inside the app store, either in keyword/search results, app store featuring or top charts positions. Next to it, there are also user ratings and reviews that serve as the social proof of your app’s quality and reliability — a download motivator — among others.
2. Using a low volume of data
Optimization attempts in ASO often aren’t best-grounded because they’re based on reports with insufficient data. A lower volume of data means a lower likelihood of accurately identifying patterns in ASO KPIs. For example, you can’t find patterns in a keyword report without historical data across months, if not years, or in a competitive report without a vast amount of data about multiple brands to compare and contrast. In shorts, ASO reports are crippled if the volume of their data is low.
More importantly, the problem of data insufficiency in ASO reporting isn’t caused by technical difficulties. Most necessary data points are rather easy to locate, compile and store. However, the process largely depends on inefficient and frustrating manual work. What’s more, the data sources are also inconsistent with each other in terms of KPI definition and data structures.
Imagine manually retrieving data related to core growth KPIs from a mobile measurement partner (MMP) like Adjust, then keyword data from a tool like AppTweak, market insights from App Annie, and A/B test results from Google Play Console. While some of them allow data file downloads in bulk, others only allow individual file downloads, and some even disable the download option. They could turn ASO reporting into an unending torment. As we most likely want to avoid such torment, we’d tend to neglect it as well. In the end, it’s safe to say this is a mistake many are willing to make.
In order to save yourself the pain while also ensuring enough volume of data in ASO reports, you’ll need the following:
- A data warehouse: Building a data warehouse is not only useful in ASO. It could help in data automation across multiple departments in and outside of marketing. It aggregates data from several sources, on a transactional level, in large amounts, which is stored in the cloud and usually operated programmatically. This is how it enables quick data gathering at scale — but it requires advanced expertise and programming capabilities.
- A data team: Ideally, your data should be managed by a team of data scientists, engineers, analysts and architects. At a minimum, you should have at least one of them. They can not only construct data automation solutions that efficiently manage large volumes of data, such as data lakes or data warehouses, they can also contribute in overall data-related expertise to make sure your ASO is well-informed.
- API access: Most raw data sources come with API provisions. With the right kind of knowledge and technologies, you can make API calls where data is automatically siphoned and brought into central storage. The biggest advantage of this is you can focus entirely on the analysis and leave the data gathering to computers. Its downside, however, is the cost — APIs are usually costly and not all teams can afford them. This means they come in most handy when your ASO projects and teams are ready to scale.
3. Counting on data that isn’t reliable
One of the most dreaded reporting-related phenomena not only in ASO but also in mobile marketing is false data. They could be caused by many factors, from fraud to corrupted data sources and errors in data cleansing. Either by accident or on purpose, such flaws make your raw data unreliable. They will eventually misinform your decisions in ASO.
Unlike the problem with low data volume, the mistake related to data reliability is one many aren’t aware they’re making. When you pay a vendor for the data they give you, and it isn’t up-to-date, you couldn’t always see it coming. Similarly, when the data they give you comes in a disorganized and inconsistent structure, it tends to happen under your radar. Still, it’s your responsibility to make sure you have reliable data to report correctly. Here’s what you can do:
- Use a trusted attribution solution: Well-known attribution providers like Adjust and AppsFlyer are the forerunners in anti-frauds. On top of the ability to attribute your growth KPIs to the correct sources, they can help you reduce the impact of fraudulent activities surrounding your app’s ASO strategies. While it’s difficult to guarantee fraud-free data, they do prevent the majority of suspicious entries from entering your reports.
- Perform regular data audits: Data processes are becoming extensively automated. There are fast and dependable tools, technologies, systems and methods to pull and store data in bulk without manual labour. Yet, sometimes certain systematic errors such as misattribution may render the collected data unusable. Auditing the backends of such data could help in early detecting said errors and reducing their damage.
- Combat data discrepancies: One of the most common and also difficult to avoid issues in ASO data is the discrepancies that could occur to any type of metrics and data sources. As a rule of thumb, all numbers in mobile marketing are only final when the attribution process is complete, which could take days or weeks depending on the selected attribution window. The best way to avoid this is to look further into the past with historical data — or trends.
4. Forgetting to update the data frequently
Like many other marketplaces, the app store is very dynamic. Different users respond to apps in different ways, for different reasons. What’s more, they change all the time. This means the data you have about them may quickly lose its relevance. Your ASO reports, hence, will become outdated — and they will lead you to make untimely decisions.
For example, keyword optimization (KWO) is ideally done based on a keyword report so you know what keyword has high volume, among other attributes. The thing is, KWO happens once a month most of the time, based on data from the last few weeks, so when changes in the app store, such as seasonality, occur and alter keyword search volume, they wouldn’t be reflected on the reports. Eventually, your keyword strategy would follow a direction that’s no longer valid.
To make sure your reports are up-to-date, you need to do the same to the data. The best solution for this is real-time data processing or streaming, where updates are made as often as you want. However, this is often very costly and time-consuming, not to mention the high technical expertise that it requires to operate. This means it makes more sense when you’re ready to scale. Until then, it’d be best to consider more affordable and manageable alternatives:
- Pick the best time to update data/reports: Most of the time, you can only afford to update the reports once a month. To make the most of the situation, make sure the date and time you pull data in each month is the optimal one. For instance, due to discrepancies, Google Play Console usually shows relatively complete data for a month when the next month is already a week old. Therefore, it makes more sense to update your reports from there in the second week of each month.
- Anticipate seasonal changes: Seasonality tends to have a significant impact on multiple factors influencing ASO. Such influence will reflect on the data. To avoid getting surprised, you need to anticipate all upcoming seasonal changes with any implication for your app. Of course, you can’t precisely predict the future. What you can do, though, is look at trends in the past, such as what happens around summer to travel apps, during Black Friday to shopping apps, or in the winter break to education apps. If patterns in historical data are identified, chances are you’ll be able to anticipate their future behaviours.
- Review third-party reports/forecasts: Watchers in the mobile industry constantly keep an eye on recent events across multiple markets. What’s more, they often have contact with insiders who have access to rather exclusive information. If there are signs of potential changes in the future that may affect your app, they’re likely the first to notice. Related reports or forecasts may be published accordingly. Follow sources like Business of Apps, App Annie and Sensor Tower to stay informed.
5. Failing to combine diverse data types
This mistake goes back to having insufficient data. However, it doesn’t only mean low volume. It also means your reports contain data of too few types. This is a major problem because ASO involves too many different factors. Each of them has its own metrics, which come from several sources with inconsistent KPI definitions. Sometimes, lacking a certain type of data means it’s impossible to calculate the metrics that actually matter.
As an example, it isn’t possible to measure the number of organic installs of your iOS app based on data from App Store Analytics alone. Unlike Google Play Console, this platform doesn’t separate installs from organic traffic sources from paid. You’ll have to split them yourself with a calculation that contains installs from app store search and browse, and those from Apple Search Ads. This means access to reports on performance marketing or paid user acquisition campaigns is required on top of organic activities.
For another example, most attempts to measure the effectiveness of keyword optimization stop at keyword-level data. This includes keyword ranking positions, distribution, visibility score, and their movements over time. However, while keyword reports are helpful, they aren’t enough. They can’t show how much traffic the keywords have helped increase. They can’t explain how much more installs are generated because of them. And they definitely can’t estimate the resulting revenue or profit. A combination with growth KPIs reports is a must.
Much like seeking a second opinion in complex analyses, fixing the mistakes in data variety can be simple. All you need to do is:
- Clearly and accurately define KPIs: The best way to see what’s missing or how your reports lack diversity is to identify what’s already available. This is beyond naming the labels such as “App Units” or “First-time Installers”. You need to thoroughly understand the meanings behind them, what they consist of, and what roles they play in ASO. Metrics like CVR could easily become tremendously confusing without proper, concrete definitions. Are you tracking the CVR from Search Impressions to Installs, or from Store Listing Visitors to Installs? If you don’t know which one your reports are tracking, it’d be difficult to realize what they aren’t.
- Understand how each KPI is counted or calculated: A distinctive trait of ASO is that its relevant data sources are rarely ever consistent with each other. They are best known for counting and calculating the same metrics in different ways. To exemplify, while Google counts the number of times your app is viewed in the Play Store by a unique user (as “Store Listing Visitors”), Apple counts the same metric in the App Store either by unique device or with no filter at all (as “App Impressions”). If you’re aware of this difference and want to track your KPIs by a unique user, then you’ll know what’s missing from App Store data.
- Gain access to reports from all marketing channels: Data in ASO reports alone isn’t granular enough for in-depth analyses. To illustrate, reports on App Store Connect don’t differentiate organic numbers from paid UA. If Apple Search Ads (ASA) campaigns are active, all data it shows from the app store search will include numbers that can be attributed to both keyword optimization and advertising efforts. This means you’ll need to calculate the organic statistics yourself with additional data from ASA reports. In fact, as ASO involves multiple contradicting raw data sources, you’ll need to calculate many metrics and KPIs yourself. Make sure you have access to the right external reports to feed those calculations.
Final thoughts
ASO reporting often makes the difference between a failing and a working strategy. Whether it fails or works heavily depends on the insights of its reports, and the quality and value of their data. To determine such quality and value, ask yourself five questions about your data:
- Is it relevant?
- Does it have enough volume?
- Is it reliable?
- Is it updated frequently enough?
- Is it diverse enough to answer all important ASO questions?
Having clear and correct answers to these questions is the first step towards effective ASO reporting. If you have doubts, it’s time for digging into your spreadsheets and databases.
