Meaningful Engagement & Retention Panel

Current context

Murad (Full HP)

• Downloads and retention like never before (61% D1 retention when it was around 40% before)
• Monetization is skewed towards ad monetization side.
• Increased the number of events in the games, introduced a "property" for players with the goal to bring them back in the game after the crisis, more social features, etc.

Tatu (Rovio) - boost for Puzzle games with covid.

• Trying to enable users to spend as much time as possible in their app (e.g. with unlimited lives)

Steve (Product Madness)

• Going very well, seeing KPI uplifts across the board. Fortunate.

Ada (Clue)

• Initially thought they might get more users into the app but seeing similar numbers on engagement/retention: it seems that users that have stress-related concerns balance the ones that are too overwhelmed by health info.
• In the long run they will be able to better understand how the situation is impacting people because you can track stress, motivation, etc.

Keeping newly engaged users

Steve (Product Madness)

• Looking at the product management side. New users come from paid and organic.
• Product Madness has a long-standing push towards the new user experience, curating the different flows and outcomes. Doing some fine-tuning now: looking at improving loading times to make the experience better and how to give more value in (shorter) sessions.

Tatu (Rovio)

• [💎@13:17] You want to understand where the user journey starts: the creative ad that you're showing and to which audience. Because depending on the kind of audiences you're driving to your game you will get a totally different kind of engagement and retention numbers.
• Example: misleading ad

Murad (Full HP)

• Typical audience is 13-25. But now have seen 30+ yo new users. They need to find what would keep this new audience in the game.
• That new 30+ audience is accumulating a lot of items in the app so Full HP created a new type of daily reward system based on their engagement analytics.
• Plan to also run retargeting campaign.

What influences engagement

Ada (Clue)

• [💎@19:00] Understand why the user is there in order to influence engagement. You can have a poll asking users early-on what is their main focus. This informs the best time to engage with them or the additional information you want to expose them to.
• Examples of user focus for Clue: next period and symptoms, pregnancy related, birth control, sickness, mental health, stress levels, levels of energy, etc.
• Clue has a hub of content they can pull from to push it through the app (notifications, etc.)

Attracting the right audience

Murad (Full HP)

• Acquiring loyal users is the biggest challenge, and there are a lot of creatives that are misleading in the industry.
• [💎@22:15] For Full HP the best way to acquire loyal users is App Store Optimization: keywords, localizations, icons, screenshots, description, etc.. A/B testing and localization to different markets makes a huge difference in performance and following engagement.

Tatu (Rovio)

• Do a lot of A/B testing
• [💎@23:10] Look at performance of creatives on the creative level and on the placement level.
• [💎@25:05] Some ads that works in FB might not work on ad network. Understand the user experience of watching the ad. Is it skippable or are you forced to watch it until the end?

Steve (Product Madness)

• Not too involved on the creative side of it but have experimented localization of app store assets.
• Try to design things so players are happy and feel comfortable paying because this influences what your LTV is and therefore what your marketing budget is.
• [💎@27:30] It's important to have your product and marketing teams work together. A misleading ad might give you a ton of users but if users are not performing once in the app it is not that valuable.

Keeping users beyond the (personalized) onboarding

Ada (Clue)

• Making sure that their users understand how the app treats their data and privacy is very important for Clue. They have a series of blog posts explaining how the data is tracked and link to it from onboarding material, including an email solely dedicated to this.
• The CRM team doesn't know what is going on in terms of personal tracking at the user level but can look at the engagement with the app and how much you are tracking data. They encourage engagement based on that.
• Main metric in every experiment they are running is always the top level engagement and retention metric. Example: did they set a reminder after they've been explained how to do this?

Predicting engagement

Tatu (Rovio)

• [💎@32:25] Rovio is using bots and machine learning to try and predict churn. In casual games and puzzle games churn is mostly related to the level funnel so they use the data from bots to make better funnels and use the churn predictions to keep players engaged (e.g. sending them to another app)

Steve (Product Madness)

• [💎@33:30] Product Madness has models for churn prediction that might trigger: CRM offers, free virtual currency, deal for a purchase.
• Important to draw a line: they never change the odds and outcomes based on player profile or history (odds are always random).
• People want to see the value if they're invested their money, so it's fine for players to win some and save a lot of time in the game if they paid for something.

Murad (Full HP)

• [💎@36:10] Full HP divides its retention politic in 2 parts:
  Game designing elements: new contents, systems to progress through the game
  Which is feeding social media campaigns: target audience is Gen Z and they reach them through TikTok, FB, etc.

Metrics

Tatu (Rovio)

• A new metric they have started to look at is the cost per payer to understand what is the actual cost to acquire the right audiences.
• Also LTV
• Compare cost per payer vs. LTV

Steve (Product Madness)

• [💎@40:30] It's important to have a solid LTV model because if you can accurately predict D90 or even D365 it makes it much easier to be confident in your campaign budgets.
• [💎@40:50] Track LTV and other KPIs by campaign.
• [💎@41:15] Technical KPIs are often overlooked. Example: framerates, loading times, size of device, ANR on Google Play.
• [💎@41:30] Track users' sentiment via NPS or volume of customer support tickets to understand if your players are happy in general because in the end it is a service and you want customers to be satisfied. It can help you better anticipate retention issues.

Ada (Clue)

• Clue's typical retention metric is month over month so they use a proxy retention metric. When conducting tests, instead of looking at the first 90 days (onboarding - 3 cycles) they look at the first week or two weeks and then use predictions.
• For lifecycle campaign metrics they do their A/B tests iteratively in order to target people on a certain day of their lifecycle.
• Look at the types of subscriptions and optimize for 12 months (they have 1/6/12): bring more value and get the money upfront for cashflow.
• Engagement metric: how many times a user has tracked data per month.

Q&A

Ads on questionable websites? How to do deal with groups like Sleeping Giants? How to be politically correct and economically profitable?

• Steve - you need to stay on top of that and try to work with more reputable partners that give you more options

How to cope with trying to grow many apps in many countries at the same time from a localization perspective?

• Start by focusing on the countries that are more likely to make a big impact: France, Germany, Japan, etc.
• If your app is already localized you should have some content that can be shared with your ASO manager.
• [💎@48:28] Clue had success grouping some regions that behave similarly when running A/B tests concepts (translating but not localizing for each country) : countries in the EU vs. US vs. Korea/Japan.

Paid retargeting for users predicted to churn?

• Ads (Clue) [💎@49:17] If Clue sees a drop off in tracking or drop off in sessions per month for a user they try to reactivate with push. Example: "Are you getting the most value? Do you want to activate a reminder?"