Schema and Standard Reports

• Adobe Analytics organizes data using a standardized data schema with a hierarchy of Visitor > Visit > Hits.

• Hits can be subdivided into two main categories: Page Views and Non-Page Views (Custom Links, Downloads, and Exit links).

• Data is organized using two major blocks: metrics and dimensions. A variety of built-in dimensions/metrics are available out of the box.

• The Adobe schema allows for data to be cross-tabulated across metrics and dimensions from different schema levels. (e.g. it is possible to freely combine a hit-based dimension with a visitor or a visit-level metric and vice versa a visitor-level dimension with a hit or a session-based metric). 

• Classic Google Analytics organizes data in a similar hierarchy: User > Session > Hit.  

• Hits can be also divided in two categories: Page Views and Events, with a special consideration for events that can be further classified into interaction vs. non-interaction events. The same non-interaction flag can be also applied to Page Views

• The Classic Google Analytics schema has limitations in allowing combinations of metrics and dimensions from different scopes/levels. 

• Built-in GA reports guard against such combinations by disallowing them, while custom reports allow them but require advanced understanding of the scope of different dimensions/metrics for proper data interpretation.

Data Organization

• Adobe uses the concept of report suites to group/organize the data collected from different website(s)/app(s).  

• Virtual report suites allow segments to be applied to the data inside a standard report suite; coupled with granular user access permissions this option makes it easy to curate views of the data that are relevant to a particular business need/unit.

• Virtual report suites are a good alternative to the concept of "multi-suite" tagging providing a good way to organize data from different digital properties into a single standard report suite entity while maintaining the capability to curate different views of the data. 

• Adobe offers a "roll-up" report suite type which can be used to aggregate high-level data from other report suites. Roll-up report suites do not deduplicate visitor/sessions from different report suites, nor is there a way to segment/break out data contained in different custom dimensions. 

• Google Analytics organizes data using properties.  

• Within properties different views can be configured with options for filtering data, applying different reporting configurations, and enabling different access permissions. 

• GA360 accounts have a special category of roll-up properties that allow data from multiple properties to be integrated together while maintaining sessionization/visitorization and  allowing mapping of dimensions/metrics from related properties.

• Data from mobile apps can be sent to the standard "Web" family of properties,  to specialized "App" properties, or to the new style of "Web + App" properties.

• Web + App properties allow different feeds to be configured representing ways to sub-categorize data 

Custom Metrics

• Advanced support for different types custom metrics (counters, numeric, revenue) to capture the success events of digital properties. 

• Built-in capabilities to configure advanced deduplication logic using event serialization. 

• Coupled with advanced segmentation capabilities, metrics can be configured to deduplicate activity across custom visit and visitor scopes. Built-in configurability for event participation settings allowing more advanced analyses. 

• Google Analytics offers the creation of custom metrics with two scopes "Hit" and "Product". 

• While not part of the "Custom Metrics" setup, through the creation of "Goals", Google Analytics makes it possible to capture sessions where a particular success event was accomplished. 

• Goals are natively available for inspection in most standard GA reports, while custom metrics are only available within custom reports. 

Custom Dimensions

• Robust support for different types of dimensions (conversion and traffic variables) allowing to assign meta data at different levels in the schema to meet even the most custom requirements. 

• While ongoing developments in Analysis Workspace have diminished the value of Traffic Variables (props), Conversion Variables (eVars) offer advanced configurability for setting expiration and allocation capabilities.

• Dimensions with multiple values are supported through the use of the reserved s.products variable as well as through list vars. Usually only three list vars are available per report suite.

• List vars do not support full correlation of multiple values in a dimension with multiple values in a corresponding metric. 

• Google Analytics offers a single class of dimensions that can be configured with four different scopes (Hit, Session, User, Product).  

• Google Analytics dimensions capture the last (not most recent) value for Session and User-level dimensions. However, the last allocation is changed to first if the dimension is set via advanced filters.

• Custom Dimensions are available as secondary dimensions in most built-in reports and also available in custom reports. 

• Reserved product-scoped dimensions allow multiple values to be passed into a particular dimension on the same tracking HTTP request. 

• Extensive support with out-of-the-box dimensions and metrics to capture key eCommerce interactions. 

• Specialized dimension settings allowing expiration of credit allocation at the time of a custom eCommerce event (i.e. Add to Cart or Order).

• Merchandising eVars enable advanced analysis techniques, especially in the context of eCommerce-related attribution. The most common example is with attribution of purchases to a particular placement where a product was added to the shopping cart. Multiple blog posts by Adam Greco describe the process and logic behind the use of merchandising vars. 

• Through the introduction of Enhanced eCommerce Google Analytics has created a robust framework for tracking eCommerce interactions similar to capability in what Adobe Analytics offers. The framework comes with a fully instrumented demo site.

• One interesting aspect is that eCommerce tracking has its own scope sitting outside of metrics/dimensions with a scope of Hit and Session.

High-cardinality Dimensions

• Dimensions with a high number of unique values can show a special item labeled "(Low-Traffic)" that groups long-tail dimension values.

• This limitation usually begins to occur after a dimension has reached 500,000 unique items during a given month, but the different limits can be reconfigured. 

• High cardinality tables can also sometimes lead to hash collisions that can make the interpretation of data quite challenging.

• Adobe Data Warehouse and Data Feeds are not affected by cardinality limitations. 

• Google Analytics are also not immune to limits associated with high cardinality. Items beyond the given limit are grouped in a bucket called "other".

• Free and GA360 accounts have different cardinality limits that depend on the type of table that is queried and the selected date range, among other factors.

• Custom tables and queries of raw data in BigQuery are two possible ways to bypass issues with cardinality limits. 

• Adobe Analytics does not perform any sampling and processes all available data. 

• Google Analytics has different sampling thresholds for Analytics Standard and Analytics 360 accounts.

• Adobe offers a built-in dimension slot for tracking campaigns. The slot can be configured (like any other eVar) to fully customize its expiration and allocation settings. 

• Adobe's standard relies on a single campaign query parameter. There is built-in support for "cid", but through javascript modifications or processing rules any parameter(s) can be used.

• The parameter value can be enriched to include other campaign meta data through the use of classification techniques.

• Google Analytics's campaign tracking standard involves the use of five campaign parameters (utm_source, utm_medium, utm_campaign, utm_term, utm_content) and five built-in dimensions that use a "most recent" allocation logic.

• If other parameters are used they can be remapped to the built-in reports/dimensions through the use of advanced view filters. 

• If a session has more than one campaign, Google Analytics will increment a Session count for each campaign instance. 

• Using a waterfall rule-set, Adobe can organize different traffic sources into up to 25 marketing channels.

• Information about specific campaigns/drivers within a marketing channels can be extracted through configuring a separate dimension called "Marketing Channel Detail".

• The built-in configuration generates two dimensions "Last Touch Marketing Channel" (more accurately described as "Most Recent") and a "First Touch Marketing Channel".

• Using the specialized Attribution panel type in Analysis Workspace the default Last and First methods can be changed on the fly to include a variety of other attribution models (Linear, U-shaped, J Curve, Time Decay, etc)

• Advanced settings include an "override" toggle which can determine if a given channel can override other channels, as well as an expiration window. 

• Google Analytics has a similar implementation with Channel Groupings. 

• Multiple Channel Groupings can be configured to build different channel models.

• The Default Channel Grouping integrates most easily with built-in reports. 

• Channel Grouping configurations are specific to each GA view making it highly configurable, but also difficult to deploy and manage across an enterprise setting with multiple properties/views. 

• The primary method for enriching dimension data is through two classification techniques: lookup-based classification (file imports) and rule-based classification (rules that can extract and match patterns/RegEx and return hardcoded values).

• Another method for data enrichment is the use of DB Vista rules.

• Dimension enrichment can be achieved through the Data Imports feature. 

• Using GA Advanced Filters - users can also apply rule-based classification similar to the functionality Adobe offers.

• Unlike Adobe, data fields added through the lookup require their own dimensions.

• Each report suite can be configured with its own time zone, currency, and custom calendar setting.

• Time zone and currency support down to the view level.

• Visitor identification (visitorization) relies on the cookies set through the standard JavaScript tracking libraries. 

• Google Analytics's primary visitorization method also relies on the cookies set via JavaScript. 

• It is also possible to modify the visitorization methodology through specialized views and a custom identifier (user id).

• Bot definitions maintained by the IAB can be enabled on the level of the report suites to filter out known bots.

• Traffic identified as coming from bots does not count towards various metrics/dimensions but is available for inspection in two built-in reports - Bots and Bot Pages.

• IP and User Agent based filters can be custom-built to filter additional traffic. 

• Google Analytics uses a similar process to exclude bot traffic from reports. 

• Custom filters based on IP and other fields can be also built. 

• Processing Rules allow a good deal of flexibility for processing data after it has been transmitted by the tag and before it enters the reports.

• VISTA Rules also allow data preprocessing but come with an additional cost and usually require custom engagements

• Google Analytics can use advanced filtering rules which can be used to execute data cleanup and meta data reassignment similar to the functions performed by Processing Rules in Adobe.

• Through IP lookups, Adobe generates automatically Geo dimensions for Country/State/City/DMA/MSA. 

• Google Analytics uses the same method to generate predefined Geo dimensions. 

• Adobe offers two separate options for masking the visitor's IP (last octet replacement with 0 and IP obfuscation).  

• These settings have consequences on filters relying on IPs well as the the accuracy of Geo-based dimensions.

• Google Analytics provides a special flag "anonymizeip=true" that can be set to remove the last octet of the visitor's IP from the data written to GA.

• Real time data is available within a few seconds. Different dimensions/metrics can be configured for real time data review. eVars are generally not suitable for real time exploration.

• Current data is usually available within 30 minutes after the data has been generated. 

• Fully processed data is usually available an hour or so after the data has been generated. 

• Real time data is available within a few seconds of generation. Real time data is limited to a selection of built-in dimensions/metrics.

• For smaller data sets, fully processed data is usually available within 30 minutes after the data has been generated. For larger data volumes, it may take several hours before fully processed data is available.

• The default data retention period is 25 months but Adobe also offers purchases of extensions.

• The data retention terms depend on the type of GA property as well as on the type of data.

• Settings can range from 2 months to no expiration date for some types of data.

• Adobe provides ways to bulk-manage the configuration of reporting settings across multiple report suites directly in the UI.

• Most reporting settings can be copied from existing report suites to new report suites.

• Selecting the metrics/dimensions across multiple report suites also provides a handy diff of the configurations across different report suites.  

• Google Analytics provides a management API that comes with daily utilization limits.  

• Bulk management through the UI is not available.

• Data organization/processing considerations for Google Analytics App + Web properties are somewhat different from Classic Google Analytics with a focus on an open-ended parameter name/value pair schema. The beta version of the product offers a limited number of parameters that are available directly in the App+ Web UI, but promises a more streamlined and modern approach to data organization. 

• On Adobe's side, legacy products such as Adobe Data Workbench (from the Visual Sciences acquisition) have been able to offer a fully customizable schema for clickstream data and joining of other data sources for over a decade. Adobe's latest developments in this direction include the introduction of the Alloy.js tracking library and the customer journey analytics offering.

  About QA2L

QA2L is a data governance platform specializing in the automated validation of tracking tags/pixels. We focus on making it easy to automate even the most complicated user journeys / flows and to QA all your KPIs in a robust set of tests that's a breeze to maintain. We take a smart, targeted approach to issue discovery via auto-selective site audits and scans and can help your organization on the path of improved data governance and data quality.

Contact Us