No content

Facebook advertising best practices How to start, measure and scale Evgeny Marchenkov CEO @ AdSide Media [email protected] 

About AdSide Media We help developers to bring revenue through Facebook (create ROAS-positive campaigns) ■ Exclusively focused on Facebook and Instagram ■ Deep analysis of each client’s business ■ Benchmarking for in-house FB media-buying teams ■ Media buyers from product companies ■ Video and banner production included 

This presentation will… …get you ready for media-buying on Facebook, and teach you: ■ How to set up analytics ■ What metrics to analyze ■ How to apply metrics to the media-buying process ■ How to start buying data on Facebook 

How to set up analytics You need to integrate: ■ Facebook SDK ■ 3rd party tracker ■ AppsFlyer ■ Adjust ■ Kochava ■ Server analytics 

Where do people fail with analytics? ■ Sending incorrect & insufficient data ■ Sending the same data from FB SDK and a 3rd-party tracker ■ Setting up a different attribution window between FB & Tracker ■ Choosing different time zones between FB & tracker ■ Not verifying transactions properly ■ Failing to specify currencies Remember: There will be discrepancies, but no more than 10%! 

How to start testing… even in the dark Testing means making sure that Facebook is a profitable channel If you already have data: ■ Run FB campaigns and make sure that CAC < LTV If you’re just starting out: ■ Research the existing market ■ Example from the market: Average CPI for mobile MMO 3d shooter in the US is $2.0 ■ Client’s data: Deploy & Destroy by AppsMinistry/ LionsGate: US CPI $0.3, EUR CPI $0.05, WW CPI $0.01 (Watch the game’s FB Ad here) ■ Measure LTV from organic traffic ■ Start buying data on Facebook 

Product metrics? Not just LTV Measure these… ■ LTV ■ ARPU ■ ARPPU ■ Lifetime ■ LTV Curve …with cohort analysis 

Getting data buying right What to buy: ■ Installs ■ Paying users ■ Revenue from ad spend How to buy: ■ Targeting ■ Creatives ■ Optimization 

What to buy ■ Installs ■ CPI should be less than ARPU ■ Buying cheap installs means getting less revenue ■ Example: We used a killer FB Ad (watch it here) and got very cheap installs, but not much in return ■ Paying users ■ CAC should be less than ARPPU ■ Example: We used organic ARPPU, but found out that different cohorts of paying users brought different revenue ■ Revenue from ad spend ■ ROAS > 1 for the specific timeframe ■ Remember to add revenue generated by organic traffic from paid UA and to exclude the store cut (30% from revenue) 

The holy trinity of FB media-buying Targeting Creatives Optimization 

5 ways to target audiences ■ Set a wider target audience ■ Run additional tests for the best performing narrow audiences from the first test ■ Reuse audiences from other apps ■ Create Custom Audiences and LALs based on the significant events in your app (e.g. Tutorial completion, Purchase, Level X achievement etc.) ■ Target users who like your competitors 

4 ways to optimize media-buying ■ Test different FB optimization types - Installs, Events, Value ■ Start with small budgets ■ Duplicate adsets ■ Duplicate creatives within adsets Remember: Facebook SDK sends data about ALL users in the app 

3 ways to create media-buying assets ■ Video works best when ■ Aspect ratio 1:1 and 4:5 ■ Up to 15 sec ■ Attention is grabbed in the first 3 seconds ■ Use banners to test ideas, concepts ■ Create videos based on the best banners ■ Use Playable ads ■ The new way to engage users 

AdSide Media Showreel https://youtu.be/h35Vkdzs2wE 

Let’s wrap it up ■ Set up analytics ■ Measure key metrics – LTV, Lifetime, ARPU, ARPPU ■ Use cohort analysis ■ Start buying data ■ Look at ROAS and predict LTV ■ Set a wide target audience ■ Test different optimization types (purchases, installs, value) ■ Create videos 1:1, 4:5 up to 15 sec And you’re all set! 

Thank you! Evgeny Marchenkov CEO @ AdSide Media [email protected]