Marketing is generally the largest part of any app business’ budget - so making sure your marketing spend is used as efficiently as possible across channels is key. The way to improve that efficiency is to analyze the numbers. Getting the full picture of creative performance, from both the high- and granular-levels, is how you’re going to allocate your spend the most effectively and achieve those small gains that lead to an overall big boost to ROAS.
Niv Yemini, VP Innovation at ironSource Luna, has the tips to help you get to the right data across channels and improve accuracy so you can make data-driven decisions and spot opportunities for optimizing your creative strategy - and your overall marketing.
4 ways to gather creative insights
Before diving into how to apply the full data picture, let’s first take a look at the four common strategies app marketers use to gather creative data:
1. Track data with an Excel spreadsheet
Creating an Excel sheet for recording performance and running analyses is one approach to data tracking. It involves manually gathering data from each channel and plugging it into the sheet, then generating reports. This is one of the simplest and most affordable approaches - but it comes at a cost.
Using Excel for manual data entry takes much more time - you need to gather the data from each source and add it to the sheet, then constantly update. The lack of timeliness and exposure to human error can lead to inaccurate information, especially on the granular level where there are even more numbers to crunch.
2. Rely on an out-of-the-box aggregation solution
Buying an aggregation solution that you can just plug into and go is another approach. For example, this is one of the functions of an MMP. Using an out-of-the-box tool is an excellent option for high-level monitoring and extracting macro insights, but relying on this solution alone has restrictions on how deep into the data it can go, like offering visibility into the creative format.
3. Build a data insights tool internally
You can build a BI system or other insights tool in-house - doing so gives you greater customization opportunities and the ability to go more granular than an MMP. However, the time and resource costs of building a custom tool internally are often very high. Just a few of the considerations to keep in mind include:
- Development time
- Hiring staff to build and maintain the system
- Learning curve of training staff to use the tool
- Innovation costs of keeping up to date with regular channel changes/updates
4. Use a comprehensive marketing data platform
The fourth and final option is a comprehensive, external marketing data platform that gathers creative data from across UA channels and aggregates it into one place. Relying entirely on an external platform could come at the cost of the control you get with an internally-built tool.
However, if you use an external tool in addition to what you build internally, you can enjoy the best of both worlds. Supplementing your internal tool with an external platform lets you get the granularity and specificity that can help you get a full picture of creative performance.
Look at performance on both the macro and granular levels
You never want to make decisions based on missing data - which is why you need a comprehensive view across your entire marketing strategy. Each UA channel behaves differently, so it’s important that you’re looking at creative performance both on a high level across campaigns and channels and a granular level by taking a deeper dive into each, too.
To get high level insights, use a data platform that lets you monitor performance across channels. Often, this is where an MMP comes in handy - with a monitoring tool like this, you can track performance in a macro way that helps you spot trends easily.
To get granular insights - like exploring the performance of your video creatives vs. interactive end cards, for example - you need an insights tool that can go deeper than an MMP that can be used for building automations and making optimizations that improve ROAS.
How does it look when you combine them? You could look at MMP or BI data and see the performance on one channel and see that it’s driving high-quality users - the ARPU is high overall. But when you increase spend on this channel to try and get even more high-quality users, it doesn’t get the results you expected - despite a higher budget, there isn’t a big uptick in impressions, conversions, or revenue.
Now, when you’re able to break down channel performance by creatives on an external platform, you can see that you’re only running video ads. By running with just one ad format, you’re missing out on the scaling opportunities of additional placements, like static and interactive ads, that can help you get more out of your spend.
Getting the full picture of data lets you make more accurate decisions that can boost ROAS and scale.
By getting access to these additional creative insights, you could make the informed decision to introduce more ad formats to this channel, and that this would likely result in the scale and revenue increase you were after from the start. Getting the full picture of data lets you make more accurate decisions that can boost ROAS and scale.
Look at the data from across the funnel
Too often, we see app marketers don’t test their creatives - they simply push them live with low budgets and see which perform best based on a single KPI. Creative testing based on data is key to optimizing efficiently and reducing wasted ad spend. The key is designing multiple versions of your creative and let them run long enough to gather sufficient and accurate metrics to find what we call the hero creative. Here are a few top testing tips:
- When designing different versions of your creative, start with macro iterations (e.g. the concept) before diving into minor changes, like camera angle and length
- Use an automated creative tool to test more variations and increase testing speed
- Allocate more spend during testing so you can shorten the learning curve
- Run tests consistently with new creatives to ensure they each reach the right threshold of impressions and can gather enough data
Analyze the results from each test by looking at the full set of data. KPIs from lower in the funnel, like ARPU and retention rate, tell you about user quality. Metrics from higher in the funnel, like CTR and engagement rate, indicate how effective the creative itself is at driving scale. Comparing KPIs from across the funnel gives you a full view of performance and helps you identify top-performers - and helps you spot areas for optimization. If you’re only comparing creatives based on CTR, for example, you could end up killing a version that’s actually earning a high ARPU because you didn’t look at metrics beyond CTR.
Comparing KPIs from across the funnel gives you a full view of performance and helps you identify top-performers - and helps you spot areas for optimization.
Here’s another scenario: Looking at your top-performing video creatives, you could see that ones featuring a bakery theme have a particularly high CVR and retention rate. This indicates the theme is resonating with high-quality users that are installing and continuing to use your app. With these insights, you can apply this theme to other creatives, like playables, and start iterating to see if it makes a positive impact on KPIs throughout the funnel.
Connecting post install data and marketing performance
For many mobile marketers, it’s challenging to draw correlations between post install data (like in-app purchases and DAU) and UA performance. But understanding the relationship is key, as marketing KPIs can inform in-app optimizations, and vice-versa - like testing the character from a top-performing playable as a bonus character in the app.
To draw correlations between post install data and UA performance, you need the full picture of creative data.
To draw these conclusions in an accurate and efficient way, you need the full picture of creative data. This not only helps improve UA performance, but can also boost your in-app metrics.