What We Did:

• We tested all ad networks which were not being tapped into by Grammarly’s own internal marketing team, thus focusing on incremental increases in reach.
• We tested a mix of contextual, in-app, mobile web and regular display ad units to drive low CPAs and identify sources of high upgrade rates of free users.
• Our design team created individual creatives to complement each ad network and the context in which the ads appeared, allowing us to split test hundreds of different ads to identify optimal performance.
• Setup tracking of user retention data on a placement level and used this deeper level information to predict upgrade rates, thus minimizing spend on non-performers.
• Increased average upgrade rates month over month through continuous optimisation of lower quality placements, creatives and ad networks.
• Utilised internal anti-fraud technology to prevent fraudulent in-app traffic being driven through three different ad networks.

Summary:

Despite Grammarly’s internal marketing team having successfully penetrated large ad networks, such as Facebook and Google, we were able to drive incremental volume through numerous smaller ad networks. We were able to cope with this more resource intensive strategy and as a result, we delivered high quality users at the same or lower CPAs than the larger ad networks.