Have you ever wondered, how does in-app advertising work? What goes on behind the scenes to make mobile advertising happen today? Then this guide is for you.
Introducing the Digital Marketing Players
In the past, advertisers would communicate directly with app developers and publishers whenever they wanted to run an in-app ad campaign. Essentially, the advertiser would pay the publisher directly for access to their user base.
But, as the number of apps (and brands investing in digital advertising) rose, these one-on-one deals no longer made sense. Programmatic ad buying and real-time bidding have affected matters further, by introducing greater automation and efficiency to the mix.
So, who’s now involved in getting an ad to appear in an app?
On the advertiser side, a marketer will either have their own team members and/or a creative agency develop a mobile app advertisement and put together the creatives to use. From there, a demand-side platform (DSP) will aggregate creatives/campaigns from multiple advertisers, and then work to ensure they are offered up to app publishers.
On the app publisher side, developers and app owners will integrate with a source of advertiser demand either directly or with a supply-side platform (SSP) that aggregates publisher supply. Typically, this integration occurs through a software development kit (SDK), but sometimes it’s through an application programming interface (API).
An ad network is a bridge between the two sides. An ad network connects advertisers and their partners with publishers and their partners, ensuring that lots of brands can reach their target audience across lots of apps and that apps have lots of advertisers clamouring to reach their users.
Breaking Down the In-App Advertising Supply Chain
So how does an ad actually end up in front of someone? What happens to make sure the right mobile ads appear at the right time and for the right person? Let’s break it down:
First, marketers need to determine who they want to reach and what kind of messaging they want to bring to market. For example, an energy drink brand may decide it wants to reach male gamers between 18 and 25 in the U.S. with their campaign.
From there, marketers (and their partners) need to determine their goals for the advertising campaign, including how much they want to spend and how they will be determining success. Goals could range from brand awareness, i.e. reaching as many relevant people as possible looking at their ads, to driving a specific action like a mobile web landing page visit or an app download via a particular app store.
Concurrently, app developers and publishers need to determine how they will generate ad revenue and how ads fit into the user experience being offered. Consider this: while banner ads are fairly easy to unobtrusively include in an app, more immersive formats like video ads typically provide greater ad revenue. App publishers need to consider how ads affect app usage, while also maximizing earnings.
Now back to the advertisers for a moment. Once a campaign is finalized, it’s entered into an ad network’s campaign management system. From there, an ad ops team will review the campaign to ensure it meets the ad network’s quality standards. Creatives and campaigns that don’t adhere to compliance standards, publisher preferences or existing laws will be rejected. When a campaign is approved by ad ops, it’s placed in a queue on the ad server until the right time.
Now, when a user opens up an app, the ad network’s SDK (or API, in some instances) is triggered, to highlight the fact that a particular person is ready to have an ad served to them.
Based on the data available from the SDK, the ad network determines which potential ad would be most relevant and offer the greatest payout to the app publisher. When impressions and reach are the primary campaign goal, determining the right ad to serve is fairly straightforward. But, when a campaign tracks success using performance metrics like ad click-through rates, the ad network needs to determine which ad will likely be clicked on and engaged with by the end user, and then use that prediction to determine payout rates and ad revenue. Having historical data and advanced predictive analytics capabilities powered by artificial intelligence helps in this arena, as does machine learning algorithms.
Once an ad is selected (or served) it’s then (hopefully) rendered on the app and shown to the end user. The whole process, from start to finish, happens in as close to real time as possible – ideally, around 200 milliseconds. It’s important to note, however, that just because a mobile ad is a server doesn’t mean it’s rendered. In the time it takes for the SDK to present a potential ad opportunity, someone may have left the app environment or scrolled past the advertising opportunity window.
Finally, after all this, there’s the reporting and payout period. This occurs either during or after the campaign. Essentially, this is where advertisers and their partners determine how the campaign performed, and how much they owe based on performance. For instance, if the campaign’s goal was to ensure a certain amount of time spent engaging with an ad, then the data will determine how much an advertiser should spend with the ad network. Typically, advertisers will use third parties like AppsFlyer, Adjust, etc., in combination with the ad network’s own data to determine how the campaign went and how much is owed. This data can also be hugely influential in determining which ad network partners to use. But, certain platforms limit how much third-party verification is possible. With a Facebook ad, for example, advertisers are limited to just data from the publishing platform.
Other In-App Advertising Considerations
Getting in-app advertising right is about more than just understanding how the ecosystem works at a basic level, as there are many other factors at play in a mobile app advertising campaign.
For starters, it’s key to understand that mobile web advertising is very different than mobile in-app advertising. While many brands are often tempted to run the same campaigns for both audiences, such an approach is not primed for success. Even though both involve mobile devices, they are dramatically different. In particular, how ads are displayed and what data is available can vary significantly between a browser-based environment and a mobile app environment.
What ad formats will work best for the goal, and for the target app? Like the earlier example, sometimes lighter ad formats like banners, native ads or interstitial ads are fairly easy to integrate, and a lot of apps have the available inventory. These formats can be good for awareness, but not always for interactivity. In contrast, video ads and ads with interactive end cards are ideal for generating clicks, but not all apps incorporate these ad types. It’s important to weigh reach with cost and interactivity when deciding which ad formats to develop.
Also Read: Reinventing Display Advertising for 2019
What technology is used to serve and then render an ad on mobile devices? This is another key question to full answer, as it’s crucial to ensure ads are served and rendered in a timely and efficient fashion. After all, even the best ad will be rendered moot if it fails to appear. This is why the VAST video ad delivery standard is gaining traction over VPAID, as its precaching capabilities ensure that ads render in less than a tenth of a second and are not encumbered by loading problems.
And, another key question to answer is, how does the publisher determine which ads to serve? Typically, app publishers use more than one source of ads, and often leverage dozens of ad networks. Many apps now use waterfalls – in which different ad networks, exchanges, etc. are called in sequential, static order until all ad opportunities are fulfilled – although unified auctions are becoming more popular since they introduce more competition and reduce latency.
This article originally appeared on the InMobi blog.