IS MULTI-TOUCH ATTRIBUTION STILL POSSIBLE IN TODAY’S PRIVACY-CENTRIC WORLD?
B2C (Business-to-Consumer) businesses face increasing challenges in attributing conversions to their online marketing channels, particularly due to the increasing impact of privacy regulations and browser technologies like Intelligent Tracking Prevention (ITP) and General Data Protection Regulation (GDPR). These developments have significantly complicated the ability of businesses to track users across the web, impacting marketing attribution models and strategies.
In this guide we’ll explore what those challenges mean to online marketers today, explore some of the alternatives and ways to address those challenges and answer the question as to whether there is still any value in multi-touch marketing attribution today.
TABLE OF CONTENTS
Key Challenges FOR MULTI-TOUCH ATTRIBUTION TODAY:
Intelligent Tracking Prevention (ITP):
Third-Party Cookie Restrictions: ITP, implemented by browsers like Safari and Firefox, limits the lifespan of third-party cookies or blocks them entirely. This prevents businesses from tracking user behaviour across different websites, which is crucial for attributing conversions to specific marketing channels.
First-Party Cookie Lifespan: Even first-party cookies, which are set by the website a user directly visits, can have their lifespan restricted by ITP. For example, in some cases, these cookies might expire after just seven days, making it difficult to attribute conversions that occur over a longer period.
GDPR and Other Privacy Regulations:
Consent Requirements: GDPR and similar regulations require that businesses obtain explicit consent from users before tracking their data. This means that users can opt out of tracking altogether, leading to incomplete data for attribution.
Data Anonymization and Minimization: GDPR emphasises data minimisation, meaning that businesses can only collect data that is strictly necessary. This limits the granularity of the data collected, making it harder to attribute conversions accurately.
Browser Privacy Features:
Ad Blockers: Many users employ ad blockers that prevent tracking scripts from loading, further reducing the data available for attribution.
Privacy-Centric Browsers: Browsers like Brave and extensions like Firefox's Enhanced Tracking Protection (ETP) go even further in blocking tracking technologies, making attribution more challenging.
Impact on Attribution Models:
Reduced Accuracy: The inability to track users across different sessions and websites leads to fragmented data, making it difficult to attribute conversions to specific campaigns or channels accurately.
Shortened Attribution Windows: With the reduced lifespan of cookies, businesses have a smaller window in which they can attribute conversions, which might not accurately reflect the customer journey, especially for high-consideration purchases.
Increased Reliance on First-Party Data: As third-party data becomes less reliable, businesses must focus on collecting and utilizing first-party data (e.g., from CRM systems) to build their attribution models.
Adaptation Strategies:
Server-Side Tracking: Moving tracking from the client-side (in the browser) to the server-side can help bypass some restrictions imposed by browsers. This involves capturing data when users interact with a website and storing it on the server, though it still must comply with privacy regulations.
Privacy-First Attribution Models: Developing models that respect user privacy while still providing useful insights is key. This could involve aggregated data, probabilistic models, or the use of anonymised data sets.
Customer Data Platforms (CDPs): CDPs can help businesses unify data from various sources and create a comprehensive view of the customer, which can improve attribution despite tracking limitations.
Overall, while the landscape for marketing attribution is more challenging, businesses we work with are finding ways to adapt by embracing privacy-first strategies and innovative approaches to data collection and analysis.
Interested? Find Out More!
Rittman Analytics is a boutique data analytics consultancy that works with growth-stage, mid-market and enterprise businesses in the UK, EU and North America to help level-up their marketing analytics capabilities and build world-class web performance and marketing attribution models.
If you’re looking for some help and assistance with your marketing attribution, marketing analytics or would just like to talk shop and share ideas and thoughts on what’s going on in your organisation and the wider data analytics world, contact us now to organise a 100%-free, no-obligation call — we’d love to hear from you!
IS CROSS-DEVICE TRACKING STILL POSSIBLE?
Cross-device tracking and attribution have become significantly more challenging due to increased privacy regulations, such as GDPR, CCPA, and browser features like Intelligent Tracking Prevention (ITP) and Enhanced Tracking Protection (ETP). These changes restrict the ability to track users across multiple devices and sessions, which in turn impacts the accuracy of multi-touch attribution (MTA).
Impact on Cross-Device Tracking and Attribution
Restricted Third-Party Cookies:
Limited Cross-Device Tracking: The blocking or restriction of third-party cookies by browsers like Safari, Firefox, and even Chrome significantly hinders the ability to track users across different devices. This is because third-party cookies were a primary method for identifying and following users across sites and devices.
Reduced User Identification: With third-party cookies blocked, it becomes difficult to link a user’s interactions on different devices (e.g., mobile phone, desktop, tablet) to a single identity, which is crucial for accurate cross-device attribution.
Shortened Cookie Lifespans:
Loss of Long-Term Tracking: Even first-party cookies are now subject to restrictions, such as reduced lifespans (e.g., 7 days in Safari with ITP). This limits the ability to track users who take longer to convert or who switch between devices over extended periods.
Device-Based Consent Differences:
Inconsistent Consent Across Devices: Users may provide different levels of consent on different devices, which can lead to partial tracking. For example, a user might consent to tracking on their desktop but decline it on their mobile device, fragmenting the data available for cross-device attribution.
Privacy-Focused Browsers and Extensions:
Increased Privacy Measures: Privacy-focused browsers (like Brave) and extensions (like uBlock Origin) block tracking mechanisms that would normally allow cross-device tracking, further complicating attribution efforts.
Workarounds and Strategies to Mitigate Impact on Multi-Touch Attribution
Leverage First-Party Data and User Authentication:
Encourage User Logins: One of the most effective ways to achieve cross-device tracking is by encouraging users to log in across all their devices. When users are authenticated, you can link their activity across devices to a single user ID, allowing for more accurate attribution.
First-Party Data Integration: Use first-party data from your CRM, loyalty programs, or email systems to track users across devices. Integrating this data with your analytics platform can help bridge the gap left by the loss of third-party cookie tracking.
Use Server-Side Tracking:
Server-Side Data Collection: Implement server-side tracking to maintain control over data collection and extend the lifespan of identifiers that can be used across devices. Server-side tracking can bypass some of the restrictions imposed by browsers on client-side cookies.
Linking User Interactions: Server-side tracking allows you to link user interactions across devices more effectively, as the data is processed and stored on your servers, which are not subject to the same restrictions as browser-based tracking.
Probabilistic Matching:
Statistical Techniques: Use probabilistic matching techniques that rely on aggregated data, such as IP address, browser type, and other non-identifying signals, to estimate when interactions on different devices might belong to the same user.
Device Fingerprinting: Although increasingly limited by privacy regulations, device fingerprinting can be used to track users across devices by analyzing a combination of attributes (e.g., screen resolution, device type). However, be cautious with this method as it may raise privacy concerns and face regulatory challenges.
Utilize Customer Data Platforms (CDPs):
Unified Customer Profiles: Implement a CDP that aggregates data from various sources, including web, mobile apps, CRM, and offline interactions. CDPs can help create a unified customer profile that spans multiple devices and touchpoints, aiding in cross-device attribution.
Identity Resolution: Use the identity resolution capabilities of CDPs to match and merge user profiles across devices, ensuring that interactions are attributed to the correct user regardless of the device used.
Use Advanced Attribution Models:
Data-Driven Attribution (DDA): Implement data-driven attribution models that rely on machine learning to evaluate the contribution of different touchpoints, even when cross-device tracking is incomplete. DDA models can help distribute credit more accurately across channels.
Hybrid Attribution Models: Combine deterministic data (e.g., logged-in user behavior) with probabilistic models to create a hybrid attribution model that can better estimate cross-device behavior.
Adopt Privacy-Friendly Alternatives:
Google’s Consent Mode and Enhanced Conversions: Google’s Consent Mode allows you to adjust the behavior of your tracking scripts based on user consent, providing more accurate modeling of user behavior while respecting privacy choices. Enhanced Conversions can improve the accuracy of conversion tracking by leveraging first-party data.
Anonymized Aggregated Data: Focus on using anonymized and aggregated data to understand cross-device patterns at a higher level, rather than trying to track individual users. This can still provide valuable insights without infringing on privacy.
Analyze Aggregated and Trend Data:
Segment-Based Analysis: Rather than focusing on individual user journeys, analyze how different user segments (e.g., by device type or entry channel) behave across devices. This approach can reveal trends and patterns that inform cross-device attribution.
Time-Based Cohorts: Use time-based cohort analysis to study user behavior patterns across devices, even if you cannot track individual users. For instance, compare conversion rates of users who start their journey on mobile and later convert on desktop.
Conduct Incrementality Tests:
Controlled Experiments: Run incrementality tests to measure the impact of specific channels or devices on overall conversions. By comparing test and control groups, you can infer the contribution of cross-device interactions even without direct tracking.
Geo-Targeted Testing: Perform geo-targeted experiments where you expose users in different regions to varying levels of cross-device marketing efforts to estimate the incremental lift.
While the landscape for cross-device tracking and attribution has become more complex due to privacy regulations and browser restrictions, businesses can still derive valuable insights by adapting their strategies:
Enhance first-party data collection through user authentication and server-side tracking.
Leverage probabilistic models and identity resolution technologies to estimate cross-device behavior.
Adopt privacy-friendly tools and practices that comply with regulations while still providing useful attribution data.
By combining these techniques, you can still achieve meaningful multi-touch attribution, even in an environment where cross-device tracking is increasingly difficult.
What is server-side tracking and how can it help?
Setting up server-side tracking in Google Analytics 4 (GA4) can help a B2C business address some of the challenges posed by browser privacy restrictions and consent requirements, such as those imposed by ITP, GDPR, and other regulations.
Setting Up Server-Side Tracking in GA4
Set Up a Cloud Environment:
Google Cloud Platform (GCP): Create a Google Cloud Project. GA4's server-side tracking often uses Google Cloud Functions, Google App Engine, or other services to process data.
Server Environment: You could also use your own server or another cloud provider. Ensure the environment can handle incoming data and has the necessary security measures.
Set Up a Tagging Server:
Google Tag Manager (GTM) Server-Side Container: Create a server-side container in GTM. This container will handle tracking requests on the server rather than on the user's browser.
Deploy the GTM Server-Side Container: Deploy this container on your cloud environment or server. Google provides detailed documentation for deploying GTM server-side containers on Google Cloud.
Modify Existing GA4 Tags:
Update GA4 Tags in GTM Web Container: Modify your existing GA4 tags in the GTM web container to send data to the server-side container instead of directly to Google Analytics. This involves changing the endpoint URL to the one corresponding to your server-side setup.
Customize Tagging: Configure the server-side GTM container to process incoming data. You can filter, transform, or enrich the data as needed before sending it to GA4.
Set Up Custom Endpoints:
Custom Data Collection Endpoints: Use the server-side environment to collect data via custom endpoints. This helps bypass browser restrictions by capturing data directly on your server.
Event Transformation: Implement logic to handle specific events and map them to GA4’s data model before forwarding the data to GA4.
Send Data to GA4:
Forward Processed Data: Once the data is processed on the server, it is forwarded to GA4 using the Measurement Protocol, which allows you to send events directly to GA4 from your server environment.
Configure GA4: Ensure that GA4 is configured to recognize and process these server-side events correctly.
Overcoming Restrictions and Consent Issues
Bypassing Browser-Based Tracking Restrictions:
ITP and Other Browser Privacy Features: Server-side tracking mitigates the impact of browser-based restrictions like ITP by moving the tracking logic away from the client-side (browser) to the server-side. This means that even if a browser restricts or blocks client-side cookies, the server can still capture and process the necessary data.
Extended Data Retention:
Cookie Lifespan Management: With server-side tracking, you control the cookies and data retention policies on your server, which can extend beyond the limitations imposed by browsers like Safari that limit cookie lifespans. This enables more accurate tracking of long-term user journeys.
Consent Management Compliance:
GDPR and Other Privacy Laws: Server-side tracking still requires consent, but it allows for more granular control over data collection and processing. You can implement robust consent management processes to ensure that data is only collected and processed if the user has granted the necessary permissions.
First-Party Data Collection: Server-side tracking enables better handling of first-party data, ensuring compliance with regulations while still gathering valuable insights.
Improved Data Accuracy and Security:
Data Integrity: By processing data on your server, you can reduce data loss that may occur due to ad blockers or other client-side interruptions.
Secure Data Transfer: Data sent from the client to your server and then to GA4 can be encrypted and handled according to strict security protocols, ensuring compliance with privacy regulations.
Customized Data Processing:
Data Minimization and Anonymization: Server-side tracking allows for custom data processing before sending it to GA4. You can apply anonymization techniques, data minimization principles, or other transformations to ensure that only necessary data is collected and stored.
Limitations of Server-Side Tracking
While server-side tracking provides significant benefits, it is not a complete solution to all privacy challenges:
Consent Requirements Remain: Even with server-side tracking, GDPR and other privacy laws still require that you obtain user consent before tracking begins.
Increased Complexity: Setting up and maintaining a server-side tracking environment is more complex and resource-intensive compared to traditional client-side tracking.
Potential Data Gaps: Some user interactions that occur entirely in the browser without hitting the server may still be missed unless explicitly tracked.
What specifically is the issue around cookie expiry?
The issue around the expiry timelines of first-party cookies primarily revolves around the limitations imposed by certain browsers, particularly with Intelligent Tracking Prevention (ITP) implemented by Safari and similar technologies in other browsers like Firefox. These measures aim to enhance user privacy by restricting the duration that first-party cookies can persist on a user's device, which directly impacts how long businesses can track and attribute user activity.
Specific Issues with First-Party Cookie Expiry Timelines
Shortened Cookie Lifespan:
Default Expiry Reduction: Traditionally, first-party cookies could have expiry dates set for months or even years, allowing businesses to track user behavior and attribute conversions over a long period. However, with ITP, browsers like Safari have drastically reduced the lifespan of first-party cookies. In some cases, the cookies may only last for seven days or less.
Impact on Attribution Windows: This shortened lifespan means that if a user interacts with a website but does not convert within the short cookie window, the tracking information may expire before the conversion occurs. This makes it difficult for businesses to attribute that conversion to the correct marketing channel or campaign.
Impact on Multi-Visit Conversion Journeys:
Complex Customer Journeys: Many B2C businesses rely on users making multiple visits to their websites before making a purchase. For example, a customer might research a product, return later to compare prices, and then finally make a purchase. If the first-party cookies expire between these visits, the business loses the ability to connect these interactions, leading to inaccurate attribution and reporting.
Re-engagement Campaigns: Marketing strategies like retargeting, which depend on recognizing users who have previously visited a site, become less effective when first-party cookies expire quickly. Users may be treated as new visitors, leading to redundant ads and a less personalized experience.
Challenges in Cohort Analysis and Segmentation:
Loss of Cohort Data: Cohort analysis, which tracks groups of users over time, becomes less reliable when the cookie duration is shortened. Businesses cannot accurately track user behavior over extended periods, making it harder to analyze long-term trends or the effectiveness of retention strategies.
Segmentation Difficulties: Segmenting users based on their behavior or history is also impacted. If a user’s history is erased after just a few days, it becomes challenging to build meaningful segments for targeted marketing or personalized experiences.
Data Integrity and Reporting Issues:
Incomplete Data Sets: Short-lived cookies lead to incomplete data, where businesses might only capture a fragment of the user’s journey. This fragmented data can skew analytics, leading to incorrect conclusions and decisions based on incomplete information.
Inaccurate Metrics: Metrics like customer lifetime value (CLTV), return on ad spend (ROAS), and even basic metrics like return visits may be underreported or inaccurately calculated due to the frequent expiry of cookies.
Workarounds and Mitigations
While the shortened lifespan of first-party cookies presents significant challenges, businesses can take certain steps to mitigate these effects:
Server-Side Tracking: As discussed earlier, moving tracking to the server-side can help maintain data continuity even when browser-based cookies expire.
Local Storage and Alternative Storage Mechanisms: Some businesses use local storage or other browser storage options that may not be as heavily restricted as cookies. However, these methods must still comply with privacy regulations and may not fully resolve the issue.
Frequent Data Synchronization: Businesses can implement more frequent data synchronization techniques, where user data is frequently sent to the server before the cookie expires. This approach can help capture and store data, but it may not fully overcome the challenges for longer attribution windows.
Consent Management: Obtaining explicit user consent for extended tracking can sometimes allow businesses to set longer-lived cookies, although this is increasingly difficult with privacy-conscious users and strict regulations.
In summary, the issue with the expiry timelines of first-party cookies mainly centers on the reduced ability to track user behavior over time, leading to challenges in accurate attribution, segmentation, and data analysis. These restrictions are a direct response to increasing privacy concerns and are a key factor that businesses need to address in their analytics and marketing strategies.
CAN TECHNIQUES SUCH AS VISITOR-SPECIFIC CUSTOM SUBDOMAINS HELP HERE?
Using custom subdomains uniquely named for each visitor can be a technique to help with properly attributing referral traffic, but it’s not without its complexities and limitations. This technique can assist in bypassing some tracking restrictions, particularly those related to cross-site tracking and cookie policies imposed by browsers like Safari's Intelligent Tracking Prevention (ITP).
Unique Subdomains for Each Visitor:
By assigning each visitor a unique subdomain (e.g.,
visitor123.yourdomain.com), the traffic can be attributed more accurately because cookies set on these subdomains are considered first-party cookies. Since the subdomain is part of your main domain, the cookies set under it are not subject to the same restrictions as third-party cookies.
Cookie Persistence Across Subdomains:
Since the cookies are tied to a unique subdomain under your main domain, they can persist across sessions and be used to track the visitor's interactions over time, even if the user revisits the site after a few days. This can help extend the lifespan of the tracking cookies beyond the usual limits imposed by ITP or other privacy features.
Referral Attribution:
When a visitor is assigned a unique subdomain, all traffic that comes through this subdomain can be accurately attributed to the original referral source. This avoids the issue of losing referral data due to redirection or due to the referrer being stripped by privacy settings in browsers.
Benefits of This Approach
Improved Attribution Accuracy: By maintaining cookies in a first-party context using a subdomain, you can achieve more reliable tracking and referral attribution over time.
Bypassing ITP Restrictions: Unique subdomains are less likely to be treated as cross-site tracking, which helps bypass the short cookie lifespan imposed by ITP on third-party cookies.
Enhanced User Tracking: This method allows for better tracking of users across multiple sessions, potentially improving the accuracy of your analytics and conversion tracking.
Challenges and Limitations
Complex Implementation:
Technical Overhead: Implementing a system where each visitor is assigned a unique subdomain involves significant technical overhead. You need infrastructure to dynamically create and manage subdomains, ensure that the right subdomain is used on return visits, and handle DNS settings.
SSL Management: Managing SSL certificates for dynamically generated subdomains can be challenging, although solutions like wildcard certificates or Let's Encrypt’s automation can help.
Potential Impact on SEO:
Search Engine Crawling: Search engines might treat each unique subdomain as a separate entity, which could dilute the SEO value of your main domain if not properly managed.
Canonical Issues: Ensuring that the main domain is treated as the canonical version of the site is crucial to avoid SEO penalties.
User Experience Considerations:
URL Appearance: Some users might find it suspicious or confusing to be redirected to a unique subdomain, potentially impacting trust and user experience.
Bookmarking and Sharing: Users who bookmark or share the URL might inadvertently share their unique subdomain, which could cause issues if another user tries to access it.
Compliance with Privacy Regulations:
GDPR Compliance: Even with this technique, GDPR and similar regulations still require you to obtain explicit consent before tracking users. Using unique subdomains does not exempt you from these requirements, and you must still ensure that users are informed and have consented to the use of cookies.
Browser Behavior Variability:
Inconsistent Browser Support: Some browsers may handle these subdomains differently, and future browser updates could introduce new restrictions that affect this technique.
Alternative or Complementary Approaches
First-Party Tracking with Server-Side Processing: Server-side tracking remains a robust option, allowing for greater control over data collection and minimizing reliance on client-side cookies.
Enhanced URL Parameters: Using unique URL parameters for tracking can be a simpler alternative to custom subdomains, although it comes with its own set of limitations and risks, such as parameter stripping by browsers.
WHAT IS THE IMPACT OF CONSENT MODE IN GA4?
When using Consent Mode with Google Analytics 4 (GA4), you can still track and analyze certain aspects of each individual visitor's journey, albeit within the boundaries set by user consent and privacy regulations. Consent Mode is designed to help businesses comply with regulations like GDPR and CCPA by adjusting how Google tags behave based on the consent status of users.
Understanding Consent Mode in GA4
Consent Mode allows you to customize the behavior of Google tags (including GA4) based on the consent status of your users for different types of data storage. Specifically, it manages two primary consent types:
ad_storage: Relates to advertising cookies.analytics_storage: Relates to analytics cookies.
By configuring these consent types, you can control how data is collected and processed by GA4, ensuring compliance with user preferences and legal requirements.
Data Tracking Based on Consent Status
1. When Users Grant Consent (granted)
Full Tracking Enabled: If a user consents to
analytics_storage, GA4 operates normally, utilizing cookies to:Identify Users: Track individual user sessions and behaviors across multiple visits.
Attribute Conversions: Accurately attribute conversions to specific marketing channels or campaigns.
Personalize Experiences: Use data to personalize content and advertising.
Data Collected Includes:
Pageviews and screenviews
Events and conversions
User demographics and interests (if enabled)
Session duration and engagement metrics
Acquisition data (source, medium, campaign)
2. When Users Deny Consent (denied)
Limited Tracking: If a user denies consent to
analytics_storage, GA4 still collects certain data but with significant limitations to protect user privacy.Data Collected Includes:
Aggregated Data: Basic information such as the number of pageviews, sessions, and users, without linking to individual identities.
Event Data: Some event tracking may still occur, but without associating events to specific users.
Anonymized Data: IP addresses are anonymized, and user identifiers are not stored.
Data Not Collected:
User-Specific Identifiers: No storage or use of client IDs or user IDs, preventing tracking of individual user journeys.
Detailed Behavioral Data: Detailed interactions and paths taken by individual users are not tracked.
Retention of Cookies: Analytics cookies are not set or used, limiting the ability to recognize returning users.
Implications for Tracking Individual Visitor Journeys
Given the above distinctions, here's what can and cannot be tracked about individual visitors' journeys when using Consent Mode with GA4:
With Consent (analytics_storage Granted)
Individual Tracking: Full tracking of individual user journeys is possible. You can:
Monitor Page Visits: Track each page a user visits during their session.
Analyze Event Sequences: Understand the sequence of actions or events a user engages with.
Attribute Conversions: Link specific actions to marketing efforts that led to conversions.
User Segmentation: Create segments based on user behavior, demographics, and more.
Without Consent (analytics_storage Denied)
No Individual Tracking: Tracking is significantly restricted to protect user privacy.
Aggregated Insights: You receive high-level metrics such as total pageviews, sessions, and basic event counts without any association to individual users.
No User Identification: Since client IDs and user IDs are not stored, you cannot reconstruct individual user journeys.
Limited Attribution: Attribution models rely on aggregated data, reducing the precision in linking conversions to specific channels or campaigns.
What Specific Data Can Still Be Tracked Without Consent
Even when users deny consent for analytics_storage, certain non-identifiable data can still be collected to provide valuable insights without infringing on individual privacy:
Basic Interaction Metrics:
Page Titles and URLs: Track which pages are being viewed.
Screen Resolutions: Understand the devices and screen sizes used by visitors.
Browser and OS Information: Gain insights into the technologies your audience uses.
Event Counts:
Non-Personal Events: Track events that do not contain personal data, such as button clicks or video plays, without linking them to specific users.
Aggregated Performance Data:
Load Times: Monitor page load performance across your site.
Error Rates: Track the frequency of errors encountered by users.
Traffic Sources:
Referrer Information: Identify where your traffic is coming from (e.g., search engines, direct visits) in an aggregated manner.
Campaign Data: Analyze the performance of marketing campaigns without associating them with individual users.
Compliance and Privacy Considerations
Data Anonymization: Even with consent, GA4 emphasizes data anonymization and minimization to protect user privacy.
User Consent Management: Implement robust consent management platforms (CMPs) to handle user preferences effectively.
Legal Compliance: Ensure that your use of GA4 and Consent Mode aligns with all relevant data protection laws and regulations.
Best Practices When Using Consent Mode with GA4
Implement a Consent Management Platform (CMP):
Use a CMP to accurately capture and manage user consent preferences for
ad_storageandanalytics_storage.
Configure Consent Mode Correctly:
Ensure that your GA4 tags are properly configured to respect the consent signals sent by your CMP.
Leverage Enhanced Measurement:
Utilize GA4’s enhanced measurement features that can operate within the constraints of consent settings, capturing valuable interactions without requiring user identification.
Focus on Aggregated Data Analysis:
When consent is limited, shift your analytical focus to aggregated trends and patterns rather than individual user behaviors.
Combine First-Party Data:
Enhance your analytics by integrating first-party data sources (e.g., CRM systems) that comply with consent and privacy regulations.
CAN SWITCHING FROM GA4 TO SNOWPLOW HELP?
Switching from Google Analytics 4 (GA4) to a service like Snowplow can indeed provide more detailed information on each visitor's journey, as Snowplow is designed to offer more control over data collection, processing, and analysis. However, this does not necessarily mean that Snowplow is "less privacy-first." Instead, it provides businesses with the tools to manage their data in a more flexible way, which can result in richer insights if implemented correctly and in compliance with privacy laws.
How Snowplow Provides More Detailed Information
Greater Control Over Data Collection:
Customizable Data Tracking: Snowplow offers extensive customization of data collection, allowing you to define and capture events and user behaviors in a highly granular manner. Unlike GA4, which comes with predefined data collection schemas and some restrictions, Snowplow enables businesses to design their own tracking schemas tailored to their specific needs.
Event-Level Data: Snowplow captures detailed event-level data, which can include custom parameters and context-specific information. This allows for a more comprehensive understanding of user interactions across the entire journey.
Server-Side and Client-Side Tracking:
Server-Side Tracking: Snowplow supports both client-side and server-side tracking. Server-side tracking helps mitigate the impact of browser privacy features like Intelligent Tracking Prevention (ITP) and can allow for more persistent tracking of users across sessions and devices.
Hybrid Data Collection: By combining client-side and server-side tracking, Snowplow can provide a more complete picture of the user journey, even in environments where client-side tracking may be limited due to ad blockers or cookie restrictions.
Data Ownership and Storage:
Full Data Ownership: With Snowplow, businesses own their data and can store it in their own infrastructure (e.g., AWS, GCP, Azure). This contrasts with GA4, where data is stored in Google's infrastructure, which comes with certain limitations on access and usage.
Raw Data Access: Snowplow provides access to raw event data, enabling detailed analysis and the ability to build custom reports and models that go beyond what GA4 offers. This raw data can be used for advanced analytics, machine learning models, and other custom applications.
No Sampling:
Complete Data Sets: Unlike GA4, which may sample data in certain reports, Snowplow does not sample data. This means you can analyze complete data sets without worrying about the potential inaccuracies introduced by sampling.
Data Enrichment and Integration:
Custom Enrichment Pipelines: Snowplow allows you to build custom enrichment pipelines that can add context to your data, such as geographic data, device information, or user attributes. This leads to a richer understanding of each user's journey.
Integration with Other Data Sources: Snowplow can be integrated with other data sources, such as CRM systems, marketing automation platforms, and databases, to create a unified view of the customer journey.
Privacy Considerations
While Snowplow offers more detailed tracking capabilities, it is not inherently "less privacy-first." The platform is highly flexible, meaning that privacy compliance is largely dependent on how you choose to configure and use it. Key privacy considerations include:
Consent Management:
Customizable Consent Framework: Snowplow allows you to implement a customizable consent framework, ensuring that data is collected only when appropriate consent has been obtained. You can design consent flows that align with GDPR, CCPA, and other privacy regulations.
Granular Consent Handling: You can configure Snowplow to handle different levels of user consent, such as opting in or out of specific types of tracking, which allows for more nuanced data collection practices.
Data Minimization and Anonymization:
Anonymization Features: Snowplow supports data anonymization features, such as IP address anonymization and pseudonymization of user identifiers, which help protect user privacy while still enabling detailed tracking.
Data Minimization: You can configure Snowplow to collect only the data necessary for your analytics needs, adhering to the principle of data minimization.
Compliance with Privacy Regulations:
Regulatory Compliance: Snowplow can be configured to comply with various privacy regulations. However, since it gives you control over how data is collected and processed, it's up to you to ensure that your implementation is compliant.
Read More on the Rittman Analytics Blog
WHAT IF THE MAJORITY OF OUR VISITORS OPT-OUT OF TRACKING?
Even with a significant portion of your users opting out of tracking, there are several strategies and techniques you can use to derive meaningful channel attribution numbers from the remaining traffic data. These approaches involve leveraging the data you do have, applying statistical methods, and making certain assumptions to estimate the behavior of the untracked users.
1. Modeling and Extrapolation
Use the Tracked Data to Model Untracked Behavior: You can develop models based on the behavior of the 25% of mobile app users and 50% of website visitors who have consented to tracking. These models can then be used to estimate the behavior of the untracked users.
Look-Alike Modeling: Identify patterns in the tracked user data (e.g., demographics, source of traffic, device types) and use these to create "look-alike" segments among the untracked users. This allows you to infer likely behaviors and attributions for the untracked segments based on their similarities to the tracked ones.
Scaling and Weighting: After identifying patterns in the tracked user data, you can scale these patterns to the entire user base. For example, if 60% of tracked users from a specific campaign convert, you might assume that a similar percentage of untracked users from the same campaign would convert, adjusting for any known biases.
2. Probabilistic Attribution
Statistical Attribution Models: Probabilistic models can be used to estimate the likely distribution of conversions across different channels. This method doesn't rely on direct user tracking but rather uses statistical methods to infer the contributions of various channels.
Markov Chains: This is a type of probabilistic model that looks at the sequence of user interactions before conversion. Even with partial data, you can estimate the probability of different channels leading to conversions by analyzing the paths of tracked users.
Bayesian Attribution Models: These models can incorporate prior knowledge and assumptions about how channels perform, updating the estimates as new data (even if limited) comes in.
3. First-Party Data and Offline Integration
Leverage First-Party Data: Even if some users opt out of tracking, you can still use first-party data (e.g., CRM data, purchase history) to link online and offline behavior. This is particularly useful if you have other touchpoints with customers (e.g., email campaigns, in-store visits) where consent might be different or where you collect different forms of user data.
Customer Lifetime Value (CLTV) Modeling: For known customers (e.g., those who log in to your website or app), you can use their past behavior to predict future actions and attribute their conversions back to specific channels.
Cross-Device Tracking: If users log in on multiple devices, you can track them across devices and platforms even if they decline tracking on some of them.
4. Aggregated and Anonymized Data
Utilize Aggregated Data: Even when detailed tracking is not allowed, you can often still access aggregated data that shows trends across large groups of users. This data can provide insights into the overall performance of channels without needing to track individual users.
Google’s Consent Mode: If you use Google’s Consent Mode, it can provide insights based on non-identifying signals. While this data is less granular, it can still indicate trends in traffic sources and conversions.
Platform Analytics: Some platforms (e.g., Facebook, Google Ads) provide aggregated performance data that can help estimate channel contributions without tracking individuals.
5. Use of Consent Management Insights
Analyze Consent Patterns: Understanding who consents to tracking and who doesn’t can also provide insights. For instance, if certain demographics or channels have higher opt-in rates, you can adjust your attribution models to account for these differences.
Adjust Marketing Strategies: By analyzing the consent data, you can potentially optimize your marketing strategies. For example, if a particular audience is more likely to consent, you might allocate more budget to channels targeting that audience.
6. Incrementality Testing
Run Incrementality Tests: Incrementality testing involves running controlled experiments where you isolate the effect of a particular channel or campaign. By comparing results from test and control groups, you can estimate the incremental impact of your marketing efforts.
Holdout Groups: Create groups that are deliberately excluded from certain marketing activities to measure the difference in conversion rates compared to groups exposed to the marketing. This can help estimate the true value of a channel even when you can’t track every user.
7. Use of External Data and Benchmarks
Industry Benchmarks: Compare your tracked data against industry benchmarks to estimate the performance of untracked channels. This is particularly useful if your data is incomplete but you have access to reliable industry standards.
Third-Party Data: Consider using anonymized third-party data providers that can help fill in gaps or validate your models, giving you more confidence in your attribution estimates.
While having a large percentage of untracked users presents challenges, the techniques above allow you to still derive meaningful channel attribution insights:
Modeling and extrapolation provide a way to infer behavior for untracked users.
Probabilistic and statistical models help estimate contributions from different channels.
First-party data and cross-device tracking can supplement the gaps left by untracked users.
Aggregated and anonymized data offer a broader view of trends and patterns.
Incrementality testing provides concrete evidence of channel effectiveness.
By combining these strategies, you can create a more accurate picture of your marketing performance, even in a landscape where user tracking is increasingly limited by privacy concerns.
So, IS MULTI-TOUCH MARKETING ATTRIBUTION STILL FEASIBLE?
Multi-touch cross-device marketing attribution remains feasible and meaningful today, but it is increasingly complex and requires a thoughtful, adaptive approach to yield reliable insights. The challenges introduced by privacy regulations, browser restrictions, and user consent dynamics have certainly made it more difficult to track and attribute user behaviour across multiple devices and touchpoints. However, with the right strategies and tools, businesses can still derive significant value from multi-touch attribution.
Why Multi-Touch Cross-Device Attribution is Still Feasible and Meaningful
Continued Relevance of Understanding Customer Journeys:
Holistic View: Despite the challenges, understanding the complete customer journey across devices remains crucial for optimising marketing spend, improving user experiences, and driving conversions. Multi-touch attribution allows businesses to recognize the impact of various channels and touchpoints, helping them make informed decisions about where to allocate resources.
Strategic Insights: Even with partial data, multi-touch attribution can provide valuable insights into which combinations of channels and devices are most effective in driving conversions, allowing for more strategic planning and execution of marketing campaigns.
Adaptation Through Advanced Techniques:
Advanced Modeling: Probabilistic models, data-driven attribution, and hybrid approaches (combining deterministic and probabilistic data) enable businesses to continue attributing value to different touchpoints even when full cross-device tracking is not possible. These techniques allow for meaningful attribution by inferring patterns from the available data.
First-Party Data Utilization: Encouraging user logins and leveraging first-party data helps maintain cross-device tracking capabilities. Businesses that can link interactions across devices through user authentication can achieve accurate attribution, even in a privacy-conscious environment.
Privacy-Compliant Attribution:
Respecting Privacy: The shift towards privacy-first approaches does not negate the value of multi-touch attribution; it simply requires more sophisticated, compliant methods. Using tools like Google’s Consent Mode, server-side tracking, and anonymized data aggregation, businesses can continue to perform attribution without compromising user trust or violating regulations.
Evolving Technologies: The industry is continuously evolving, with new tools and methodologies emerging that balance the need for privacy with the demand for effective attribution. Solutions like Customer Data Platforms (CDPs) and enhanced attribution models are designed to work within the constraints of modern privacy standards.
Meaningful Aggregated Insights:
Trend Analysis: Even when individual-level tracking is incomplete, analyzing aggregated data across user segments, devices, and channels can provide meaningful insights. These aggregated insights can inform marketing strategies and help businesses understand broader trends in user behavior.
Incrementality and Experimentation: Conducting controlled experiments and incrementality testing allows businesses to measure the impact of different marketing activities across devices. This approach provides a robust alternative to traditional attribution models, offering a clear view of the incremental value contributed by various touchpoints.
Challenges to Consider
While multi-touch cross-device attribution remains feasible, it’s important to acknowledge the challenges:
Data Gaps: The inability to track every user across every device creates data gaps that can lead to less precise attribution models. These gaps must be accounted for in analysis and decision-making.
Complexity and Resources: Implementing and maintaining advanced attribution models and first-party data strategies require significant investment in technology, expertise, and ongoing management.
Privacy Compliance: Ensuring compliance with evolving privacy regulations requires continuous attention and adaptation of tracking and attribution practices.
Conclusion
In summary, multi-touch cross-device marketing attribution is still both feasible and meaningful in today's digital landscape, but it requires adapting to new realities:
Leverage advanced models and first-party data to bridge the gaps left by reduced tracking capabilities.
Utilize privacy-compliant tools and methods to maintain user trust and regulatory compliance.
Focus on aggregated insights and experimentation to derive value from available data.
By embracing these strategies, businesses can continue to benefit from multi-touch attribution, gaining insights that drive better marketing outcomes and more personalized user experiences. The key is to stay agile, continuously adapting to the changing environment while leveraging the most effective tools and techniques available.
Interested? Find Out More!
Rittman Analytics is a boutique data analytics consultancy that works with growth-stage, mid-market and enterprise businesses in the UK, EU and North America to help level-up their marketing analytics capabilities and build world-class web performance and marketing attribution models.
If you’re looking for some help and assistance with your marketing attribution, marketing analytics or would just like to talk shop and share ideas and thoughts on what’s going on in your organisation and the wider data analytics world, contact us now to organise a 100%-free, no-obligation call — we’d love to hear from you!