This is a guest post from Sophia Fen, Mobile Account Manager at 3Q Digital.
As a digital advertiser, you most likely spend the majority of your digital media dollars on Google and Facebook. It makes sense—eMarketer forecasts that together, the two tech giants will make up more than 46% of global digital ad spend this year.
However, once those platforms are set up and optimized, where do you go from there? What else should you test?
In the mobile app world, a next step and expansion opportunity is running on mobile ad networks and DSPs. These offer scale, additional inventory, and targeting capabilities to help take your campaigns to the next level.
What is a Mobile Ad Network?
A mobile ad network is a collection of multiple ad inventory sources (includes both in-app and mobile web inventory) that allows advertisers and agencies to reach broad audiences relatively easily through targeted buys.
What Is a DSP?
A DSP is a demand-side platform. This is where you can run real-time bidding (RTB) campaigns, accessing a large amount of inventory by plugging into multiple ad exchanges and ad networks.
What Are the Pros and Cons of Working with Mobile Ad Networks?
- Large amounts of inventory
- Wide variety of inventory types—inventory grows and changes daily
- Usually an untapped growth channel for advertisers
- Many ad networks will let you buy on a guaranteed cost per install (CPI) or cost per action (CPA)
- Most ad networks have sophisticated machine learning algorithms that learn your campaign and objectives over time. They also maintain databases with millions of unique device profiles.
- Often there is less transparency
- Traffic volume can be sporadic, and pacing has to be adjusted frequently
- There is a lot of fraud and distrust in the space. However, more advanced fraud protection tools and guaranteed pricing models are helping to mitigate that.
What Types of Ad Units Can You Run on Ad Networks?
Most ad networks run banners, interstitials, video, native, and offer walls. [Offer walls should be reserved for incentivized burst campaigns only, as the user is likely not highly qualified since they’re downloading the app to receive something in return (coins or points for a game, for example).]
How Is Targeting Done on Ad Networks?
Targeting capabilities vary between ad networks, but most have some type of algorithm that uses a variety of data points to determine valuable users and audiences. These data points can include types of apps installed on the user’s phone, how the user engages with their apps, demographic and behavioral information, and location data.
Once a campaign has a substantial amount of valuable converters, you can then generate lookalikes for targeting for scale. You can also target by category (e.g., serve ads for a financial advertiser in other finance apps) or whitelist based on inventory sources that have performed well for other advertisers in similar categories.
If you decide to test into the ad network space, make sure you’re asking the right questions to ensure you’re set up for success. The below checklist is a good starting point.
Checklist: Discuss/Ask about these items when evaluating an ad network:
- Main differentiators
- Pricing models and minimums: Do they offer a fixed CPI model? Do you need a certain amount of spend to run a campaign?
- Out clause terms: Usually these are 48 hours
- Targeting and re-engagement capabilities
- Operating system scale: Do they skew more heavily toward Android or iOS?
- Geo-targeting capabilities: Can they target down to the zip code level?
- Level of transparency: Will they share site/app level data and provide dashboard access?
- Customer Support: Will you have a dedicated account manager?
- Inventory overlap: Are they buying from other networks?
- Types of ad unit: Do they run video, interstitial, banner, native etc.?
- Type of inventory: Do they skew more toward in-app or mobile web?
- Fraud: What are they doing proactively to combat fraud? What’s being done if fraudulent activity is detected? If you’re interested in learning more about mobile ad fraud in general and best practices for preventing it, take a look at this useful target="_blank">blog post.