π Growth Gems #70 - UA, Subscriptions and Creatives
Hi there,
This week Iβm sharing gems on creatives, moving to subscriptions and on-device campaigns. These insights come from Claire Rozain, Piyush Mishra, Elise Zareie, Robbie Kellman Baxter and more.
Enjoy!
π₯ TOP GEM OF THE WEEK
π’ PUBLIC SERVICE ANNOUNCEMENT
Allow me to re-introduce my gemsβ¦
Every day, I share on Twitter an insight like the one below. I started doing this on January 1 and (already? only?) 300 people follow @growthgems to get #1GemEveryDay.
Join them to discover or get reminded of growth insights!
Creatives: creative testing, measuring performance
Letβs start with creatives this time!
Adam Smart (Director of Product - Gaming at AppsFlyer) hosts panels with 6 UA experts to discuss Maximising Creative for UA.
Gems below!
π Audio is quite important on TikTok compared to other platforms. You can have more narrative, more of a build up, and play more with recent trends.
by Elise Zareie (Senior UA Manager at Miri Growth)
at 05:57
π For creative testing, Miro Growth tries to give the same spend to each creative: they donβt use the TikTok split test environment but instead use a rule so that each creative spends the same amount. They try to make sure they get 100 installs per ad to make a decision, and look at IPM vs. benchmarks from the category.
by Elise Zareie (Senior UA Manager at Miri Growth)
at 08:49
π Balance between new concepts and optimization is crucial. You can get lost in iterating on a winning concept and lose creative freshness. 2 new concepts and 3 iterations per week keeps a good balance.
by Elise Zareie (Senior UA Manager at Miri Growth)
at 10:32
π The main difference post-ATT is that gaming companies do not test the creatives on iOS but on Android, then move top-performing creatives to iOS.
by Svetlana Valgina (Gaming Partnership Manager at TikTok)
at 11:34
π On TikTok, the creative doesnβt really depend on the game type. Think outside of the box and be as creative as possible. Example: the βskip to the good partβ trend can be applied to most products.
by Svetlana Valgina (Gaming Partnership Manager at TikTok)
at 14:20
π The focus on UA vs. Retention when running campaigns depends on where you are in the journey of the app. From a creative strategy standpoint, try to speak to both acquisition and re-engagement with the same creative.
by Piyush Mishra (Head of Growth Marketing at Product Madness)
at 21:40
π Consistency in creatives is very important. You need to know your game category and what users like in your game to be able to advertise, but you also need to figure out who you are in the market: you have a short term performance goal but also a long term brand building goal. Example: Merge Mansion where the characters are the same and have come alive in each creatives.
by Claire Rozain (UA Team Lead at Rovio)
at 37:54
π There is a big shift: features, product and even creative evolution is now also based on how the community is reacting.
by Piyush Mishra (Head of Growth Marketing at Product Madness)
at 40:01
π You need to be best in class in creative production cost and understand how you can automate further. You need to understand your audience but also how to produce in a smart way.
by Claire Rozain (UA Team Lead at Rovio)
at 44:45
π Revenue is not a function of creative but a function of product (playables might be an exception). When it comes to banners/interstitials/videos, you donβt see an in-app purchase payment 7 days later that is attributed to the creative.
by Piyush Mishra (Head of Growth Marketing at Product Madness)
at 45:58
βοΈ Going Deeper: because of this, Piyush recommends to limit the duration of the window youβre looking at and not to look at D7/D30 ROAS on the basis of a creative. This is controversial though, and might depend on the type of games. For non-games, my take is that the creative can have a huge impact.
π Instead of running creative testing in your main geo, you can test in a cheaper geo. Run validation tests every quarter or half a year to verify the viability: 3 campaigns (one for each tier e.g. US, tier 1, tier 3) to see if you get the same results. If thatβs the case, then you can use tier 3 countries for creative testing.
by Miaomiao Li (Performance Marketing Specialist at Rovio)
at 51:26
π For a new channel/network, start with your creatives that have been performing the best historically on other platforms. Then evolve from there based on the data.
by Piyush Mishra (Head of Growth Marketing at Product Madness)
at 52:22
π The relationship you have with a new network/DSP is important. If you have an account manager, they can give you creative direction based on what theyβve seen perform in the past for the same kind of game. Some even produce creatives for you.
by Jan Sawicki-Hughes (Director of Growth at Tripledot Studios)
at 53:43
βοΈ Going Deeper: Elise later shared that if you have very good relationships with channels you can also request industry IPM benchmarks.
π Make sure you open a competitor tool like Mobile Action or Data.ai (App Annie) to have a look at your competitorsβ creatives by network. If creatives have a lot of reach and have been live for a lot of days, it means theyβre performing well.
by Claire Rozain (UA Team Lead at Rovio)
at 55:30
π Have a very close look at creative fatigue. Exploration is only about 5-10% of a big advertiserβs budget: youβre still buying the same sort of inventory for over 90% of the money youβre spending.
by Piyush Mishra (Head of Growth Marketing at Product Madness)
at 57:30
π Focusing on IPM to measure creative performance is the most reliable, because out of the early KPIs it is the one that is the most transferrable (vs. CPIs that can be influenced by CPMs/market condition).
by Elise Zareie (Senior UA Manager at Miri Growth)
at 58:08
Subscriptions: adding subscriptions
Shamanth Rao (CEO at RocketShip HQ) recently talked with Robbie Kellman Baxter (Strategy Consultant, Author of The Membership Economy and Advisor for Strava) about How to optimize products for forever: win with subscription apps in the Mobile User Acquisition Show.
A higher-level discussion with some good reminders, especially useful if youβre considering making a move to subscriptions.
π If youβre attracting people that have no intention of staying, that is a very uphill climb. Example: streaming service acquiring users by focusing on just 1 movie and seeing poor retention.
by Robbie Kellman Baxter (Strategy Consultant)
at 12:52
βοΈ Going Deeper: Robbie also shared that attracting the right people will also give you a cleaner read on whatβs working or not working.
Same thing if you focus a lot on getting free stuff or promotions to get users in. In his presentation βSubscription Apps Fundamentalsβ, Thomas Petit was sharing that you might even want to avoid focusing too much on the free trial in external communicationsπ
π Donβt worry about nuances like your price being $9.99 or $10β¦Start with:
Making sure you have $10 worth of value
Making sure you have enough value up front to get someone to sign up
Making the first seconds/minutes/days after they sign up an opportunity for them to recognize value from you and expose them to how they can get even more value
Keeping them engaged - if you have a 6 months plan make sure that you can offer value for 6 months!
by Robbie Kellman Baxter (Strategy Consultant)
at 15:18
π When first adding in a subscription model (e.g. as another product), the first question to ask is what this is doing for your larger ecosystem/business model. Define where the focus will be: awareness, protecting from competitors, for the heaviest users, etc.
by Robbie Kellman Baxter (Strategy Consultant)
at 18:29
βοΈ Going Deeper:
Another advice from Robbie is that when considering a move to subscription, you need to test your hypothesis in a very small market and look at the worst-case scenario. You can then adjust the offering, and test again.Β Example: you have several games selling for $60 each with in-game purchases. If youβre looking to move to a $100 subscription/year for a bundle, the worst-case scenario is if all users move to the subscription (even the ones that would buy 2 games a year).
π Only reasons to have a freemium model:
Trying to change a behavior on an ongoing basis: showing them that it will be a habit
Some kind of network effect: each new person using the product for free creates more value from premium users (e.g. LinkedIn)
A viral component: making free users a marketing channel to bring premium users
by Robbie Kellman Baxter (Strategy Consultant)
at 25:54
User Acquisition: preloads
Running out of UA channels to try?
Preloads/on-device campaigns are very particular, but if your app is a good fit (more on this below) they might be worth exploring.
Below are a couple of insights from a discussion between Esther Shatz (VP Product Marketing & Consultancy at StoreMaven) and Dor Isseroff (Senior Director of Sales and Partnerships at ironSource) on the Mobile Growth & Pancakes podcast episode Driving Incremental Growth Through Mobile On-Device Campaigns.
π When using on-device as an acquisition channel, expect a different performance for each country but also for each βchannelβ/carrier within a country. Users that buy via a carrier typically have a credit card, set up an account with recurring billing, etc. which means they tend to be more βhigh-endβ users than the ones who buy directly in a shop.Β Example: Vodafone, Win Tre and Samsung in Italy will bring different users even if they buy the same device. Or in India, only being able to buy a device online will narrow down the audience significantly.
by Dor Isseroff (Senior Director of Sales and Partnerships)
at 18:45
π For mobile-only banks, the choice of device model is particularly critical when running on-device campaigns. They get significantly better performance from users buying the latest models.
by Dor Isseroff (Senior Director of Sales and Partnerships)
at 21:10
π When first running on-device campaigns, youβll be looking at the same KPIs but with a longer window: D30/D60. Start with low daily caps, get users in and give it time.
by Dor Isseroff (Senior Director of Sales and Partnerships)
at 21:58
βοΈ Going Deeper: as you probably guessed, preload/on-device campaigns are not for every product. It most likely wonβt work for apps that are not mainstream. If your app feels like a good fit (Android is an important platform, mass appeal, other channels not scaling anymore, etc.), Iβd recommend getting on a call with ironSource and Digital Turbine to learn more (both have some pretty unique adjacent offerings as well).
You can also watch The strategies LinkedIn used to accelerate app growth by Martijn Lancee (Director of Business Development, Growth at LinkedIn).
And before I leave, a quote on pricing from a Product School session:
βNever ask your customers how much to chargeβ - Rahul Hampole (Product Leader at Plaid, previously at Yelp)
See you next time. Stay savvy!
βοΈ Sylvain