In the ever-evolving landscape of mobile marketing, attribution models remain the compass for professionals who navigate through vast oceans of data.
But as privacy norms shift, these models are no longer just about leveraging data. They are about understanding and adapting to a more complex ecosystem.
These changes have not just ruffled the feathers of mobile marketers and app/game publishers, but they’ve redefined the rule book of data access and usage. We are now playing in a world where data is not always at our fingertips, a reality that presents new challenges for attribution.
In this climate of change and uncertainty, old attribution models may not serve you as efficiently.
These models relied heavily on the now diminishing stream of user-level data. That leads us to question – how can we measure the effectiveness of our marketing efforts in this brave new world of privacy-centric norms?
This article is a deep dive into different attribution models and the challenges and solutions that lie ahead. It will help you enhance your understanding of user behavior and campaign performance amidst the privacy upheavals as well as choose the best attribution platform.
Without further ado, let’s find out how each attribution model works in a privacy-conscious world.
Single-Touch Attribution: A Simplicity-Centered Approach
The single-touch attribution model is the simplest form of conversion attribution. It focuses solely on one touchpoint in the customer journey.
This model comes in two main types: first touch and last touch. Here’s how they work.
First-Touch Attribution: Mapping the Starting Point of User Journey
First-touch attribution is a single-touch attribution model that places all of the credit for a conversion on the very first touchpoint a user has with a mobile app or game. It’s like a digital version of love at first sight, where that initial click forges the connection between the user and your app.
Consider a campaign for a mobile app, launched across various channels – Google Ads, Facebook, and an in-app advertisement on a popular mobile game.
If a user discovers the app through a Google ad, and installs it as the result of seeing that ad, the first-click attribution model will attribute the conversion (in this case, install) to the Google ad.
Pros of Using First-Touch Attribution
- Identifying Effective Channels: First-touch attribution allows you to identify which marketing channels are most effective in driving initial user interest. By understanding what initially draws users in, you can optimize these channels to cast a wider net for potential users.
- Simplicity: This model is straightforward and easy to implement as it does not require complex algorithms or data processing to determine where the credit for a conversion goes.
- Beneficial for Brand Awareness Campaigns: If your goal is to expand brand recognition and reach, first-touch attribution can help identify which channels are getting your app in front of new eyes most effectively.
- Understanding Customer Acquisition Path: It provides insights into how users first find your app, giving you valuable information about your users’ acquisition journey.
- Data Privacy Friendly: In the new privacy-centric landscape, first-touch attribution can be a reliable model as it requires less user-level data compared to models that track users through multiple touchpoints.
Cons of Using First-Touch Attribution
- Overlooks Customer Journey: The model tends to overlook the entire customer journey. It does not consider the influence of subsequent interactions that may have convinced the user to convert.
- Risk of Overvaluing Certain Channels: By attributing all credit to the first touch, you may overvalue certain marketing channels or actions and undervalue others that contribute to the user’s decision-making process.
- Lack of Engagement Insights: First-touch attribution doesn’t provide information about what drives user engagement or retention – it merely tells you what sparked initial interest.
- Not Ideal for Long Conversion Paths: For apps or games with longer conversion paths, this model might not provide a comprehensive view of what works, as it does not account for interactions that occur later in the user journey.
- Potential Misallocation of Budget: Relying solely on first-touch attribution could lead to a skewed marketing budget allocation due to favoring channels that are good at initiating user journeys but may not be as effective at driving conversions.
Last Touch Attribution: The Final Step That Seals the Conversion
In the world of attribution models, last-touch attribution holds a unique position. As a single-touch model, it attributes 100% of the conversion credit to the final interaction before the conversion action.
In essence, it focuses on the last touchpoint that pushed the user to convert.
For example, let’s say a potential user first discovers your app via a Facebook post, but finally installs the app after clicking on a TikTok ad. Under the last touch attribution model, all credit for the conversion would go to the TikTok ad, because it was the last touchpoint before the user installed the app.
Pros of Using Last Touch Attribution
- Simplicity: The last touch attribution model is easy to understand and implement. That makes it suitable even for small businesses or those new to attribution models.
- Highlights Closing Channels: This model can identify the channels that are effective in closing the deal and driving users to conversion.
- Data Light: Since this model focuses on the last touchpoint, it doesn’t require extensive data tracking across multiple channels.
- Suitable for Direct Response Campaigns: If your marketing strategy focuses on direct response, where the last interaction is the most important, this model can be suitable.
- Low Cost: Compared to multi-touch models, last-touch attribution is less expensive and less resource-intensive to implement.
Cons of Using Last-Touch Attribution
- Overlooks Other Touchpoints: The last touch model disregards all other touchpoints that may have contributed to building user interest and driving them toward conversion.
- Ignores Customer Journey Complexity: In today’s multi-channel marketing environment, this model fails to capture the complex journey of a user to conversion.
- Risk of Misallocation: It can lead to misallocation of resources by overvaluing the closing channels and undervaluing the awareness-building and consideration channels.
- Limited Insights: This model offers limited insights into how different channels interact and contribute to the overall user journey.
- Favors Certain Channels: The last touch model tends to favor channels that are typically closer to the point of conversion and may undervalue upper-funnel channels.
Multi-Touch Attribution: A Comprehensive Lens on the User Journey
Multi-touch attribution is an advanced attribution model that recognizes the impact of all touchpoints in a user’s conversion journey. Unlike the simpler models, it doesn’t adhere to fixed rules for credit allocation but uses algorithms and data analysis to attribute credit based on the perceived influence of each interaction.
It treats the user journey as a collaborative symphony and assesses the importance of each instrument in delivering the final melody.
Linear Attribution: A Democratic Approach to Conversion Attribution
Linear attribution is a multi-touch attribution model that takes a democratic approach by evenly distributing the credit for a conversion across all touchpoints a user has interacted with.
Unlike the first or last touch models, this model acknowledges the influence of every interaction along the user’s journey. It treats each touchpoint as a step in a relay race, where every step contributes equally to reaching the finish line.
Suppose your mobile game is promoted through various channels like Google Ads, influencer collaborations on Instagram, or an in-app ad on another mobile game. And let’s say a user interacts with all of them before installing it. In the linear attribution model, each of these touchpoints would receive an equal share of the credit for the game installation.
Pros of Using Linear Attribution
- Holistic View of Customer Journey: By attributing equal credit to all touchpoints, this model provides a comprehensive view of the entire user journey.
- Eliminates Channel Favoritism: The linear model prevents overvaluing any single touchpoint. That promotes a balanced understanding of what influences the user decision-making process.
- Supports Multi-Channel Strategies: If you run campaigns across multiple channels, this model can help you understand the collective impact of your marketing efforts.
- Fair Budget Allocation: With its democratic approach, the linear model can guide you towards a more balanced distribution of your marketing budget across channels.
- Easy to Understand: The concept of evenly distributed attribution is simple and easy to understand, making it a good starting point for those new to attribution modeling.
Cons of Using Linear Attribution
- Lacks Precision: The Linear model doesn’t differentiate between the relative impact of different touchpoints, which might not reflect the actual influence of each interaction.
- May Overlook Influential Channels: The equal attribution might mask the performance of highly effective channels, resulting in potential underinvestment in them.
- Not Ideal for Complex Journeys: For longer, more intricate user journeys with several touchpoints, this model might oversimplify the attribution process.
- No Emphasis on Key Touchpoints: This model does not highlight crucial touchpoints that may play a pivotal role in the user’s decision-making process.
- Lack of Deeper Insights: While it provides a broad view of the user journey, the linear model falls short in delivering deeper insights into what specifically drives user engagement and conversions.
Time-Decay Attribution: Weighing the Recent Over the Remote
The time-decay attribution model, as the name suggests, attributes more credit to the touchpoints that occur closer to the time of conversion.
It works on the premise that the closer a user gets to the conversion event, the more significant the interactions become.
Let’s walk through an example.
A user initially learns about your app through a Facebook ad, checks out the YouTube review, sees the in-app ad, and finally installs your app after receiving a push notification.
With time-decay attribution, the push notification would receive the most credit, with each previous touchpoint receiving progressively less based on its temporal distance from the conversion.
Pros of Using Time-Decay Attribution
- Prioritizes Conversion-Close Activities: By assigning more credit to recent touchpoints, this model can help identify which activities are most effective in sealing the conversion deal.
- Reflects User Behavior: This model aligns well with typical user behavior, acknowledging that the impact of marketing touchpoints often diminishes over time.
- Great for Short Life Cycles: For apps or games with shorter life cycles, this model can provide accurate insights into what drives conversions.
- Balanced Approach: Unlike first or last touch models, time-decay takes into account all touchpoints while still differentiating their importance.
- Supports ROI Calculation: By focusing on the end of the user journey, it supports the calculation of ROI for activities close to the conversion.
Cons of Using Time-Decay Attribution
- Undervalues Initial Interactions: This model might undervalue the importance of initial interactions that sparked the user’s interest.
- Not Ideal for Long Cycles: For apps or games with longer user journeys, this model might fail to accurately represent the value of early touchpoints.
- Complex to Implement: The time-decay model can be more complex to set up and manage, compared to simpler models like first or last touch.
- Requires Advanced Tracking: This model necessitates more sophisticated tracking across the user journey, which can be challenging with privacy-centric norms.
- Overemphasis on Recent Activities: The focus on recent interactions may lead to over-investment in end-of-funnel activities, potentially neglecting the crucial early stages of the user journey.
Position-Based Attribution: The Power of First Impressions and Lasting Impressions
The position-based attribution model, also known as the U-shaped model, acknowledges the pivotal roles that the first and last touchpoints play in the conversion journey.
It gives these interactions more weight, typically allocating 40% of the credit to each, with the remaining 20% evenly spread across any middle interactions.
Similar to previous examples, let’s say you’re promoting your app through multiple channels.
A user stumbles upon your Google ad, checks out the influencer’s post, views the in-game ad, and eventually downloads your game after clicking on a TikTok ad. In the position-based attribution model, the Google ad and the TikTok ad would each receive 40% of the credit, while the influencer’s post and the in-game ad would share the remaining 20%.
Pros of Using Position-Based Attribution
- Highlights Crucial Touchpoints: By assigning greater weight to the first and last interactions, this model acknowledges the importance of both initiating and closing the conversion journey.
- Fairer Distribution of Credit: Unlike single touchpoint models, position-based attribution appreciates the contribution of multiple interactions along the user journey.
- Balances Acquisition and Conversion: This model provides insights into both what attracts users initially and what finally convinces them to convert.
- Flexibility: The distribution of credit in this model can be adjusted based on specific business needs, adding a layer of flexibility.
- Good for Multi-Channel Campaigns: It provides a more nuanced understanding of the efficacy of multi-channel campaigns compared to single touchpoint models.
Cons of Using Position-Based Attribution
- Overlooks Middle Touchpoints: By design, this model may undervalue the importance of middle interactions that keep the user engaged throughout the journey.
- Not Ideal for Complex Journeys: For intricate user journeys with multiple important touchpoints, this model might oversimplify the attribution.
- Requires Advanced Tracking: Implementing position-based attribution necessitates sophisticated tracking and data analysis capabilities.
- Less Suitable for Long Sales Cycles: If your user journey is drawn out over a significant length of time, this model might not accurately represent the influence of each touchpoint.
- Fixed Distribution Might Not Reflect Reality: The preset distribution of credit in this model might not accurately reflect the actual influence of various touchpoints in specific scenarios.
View-Through Attribution: Silent Influence
Next up on our list of attribution models is view-through attribution. It is a model that recognizes the impact of impressions – the number of times a user has viewed, but not necessarily interacted with, an ad.
This model acknowledges that a user’s decision to convert may be influenced by seeing an ad, even if they don’t directly click on it.
Let’s consider an example.
You have a new mobile game promoted via a gameplay video ad on social media. Many users see the ad, but they don’t click on it. Instead, they directly search for the game on their app store later and install it.
With view-through attribution, the video ad receives credit for these conversions because it sparked user interest.
Pros of Using View-Through Attribution
- Values Silent Impact: This model recognizes that impressions can have a substantial impact on user behavior, even without direct interaction.
- Enhances Understanding of User Behavior: View-through attribution can provide deeper insights into user behavior, especially in situations where direct click-throughs are low.
- Supports Branding Campaigns: For campaigns focused on raising brand awareness rather than immediate conversion, this model offers valuable metrics for success.
- More Comprehensive Attribution: By accounting for view-through conversions, you can gain a more comprehensive understanding of your campaigns’ effectiveness.
- Ideal for Display and Video Ads: This model is especially useful in measuring the impact of display and video ads, which sometimes aim to leave a lasting impression rather than prompt immediate action.
Cons of Using View-Through Attribution
- Potential Overvaluation of Impressions: It can be challenging to definitively attribute a conversion to a viewed ad.
- Relies on Advanced Tracking: Implementing view-through attribution requires advanced tracking capabilities and precise timing to ensure accurate attribution.
- Risk of Double Counting: If a user views an ad but also clicks on a different ad before converting, there’s a risk of double counting conversions in the presence of other attribution models.
- Vulnerable to Ad Fraud: View-through conversions could be inflated by fraudulent practices like hidden ads or stacked pixels.
- Needs Careful Setting of Lookback Windows: Determining the right lookback window, the period during which a view-through conversion is attributed to an ad, can be challenging and might influence the accuracy of your attribution.
Data-Driven Attribution: Leveraging Machine Learning for Conversion Insights
At the cutting edge of attribution models, we find data-driven attribution.
This model uses machine learning algorithms and statistical analysis to assign conversion credit to different touchpoints, based on their influence throughout the customer journey.
It’s a dynamic model, constantly learning from new data and adjusting the credit assigned to various interactions.
It evaluates the influence of each touchpoint on the user’s decision to convert and assigns credit accordingly.
Pros of Using Data-Driven Attribution
- Dynamic Attribution: Data-driven attribution adapts to your changing marketing environment, constantly learning from new data to provide the most accurate attribution.
- Deep Insights: This model can uncover relationships and interactions between various channels that may not be apparent with simpler models.
- Personalized Attribution: It provides a level of personalization, considering the unique aspects of your business and your marketing strategy.
- Optimized Marketing Spend: By providing granular insights into the effectiveness of each channel, it helps optimize your marketing budget allocation.
- Improved ROI: By identifying high-performing channels, this model can significantly improve the return on your marketing investments.
Cons of Using Data-Driven Attribution
- Requires Large Volumes of Data: The effectiveness of this model is tied to the volume and quality of data it can learn from. Smaller businesses may not generate enough data for meaningful analysis.
- Complex Implementation: Data-driven attribution requires advanced analytics capabilities and technical expertise to implement and manage.
- Potentially High Costs: Developing or adopting a data-driven attribution model can be expensive, particularly for smaller app publishers and game studios.
- Data Privacy Concerns: The collection and processing of large amounts of user-level data required for this model may run into privacy issues, especially with growing global privacy regulations.
- Black Box Model: The complex algorithms used can make this model a ‘black box,’ making it hard to understand exactly how credit is being assigned, which can complicate transparency and accountability.
SKAdNetwork: Apple’s Response to a Privacy-First World
In a world where privacy is becoming a paramount concern, Apple’s SKAdNetwork attribution model is a direct response to the changing landscape.
It aims to balance privacy with advertisers’ need for user acquisition data.
SKAdNetwork, rolled out with iOS 14.5, attributes app installs to their advertising sources, without revealing any user-level or device-level data. When an app is installed, the ad network receives a postback containing information about the ad campaign that led to conversion, but no data that can personally identify the user.
When a user installs the game, Apple sends a conversion postback to the responsible network. The postback includes campaign and ad group identifiers, but no user-specific data.
That aligns with Apple’s commitment to user privacy.
Pros of Using SKAdNetwork Attribution
- Privacy-Centric: SKAdNetwork respects user privacy, aligning with growing global privacy norms, and adhering to Apple’s privacy-first approach.
- Reduced Risk of Fraud: By directly managing attribution, Apple eliminates the risk of third-party fraud.
- Standardization: SKAdNetwork helps standardize the attribution process across all iOS apps.
- Transparent and Trustworthy: The system is considered reliable and trustworthy, as it’s run by Apple, a globally respected tech giant.
- Compliant with App Tracking Transparency (ATT) Framework: SKAdNetwork operates within the guidelines of Apple’s ATT framework, making it compliant and future-proof.
Cons of Using SKAdNetwork Attribution
- Limited Data: SKAdNetwork provides significantly less user-level data compared to traditional attribution models, limiting granular analysis.
- Delayed Reporting: There can be a delay of up to 24 hours before conversion data is reported, potentially slowing down optimization efforts.
- Reduced Precision: It does not support granular targeting or measurement, which reduces precision in marketing efforts.
- No User-Level ROAS: The inability to track user-level data hinders calculating user-level return on ad spend (ROAS), impacting campaign optimization.
- Dependent on Apple: As it’s an Apple-owned framework, any changes or updates are solely at Apple’s discretion, giving advertisers little control.
Attribution Models: Adapting to a Privacy-First Environment
With the rise of privacy concerns and subsequent regulations, navigating the mobile attribution landscape has indeed become a more complex task. App and game developers, marketers, and advertisers find themselves grappling with a new normal, where the goalposts have shifted.
The need of the hour is to embrace attribution models that respect user privacy while providing effective insights into user acquisition and conversions.
Firstly, last-touch attribution and first-click attribution models, despite their limitations, can provide significant insights for those beginning their attribution journey. They are straightforward, relatively easy to implement, and do not necessarily require extensive user-level data. They are a starting point for developers and marketers who are just dipping their toes in the realm of attribution.
However, for a more holistic understanding of the customer journey, considering multiple touchpoints, multi-touch attribution models like linear, time-decay, and position-based attribution could be beneficial. These models can provide a broader view of the customer journey. But they must be used with care and ensure user data privacy is maintained.
In the case of more sophisticated analytics capabilities, algorithmic attribution could provide an edge by using statistical analysis to assign conversion credits. Still, its dependency on extensive user-level data could pose privacy concerns.
Privacy First Attribution Models
Apple’s SKAdNetwork is a model that deserves special mention.
It is a privacy-centric attribution solution, designed explicitly for the Apple ecosystem. Developers and marketers focusing on iOS apps must become familiar with this framework, given Apple’s push for privacy and its widespread user base.
Data-driven attribution, a more complex model offered by Google, could be a good fit for those working primarily within the Google ecosystem. Like SKAdNetwork, it respects user privacy while providing valuable insights for marketers and developers.
In essence, the choice of an attribution model must be a careful balance between the need for actionable insights and respect for user privacy. Adapting to privacy changes isn’t merely about compliance; it’s an opportunity to build trust with users and show that you respect and protect their privacy.
As the mobile industry evolves, so must our understanding and application of attribution models, always with the user at the center of our strategies.