What is Retargeting and Why is it Important?


Retargeting mobile programmatic campaigns

The importance of mobile has grown exponentially over the past 5 years, and people are engaging with their mobile phones and apps throughout their day and night. Initially, app marketing success was measured by the number of ad clicks, but then the focus shifted to the number of people who installed the app. Now, the emphasis is on engagement, and retaining valuable users is the key success metric in the mobile app ecosystem.

So how do we deal with this ever-changing landscape? The strategy to master first is retargeting.

What is Retargeting?

Retargeting allows app marketers to maximize existing app users. For active users, this can be used to make them convert or do a new engagement in the app, and to prevent them from churn. Rather than trying to find users blindly, you spend your budget on acquiring users who are most similar to your high Lifetime Value (LTV) users. This can help drive your mobile app marketing return on ad spend (ROAS).

This mobile retargeting strategy also aims to drive previous users of an app to become active app users again. Whether these users haven’t opened the app in a while, or have uninstalled the app, the goal is to retarget those previously active users and convert them into loyal users again.

At Aarki, we have 5 pillars of retargeting: audiences, creative, machine learning models, incremental lift test, and match rate.

Quick links: 

  1. Audiences 
  2. Creative
  3. Machine Learning Models
  4. Incremental Lift Test
  5. Match Rate

Audiences

Running a retargeting campaign is a way to rekindle the trust of your target audience and deepen your relationship with your users. The first step is to get to know the audience you want to retarget. Audience data serves as your campaign’s backbone and enables the delivery of the final product (the creative ad) with the right target fit. The better you know your audience, the better you understand their interests, and the more likely you are to drive conversion rates.

Audience data includes information about the user, such as their spend level, number of days since they last opened the app, items they have added to their cart, recent product views, uptake of special offers, and many more. At Aarki, we leverage audience data to understand users and segment them according to their different types and varying needs.

Creating Retargeting Audience Lists

The outcome of a retargeting click is simple: the ad redirects the user to the app store if they uninstalled the app, or it will launch the app if it remains installed on their device. But to deliver this outcome, the most important step is to create retargeting audience lists and segments.

There are two ways to create retargeting audience lists:

  • Create user lists using postbacks (installs, opens, revenue events) sent by the advertisers
    • Advertisers enable non-attributed data to identify user behavior
    • Using open events, Aarki will build user lists that includes users who have been inactive for the last X days and will set different targets per user list
    • Advertisers can also provide criteria for user list building, so long as postbacks are available
  • Aarki has also made it easy for advertisers to create user lists internally, or via data management platforms (DMPs), to share them with Aarki
    • Aarki receives segments via its audience API
    • Aarki is fully integrated with mParticle to receive audience segments
    • Aarki can receive the retargeting segments via the cloud or email
    • Segments created on MMP’s audience builder can be shared with Aarki

The fully automated data transfer process via API integrations makes audiences dynamic, updating segments based on new user engagement. Say goodbye to manual updates of .csv files and enjoy these technological advancements.

Identifying Segments and Specific Targeting Groups

Once you have your retargeting audience lists, it's time to define how to slice and dice your target audience data to identify lucrative segments. The right segmentation strategy is an essential component to determining the success of your app marketing campaign. But how to best identify which segments to target?

Sharing as much user behavior data as possible makes it easier to analyze and identify audience segments. Based on the data provided, coupled with Aarki’s deep expertise, our team will model your data and make suggestions on which segments will be the most effective to target.

For instance, if the advertiser shares two data points with us — days lapsed and spend level, we can intersect the data and arrive at a segment of inactive medium engaged users, as illustrated in the first Venn diagram. By adding additional data points we can identify further specific segments and we can use these to target users who are more likely to spend more and engage with the app for longer.

fig 1fig 2

Whether we are discussing non-spending users or VIP users, keeping them loyal and engaged is an ongoing challenge. Thus, it’s essential to really understand the user, predict their behavior, and create personalized messaging to win their hearts. Having more segments allows us to drive better campaign performance and a higher ROAS.

Best Approaches for Retargeting Your Audience

Some approaches work better than others when running retargeting campaigns. Below, we highlight some tips to effectively retarget your audience, maximize your campaign’s performance, and successfully drive growth.

  • Don’t just retarget lapsed users - Most advertisers use retargeting to bring back users who either have stopped playing or uninstalled the app. But you don’t need to wait for users to lapse—you can also use retargeting to prevent this from happening in the first place. Users at risk of churn can be identified using different engagement data points such as retention and purchase.
    Active users can be retargeted to encourage them to re-engage with your app or perform specific new-to-them actions, like register, purchase, play a specific level or try a new feature of the app. This will help you boost user engagement.
  • Choose the right lapsed window - Another important factor to examine is the lapsed window. This tells us how long the user has been unengaged with your app. The lapsed window will vary based on your app category and the action you’d like the user to take. For example, we use a lapsed window of at least 2 days for the social casino category, but for casual apps, the window is at least 7 days.
  • Consider clustering - Performance may still vary based on how you group your users. Clustering is helpful for both optimization and data analysis. It will determine why you are not seeing an incremental lift in your retargeting campaigns or which area has caused the underperformance.

Creative

The next essential pillar of retargeting is producing the right creative. It is the cornerstone of any effective campaign and it determines whether or not users will remember your ad message and take action.

At Aarki, we help app marketers achieve their retargeting goals by leveraging data to make highly interactive and personalized creatives, as personalization works best for retargeting campaigns. 

Once we have the audience data, our analytics team identifies the intersecting segments and targeting groups to get to know the users better and serve the right ads. Creative groups are then planned based on the target group’s user behavior and specific creative strategies are applied to the ads.

Non-gaming Creative Groups and Strategies

The different spend levels of your users can be used to determine your creative strategies. For example:

  • Non-spender - Remind the user of the app’s core features
  • Low-spender - Promote app features that are not regularly used, such as payment methods
  • High-spender - Offer loyalty bonuses or advertise special discounts such as private event sales

You can also try the other creative strategies for your non-gaming app below:

  1. Offer special discounts - Special holidays and paydays are perfect opportunities to offer users special discounts. You can also give them discount vouchers for wish list items or recently viewed products.
  2. Updates to wish list - Try sending them notifications when any of the items on their wish list are on sale, back in stock, or almost sold out.
  3. Provide other payment methods - Having many methods to pay makes it easier for the users to complete a purchase. 
  4. Highlight new features - Showcase elements of the app that the target users have not been exposed to before. Let them feel that something exciting awaits them. 
  5. Offer loyalty promotions - Make highly engaged users feel special by offering them loyalty promotions and encouraging them to remain engaged.
  6. Feature the newest brands in store - Update users with your latest brands and other new offerings so they won’t miss out on anything.

fitness-RT-newworkout-wphoneRT-Personalized

Gaming Creative Groups and Strategies

In the case of gaming apps, the target groups are based on data points such as the number of days since the user opened the app, level reached, and spend level. With this information, we can inform a detailed personalization strategy. For example:

  • Users new to the game - Show parts of the game yet to be explored
  • Users who stopped playing early on - Challenge the user to beat their high score, or beat more difficult levels
  • Users who ran out of free life to continue playing - Offer ways to get free energy/life, boosters, power-ups, etc.

Based on the multivariate testing and creative iterations we’ve done, here are some other creative practices which deliver successful results:

  1. Show users what’s new in your app - Continuously spark the interest of your users by updating them with your app’s latest levels, themes, quests, characters, or gameplay strategies. This has proven to be effective in keeping long-time or high-level players engaged. Showing lapsed users the progress in the game is also an effective way to keep players interested.  
  2. Use gameplay boosters - Gameplay boosters provide enhanced gaming experiences to users. Showing how special in-game currency and boosters can help them finish their quests or progress levels faster consistently produces good results. 
  3. Provide daily game bonuses - These can take the form of daily rewards (rewards for logging in consecutively) or daily spin (randomly give items/boosters/currency). This is effective with players who don’t open the app frequently as they will be enticed to log in more often for the daily bonuses.
  4. Create game events - Events that are readily accessible to new players will encourage them to play more and perhaps purchase in-game items and boosters. Events that have strict entry requirements can provide enticing competition to long-time and paying users, which will boost your ROI. 
  5. Make game characters relevant - Lapsed users will more likely re-open the app and be encouraged to play again if they resonate with the game characters. Memorable game characters used in ads can boost CTR significantly. It is essential to plan the characters well and successfully execute the creative elements that go with the characters.
  6. Help the users with some tips and tricks - Encourage users to come back by showing tips and tricks for clearing levels or by highlighting some boosters.
  7. Show unexplored game components - Feature exciting game components that may be unexplored by the users because of early abandonment, to show the users that the most interesting part of your app is yet to come.
  8. Appeal to players' emotions - Show that your audience is important to you by using "we miss you, come back" message themes, and make them feel your loyalty.  

RPG-RT-personalizedbyactivity-wphoneadventure-RT-gamemode-wphone

 

Learn more about creative strategies for specific verticals below: 

puzzle (1)storytelling (2)solitaire (1)shopping (1)

Case Study: Creative Strategies Using Users' Post-Install Data

Use different messaging for segments based on engagement

An effective way to structure your campaign is to segment your users based on their levels of engagement and produce creative that is personalized to those segments. For instance, highly engaged users will respond to an ad creative that is focused on the game’s new features, while low-engaged users respond well to reminders on how they can get freebies or daily rewards.

In a puzzle game campaign, we targeted 7 days lapsed users, who were segmented into six targeting groups based on the lifetime purchase value. For each targeting group, we ran one non-personalized and one personalized creative. The non-personalized creative was the same across all targeting groups while personalized creative varied for each targeting group. This engagement level strategy helped Aarki generate a 6% lower cost per reactivation and improved ROI by 54%.

table 1 (1)

Different messaging based on user-level attributes

A more personalized way to reach your target audience is to have dynamic creative per user, incorporating user-level attributes into the creative. Some attributes to consider for a gaming app are the highest level achieved, remaining coins, or boosters, while an eCommerce app should consider incorporating the last searched product or last bought product. With this post-install data, we create ads that dynamically optimize messaging to an individual user at ad serving time. 

This strategy is mostly used for eCommerce apps but we at Aarki have found this strategy successful with gaming apps as well. We’ve seen this type of creative strategy is most effective on highly engaged users. Do take note that hyper-personalized creative ads can also have a negative effect on the users, so choosing the right data to incorporate in your ads is vital. Don’t forget to test, test, test!

The campaign targeted lapsed users, grouped according to their purchase engagement. Using user-level attributes helped us increase ROI by 53% and improved cost per reactivation by 17%.

table 2 (1)

Machine Learning Models

Great ad creative is a waste if the campaign is not run properly. Aarki builds custom retargeting models that utilize in-app activity data from the user (gathered from before they lapse), which informs user conversion and post engagement activity predictions. Bid models are trained specifically on retargeting campaigns for the best results. These highly specific models, in turn, result in improved optimization. Advertisers can automate optimization and targeting at scale, with machine learning systems automatically learning and improving. The custom models are combined with bid optimization on a per user/per impression level to ensure the best ad is shown at the right price.

When retargeting lapsed users, we at Aarki consider the following models:

  • User Response and In-app Behavior Prediction - This model predicts user response to a specific creative as well as post-install in-app behavior (for example, retention, registration, and purchase), all in real-time. These predictions are used to calculate optimal bids to deliver against advertiser KPIs while maximizing reach.
  • Segment Membership - This model uses segments with pre-lapsed user activity to inform the prediction.
  • Cluster Category - This model utilizes app embeddings with neural-net auto-encoders and performs automatic taxonomy discovery with clustering.

Incremental Lift Test

To see the impact of your retargeting strategies on your ROI, run an incremental lift test. At Aarki, we divide the audience into two targeting groups, the treatment group and the control group. The control group is shown placebo ads or no ads at all, depending on what type of incremental lift test the client would like to use. The treatment group is shown the actual campaign. Calculate the campaign lift by comparing the outcomes of both groups. 

PSA testing

To avoid competition and a split in data, run the targeting group that is being measured using just one partner. You should also dynamically exclude active users to avoid retargeting users that have already been reactivated by other partners. 

In some cases, advertisers opt to run an incremental lift test simultaneously with different partners to see who will deliver the better performance. Below are Aarki’s recommendations for setups like this:

  • All the partners should have the same control group.
  • The targeting group should be split equally between partners.
  • When evaluating the incremental lift, you should look at the performance of the targeting group as a whole, and not each of the partners’ targeting group vs the control group.
  • To know which partner gives better performance, you must look at the return on investment, or return on ad spend (ROI/ROAS) performance, instead of the incremental ROI/ROAS.

Only you can measure the incremental lift from the whole targeting group and conduct the analysis and calculations, as each partner will not have visibility on the data of other partners running the test.

Match Rate

To know your ability to reach as many of your defined target users as possible, refer to the match rate. Aarki maximizes your match rate through our bidder and broadens your reach via our inventory supply and our wide range of creative formats. Match rate is monitored by advertisers as the basis for running any type of campaigns for iOS.

2 Extra Tips to Make Retargeting More Effective

  1. Leverage frequency caps - By limiting the maximum number of impressions a user can see within a given period of time, you will ensure that your target audience will not become ad fatigued.
  2. Optimize your conversion funnel - Establish a deeper connection between your app and the target audience through sequential advertising. Let your user progress through a sequence of ad story “experiences” across various devices. Show them relevant messages as they move through the conversion funnel.

Conclusion

Aarki has run successful retargeting campaigns for a variety of app categories. Our winning formula is simple; it is the combination of three cornerstones that drive performance at scale - machine learning, engaging creatives, and programmatic.

  • Machine learning - Leveraging proprietary machine learning, Aarki’s data scientists conduct analyses of how the users engage with an ad and how their engagement correlates with their behavior in the app. This enables Aarki to consistently run ROAS optimized campaigns for successful retargeting.
  • Engaging creatives - Creative strategies include leveraging more creative types to increase match rates and utilizing audience attributes to inform creatives. Aarki employs strategies like personalization and localization of creatives to increase ad engagement.
  • Programmatic scale - Using our integration with all the major global exchanges for programmatic/RTB, and our five global data centers, we access high-quality inventory for our clients’ campaigns at scale.

The Impact of Apple Privacy Changes on Retargeting

2021 was a year of massive change for advertisers. The Identifier for Advertisers (IDFA) was at the heart of media buying for in-app advertising, but this reality has changed with Apple’s privacy changes, specifically with the release of iOS14.5. Because of this, advertisers need to receive the user’s permission through the App Tracking Transparency (ATT) framework to track them or access their device’s advertising identifier. So how will this change affect retargeting campaigns?

While the user data is more delayed and anonymized to protect user privacy now, the means of optimization for a buying model do not fundamentally change. A buying partner can decode conversion values into in-app events, and train and deploy models optimizing for these events.   

Another source of data for campaign optimization and models comes from mobile measurement partners (MMPs). While the scale and utility of the data may vary per app campaign (and bias needs to be accounted for) privacy-compliant postbacks from users with the necessary consent can provide rich optimization data.

You can learn more in one of many articles where we interviewed members of the Aarki team to get comprehensive answers to the questions that have arisen with this change.

Why Partner with Aarki

Aarki is a 100% real-time bidding (RTB) demand-side platform (DSP). We help companies grow and re-engage their mobile users, using machine learning, data, and large customer reach. Aarki has run many retargeting campaigns utilizing a variety of ad formats, including native, video, interactive, playable, and interstitial. Leveraging a rich database of audience and user engagement data, Aarki’s data scientists use robust machine learning algorithms to find audiences who share similar interests and are most likely to engage and spend with the app. This enables Aarki to consistently deliver strong app marketing performance no matter what the app marketer’s retargeting goals are.

back to top

Interested in running a retargeting campaign? Shoot us a message here and see the results for yourself.

Topics: Marketplace Insights