You just launched your new game or IP. You are already acquiring new players and considering push campaigns and in-app messages to interact with your players and keep them engaged in the game.
In this article, I’ll put together some thoughts on what you need to consider before establishing your CRM strategy: From choosing a tool, to defining how to tackle your player’s different lifecycles.
I will also show you some easy-to-start campaigns that can already drive decent impact on engagement, retention & conversion.
⚠️️⚠️️⚠️️Building in-house tools?⚠️️⚠️️⚠️️
Some studios consider building their own CRM capabilities in their backend. Why not? After all, your team is talented and could build this in a matter of months.
But the truth is, only the biggest brands can tackle this challenge, and it doesn’t mean they won’t end up going back and picking a third party tool at the end.
From experience, here are the main reasons why I consider it a mistake (in my humble opinion):
– Resources & Time consuming: It will be extremely hard to prioritize your development resources internally to build the tool. Usually, the tech teams are already packed and won’t be able to spare more than a few days to work on this. You can also consider hiring new engineers to focus on that but it will be costly as well.
– Maintenance & reliability cost: Once you build the tool, you’ll need to maintain it by updating the code at every OS update, as well as game updates. It is also extremely important that it is reliable (and monitored) so you’re sure you’re not missing on impact.
– Limited capabilities: The main factor why I believe it’s not worth the struggle. You will never get close to 50% of the capabilities available with a third party tool in terms of segmentation, triggers, tracking/reporting, etc..
Which tool to pick?
It will mainly depend on your budget, MAU volume and ambition with CRM efforts, but if you want to get serious about it, I’d recommend going for one of these
Braze, DeltaDNA, Leanplum, Clevertap, Mixpanel, Batch
Here are some of the features these tools will allow you to explore:
– In-app messages: I’ll talk about it later, but In-app message is the main channel you’ll be able to leverage for education & conversion campaigns. Extremely effective for any engagement goal as well (e.g.:feature adoption)
– Triggers: Being able to send a campaign based on what the user just did will be valuable when you’re trying to run action-sensitive campaigns (e.g.: purchase funnel winback, education campaigns, etc.) as well as time-sensitive campaigns (after X days)
– Advanced segmentation: Being able to segment users based on what levels they finished or how much currency they have left will allow you to tailor your messages based on each user.
– Variables: Another opportunity to tailor messages is to use data from the player (e.g.: username, friend’s name, offer name, etc.)
– Control Groups & Conversion events: Even if it’s feasible with an in-house tool, it’s way easier to set it up in a third party tool that will monitor the real impact on specific events (e.g.: purchase within 3 days after receiving the campaign).
– Retargeting: Sending one-off message will be helpful for a specific engagement goal, but once you want to create a “lifecycle” for each user, you’ll need to chain messages based on a user’s interaction with each campaign.
These are the top tier tools in my humble opinion. As features available differ between tools, you may find one to be the best fit (based on your tech stack & goals).
In the low tier, you can find tools like Firebase and OneSignal, which are still lacking greater capability in segmentation, campaign analytics and triggers.
If you’re on a budget, you can go for these. You will still get decent capabilities to get started and see what type of impact you can make on your core metrics — but don’t expect any magic.
Splitting the messages
Mobile games always use in-app messages, banners, newsfeed, highlights to educate the player about core mechanics and walk them through their progress.
I suggest that you keep everything related to progress/game loop (level unlocked, someone attacked you / etc..) deploying in-house. You’ll still be able to consider moving everything to the CRM tool later, especially if you want to experiment on copy or frequency of these messages.
You should handle all the monetization / engagement / education & marketing campaigns from an external tool that allows you to send campaigns without updating the app.
In-app as an obvious win
Referral in-app from RAID Shadow Legends
In-app is a great channel, the best for engagement-focused campaigns, you can use:
– Education: Show unused features to players or educate them on new ones.
– Surveys: get feedback from players based on new additions or a/b tests on the game loop.
– Rating: App rating is a strong tool to improve your ASO strategy (come read about our friends here)
– Winback: Users may lose their way through a specific action (e.g: onboarding funnel/purchase funnel), you can get them back on track with an in-app & adequate deep link.
– Engagement: Keeping the player hooked will involve getting them to interact with the game loop (E.g: spending coins they just collected or finish daily quests). You can use in-app for one-off engagement or build a habit around a specific “action”.
– Social features/account saved: Linking accounts add a social layer to your game, and give you a good opportunity to get players more engaged if they’re playing with friends.
– Events, marketing & offers: Live Ops & offers can be tweaked and experimented on. You can even build everything out of the box so you can switch your shop anytime without updating the game (leveraging game variables + deep links, it can also be on a one-to-one fashion if you link this to your DB or CRM).
– Pre-permission: Last but not least, you can actually build a full strategy around pre-permission for push notifications (required on iOS) that will allow you to improve the reach of your push campaigns to more players (more reach = more impact). You can find an example here from ONXMaps.
Everything I’m talking about can be easily applied to other apps as well as games. You can read more about it here from my former colleague Alex Gorius.
Step by step
Now that I’ve convinced you, I just want to talk a bit about the approach around identifying opportunities in your player’s lifecycle stages. Think of it as a game tutorial, you need to learn how to run before you can slay the dragon.
1)Audit your tech Stack
A clear understanding of your tech stack is the first step towards great CRM capabilities. Source
Don’t expect your marketing team to be able to significantly improve any metric if you can’t get accurate data and good capability from your tools. You should review every single tool and make sure you have good capability and accuracy in your tech stack. I talk about Tech stack here and mobile tracking here.
TL;DR: Make sure you can track & analyze user behavior (funnels, retention & revenue), you have good CRM capability (push & in-app messages), deeplinks (to levels, shop, daily quests & specific offers at least), and a good way to see your campaigns’ impact on several levels of engagement (and not only app open).
2) Define current state
What you want is to understand what the biggest user flows in your game are:
– Any progress wall?
– Any drop pattern? (e.g.: after specific level / time)
– What is your activation rate? Is it in line with industry benchmarks?
– What is your conversion rate? Is it in line with industry benchmarks?
To find benchmarks, you can leverage ASO / AppStore Analytics data, use Game Analytics (Automated benchmark available within the tool), or read different reports from GameAnalytics, Tenjin, DeltaDNA, Liftoff, and from other tools in the industry.
Next, you want to identify opportunities that could help you build new campaigns:
– What common paths are users taking?
– What features, apart from your game loop, are most popular (daily quests/shop/leaderboards)? What are the common usage patterns?
– What are now dormant or retained players doing after the onboarding?
– Are there emergent clusters in player behavior? (Looking into communities & social media can help.)
3) Basics: Activity Notifications
Activity Notifications examples from Candy Crush, Headball 2 & Coin Master
The first thing you want to do is to make sure your activity notifications are up and running (as discussed, when you start it will be probably on the back end side).
The key difference is that activity notifications depend on time-based reaction to something the player did or something other players did.
On Instagram, people liking a picture is something we’ll define as an opportunity for activity notifications.
On your side, you want to set up & test notifications for :
– Daily quest reset
– Chests / item unlocked
– Other player interactions (message / attack / etc..)
– New ranking
– Any time-based mechanics (building finished, end of season, etc..)
Also, by default, receiving an activity notification should be extremely relevant or psychologically rewarding. This will allow you to pump up the frequency and potential impact. For example, Clash Royale sends me a notification every time a chest is unlocked, which is fine as the reward is high.
4) Basics: Live Op
Live Ops Notifications examples from Headball 2 & MarioKart
Live Ops notifications will involve all the different events/seasons taking place in your game. These campaigns should be planned in advance, to make sure you’re taking full benefit of your messaging channels to improve the event’s engagement.
5) Lifecycle Marketing
Once these basics are set up, you want to consider Lifecycle marketing, which are campaigns that you’ll send to the player based on their actions. The trick is that lifecycle marketing should be adaptive and rarely repetitive (different from activity notification).
A) Funnel Winback
When considering CRM experiments, you first want to understand what are the main drop-offs in your funnels, looking into:
– Onboarding funnel: If players are stopping in the middle of the funnel, you can already send them a push notification (this is easier in android, as users register for push notification when they install the app, so you can send a message right after users left the app and didn’t complete the onboarding)
Again, it’s all about testing. The messaging should be aligned with the fact that the player didn’t experience the game yet, as opposed to a generic engagement message.
You may think the player didn’t have time and will come back later to finish the onboarding. But how would you know?
You should test different timing (30mins, 1h, 2h, 1d, 3d, etc..), set up a 50/50 control group and send this push, see the impact on early engagement metrics and retention and you’ll have your answer. This is an easy one and can really improve your activation rate.
– Purchase funnel: If a player expressed an intent to purchase a specific product, you can reach out to them with the exact same offer (or a discounted one). This could prove valuable as this user is already close to making a purchase (vs more generic conversion campaigns).
– Milestone funnel: Let’s say that players unlock features as they progress in the game. You want to make sure they’re aware of new features they may have unlocked. For example, if the PVP (player vs player) is now available because this player is lvl 30, you want to make sure they try it at least once (or that at least they know about it).
Example of Milestone campaign: We’re educating activated players to the Kitchen upgrade. Cooking day
B) Re-Engagement
– Time-based re-engagement: Let’s say a player starts playing for a couple of days, then doesn’t come back again. The closer to the latest session you try to re-engage the player, the easier it will be to get the player back in the game.
That’s why re-engagement campaigns are easy wins and can be optimized to reduce significant churn. Your messaging delay after the last activity complete, whether it be +1 day, +3, +7, or +14, will depend on the usage frequency for your game. It can be experienced on, but I’d suggest to tailor it for each user.
This can be achieved with calculated user attributes in your CRM tool that serve as a trigger (or segmentation if you want to simply split users on frequency group for analysis purpose).These attributes can be:
– App open frequency (since the first session).
– App open frequency in last X sessions, if you have a really visible cluster of engagement players go through (e.g.: unlocked tournaments so the frequency bumps up).
– Feature X frequency: If you have a specific game mechanics you want the player to do on a specific frequency, you can run re-engagement campaigns around it (e.g: daily quests, PVP, season pass challenges, etc..)
Again, this is more advanced. I’d suggest starting with generic +X days re-engagement campaign and iterate on messaging before moving forward.
Speaking of iterations, you should consider what the last player’s action was, or what’s their favorite activity. Aligning the re-engagement message with it will improve the conversion.
E.g.: If the player hasn’t experienced the PVP yet, you don’t want your re-engagement message to be about the latest tournament. Instead, if the player spends their time level-ing up their main character in the story, you want to talk about the next story chapter that lies ahead (or latest incomplete chapter).
Piffle use the last level as an incentive in their re-engagement campaign.
Another hypothesis is that the user got bored with the product and sending a user a re-engagement message which talks about a feature they never used will keep everything fresh for them. You just need to be mindful of the fact that the player never experienced this feature (Education copy).
C) Re-Activation
You want to consider reaching out to churned users that may come back in the game. The difference is in the targeting: For re-engagement users, we’re talking to early churned users that were active less than 14 days ago.
For re-activation, we’re talking to players that churned for 1–2 months or more. This definition may differ for other products (e.g: for subscription products, we are considering users who have stopped paying).
The messaging should be aligned accordingly: You need to “educate” these users again, or at least remind them of what the game/app is about.
D) Feature education
An easy one. By running a quick analysis on your main features, you may discover that daily quests have been accessed by only 20% of users after 5 sessions.
If you also know that doing daily quests correlates with high overall user retention, you want to use this information to create education campaigns around this feature.
You don’t want it to be too generic and sent to all users. Instead, you should consider when it is the “good time” to educate a specific user about this feature, and how to follow up if your first attempt on education didn’t work (e.g.: game education mechanics within the game may not be enough to get users to adopt a specific feature).
There are many other campaigns you’ll want to add here, these are just some examples. Some thoughts can be added around:
– Shop education campaigns: Explaining how each pack can benefit the player, explaining currencies (many people skip the tutorial).
– Progress friction: If the player is losing a lot of battles, you may want to offer them a specific power-up or equipment that will increase the percentage of wins.
– Flow campaigns: If the user is missing a specific currency, you may want to educate him on how to get more of it (or purchase it)
– Direct intents: Instead of firing different offers interstitials, try to understand what is the player interested in and display an in-app with one single offer aligned on it. (E.g. If the player looked at the description of a specific card more than once, you can offer a pack with this specific card instead of a generic one).
Again, this is really the “basics” on campaign ideation, but don’t forget to consider all the things that differentiate your game. That’s what you want to leverage to improve your messaging strategy.
P.S.: Please send us your game! We’d be happy to try it 🙂
Cheers.⚔️🧙♂️
Thanks to Andy Carvell, Hicham Massamba Wa & Natalie Jane Fitch!
