It's been a little over seven years since the mobile apps Big Bang, and in that time, they've become an essential part of most of our lives. However, a recent report by eMarketer has shown that although the number of apps and app users continues to increase, downloads and loyalty are becoming harder to achieve in a crowded marketplace. While the universe of apps is ever-expanding, our capacity to navigate that universe is limited, leading to a winner-takes-most scenario where a privileged few downloads consume more and more of our attention.
So, how do you make sure that it's your app that gets noticed, and more importantly, downloaded? At Google's Media Lab, our performance marketing team manages media campaigns to promote Google products. In doing so, they have developed best practices in using a range of platforms and established a test, learn and optimise philosophy that underpins all of their campaigns. Danish Bagadia shares the team's key learnings from promoting the Google App on AdMob.
The Google App is our search app for mobile devices, allowing phone and tablet users to access Google Search as conveniently as they do on the web. The core objective of our Google App campaign is to drive downloads and ultimately usage of the app.
Target & Test
When launching AdMob campaigns, we use run of network targeting and Conversion Optimiser to help us surface the best performing targeting options. After a few weeks of testing, we found that targeting people interested in technology was among the most successful methods of driving downloads.
In addition to finding the right people, we also want to show the right message to motivate someone to download the Google App. That's why we use both broad and specific messaging to layer our communications. For example, we highlight specific benefits of the Google App such as "find out the weather" or "what's the score" to help bring the app to life for potential users. Testing plays a big role here. We recently tested two different headlines in text ads and discovered that "Search with your voice" outperformed "Talk to Google" by 29% in Net Response Rate (NRR).
The ad formats you select should be driven by your marketing objectives - for an app campaign, this could be driving installs, engagement, or traffic to your website. Once you've defined your marketing objectives, selecting the features that will help you achieve your campaign goals becomes much simpler.
Through testing, three ad types emerged as delivering the most efficient results for the Google App campaign: text ads, Mobile App Install ads, and Video App Install ads. Text ads with tailored copy returned a cost per download that was 35% more cost efficient than other formats. The Mobile App Install format, which allows consumers to download the app straight from the ad, delivered a 35% lower cost per download and a higher conversion rate when compared to the Ads in Mobile App Install ads. Users responded well to video content in the Video App Install ads, leading to a high rate of downloads from this format. Testing various formats allowed us to identify what would truly deliver incremental value for our campaigns.
The length and duration of our tests depends on the impression threshold required to achieve statistical significance, which can vary based on the type of test. For the Google App campaign, it took about two weeks for us to see performance differences between ad formats. We continued to run all formats for another few weeks in order to establish significance. We recommend not limiting yourself to strict testing timelines - give yourself the flexibility to test as long as you need in order to gather learnings and establish confidence in the data.
Understand Market Variations
If you're running pan-European campaigns, understanding and responding to market differences can be challenging. In our experience, measuring delay to action (the amount of time between a user seeing an ad and downloading the app) by country adds context to our understanding of audience response rates, enabling us to optimise our campaigns according to regional dynamics. Measuring performance across mobile OS, demographics, browser, location and more allows us to examine additional audience nuances across markets. We then take what we've learned and feed it back into our campaign parameters.
In the Google App campaign, we used the insights we had about each country and applied it to our planning process. For example, we knew that smartphone penetration in Germany is lower than that of the UK. This came into play when we launched the campaign in Germany and found that conversion rates were lower than in the other markets. As a result, we realised that we were going to have to let the campaign run longer, and therefore budget appropriately, in order to get the same number of downloads as in the UK.
Without measurement, accurate optimisation would be impossible. The right measurement is delivered with the right goals and tracking in place. It's important to understand the impact of your investment and in mobile app promotion, it's crucial to have the right SDK implemented and integrated with the platforms that you use. Employing the tactics above have allowed us to achieve a cost-per-download that is over 50% lower than our target, despite increased spend.
Media Lab works closely with a digital agency to coordinate and manage campaigns. The relationship is a partnership, and we trust them to bring their expertise to bear in all aspects of planning and execution. If you're working with an agency, chances are they're already familiar with much of this methodology, so why not use this article as a basis for your campaign optimisation discussions?
Test, Learn & Optimise
As reports such as eMarketer's show, the crowded apps ecosystem is becoming more difficult to navigate each year, but having the right tools and approach can make achieving your goals easier. As we run future campaigns, we'll continue to test, learn and optimise our methods, and will be back to share more learnings from Media Lab soon!