Exploring New Features: Google Analytics App + Web In-depth

Recently, Google introduced a new App + Web feature that allows you to combine data from sites and mobile applications in one Google Analytics property. This new functionality should make life easier for companies that want to track and analyze user actions across different platforms.

For now, App + Web is in beta testing, and it will certainly be further developed and improved. However, we’ve already tested it, checked out its data structure, and tried the built-in export to Google BigQuery feature. We’re eager to share our first impressions with you.

In the standard Google Analytics, everything is built around sessions. App + Web departs from this and uses a data model based on events, parameters, and users. In principle, it uses the same data scheme as Google Analytics for Firebase and has the same functionality, only supplemented with new features.

On the one hand, this approach will make life easier for product analysts, since data from websites and applications will be collected in the same format. On the other hand, due to the new structure and data processing logic, existing reports will need to be redone.

Instead of the usual custom parameters at the hit, session, product, and user levels, the beta version of App + Web uses event parameters and user properties. At first glance, this appears to limit your capabilities. And in fact, only user- and hit-level options remain available. In addition, it’s not clear whether it will be necessary to develop and implement a separate metric system for existing customized tracking such as Custom Dimensions.

We’re thrilled to see Enhanced Measurement – a cool option that eliminates the need to add new tags to your site. In addition to events (which are automatically collected in App + Web), you can track page views and scrolling, clicks, site searches, video views, and file downloads. To configure their tracking, just activate the option in the resource settings.

If you’re using the SDK or gtag.js, you can use App + Web to collect some user properties without adding any additional code to your site. For example, you can collect demographic data, device brand, device model and type, user language, and user interests. These properties can be used as parameters for audiences and general filters in reports.

Another App + Web bonus is a custom sales funnel report (aka “Special Funnels”), which was previously available only to Google Analytics 360 users.

Moreover, in the new App + Web property, this report provides more opportunities. For example, instead of five steps, a maximum of ten are now available. You can set a time limit between steps, use optional steps, etc. Such flexible settings will help you determine the important steps for conversion and find out at what stage of the funnel users get lost without creating reports using third-party tools.

The familiar real-time reports are replaced in App + Web by the StreamView function. On the one hand, it provides greater detail: you can see more data about users who have visited your site or application in the past half hour. On the other hand, StreamView provides no data on traffic sources that are available in real time in Google Analytics.

The DebugView tool has also migrated from Firebase to App + Web, which greatly simplifies testing and introducing new functionality to the site.

Compared to the standard Google Analytics, App + Web has stricter limits on settings as well as on the number and names of events and parameters.

For example, in App + Web you can use:

• Up to 500 unique events (excluding automatic) (max 40 characters for an event name)
• Up to 15 conversion events (similar to goals in GA)
• Up to 25 parameters per event (max 100 characters for the value of a transmitted parameter)
• Up to 25 user properties
• Up to 50 audiences

Such restrictions can interfere with the operations of large and medium-sized companies with large sites that collect many additional parameters. For example, fifteen conversion events will not be enough for businesses with a large number of micro-conversions.

The App + Web interface is more like Firebase than standard Google Analytics, so you won’t find some reports that are important for marketers. For example, there are no reports on traffic sources, cost analysis, e-commerce, or site search, no grouping of advertising channels, etc.

It’s also still not possible to upload data on advertising costs and offline data into App + Web. And there’s no way to adjust the duration of the session, which by default is 30 minutes. Perhaps when App + Web comes out of beta testing, some of these features will appear.

So far, App + Web has no direct integration with Google BigQuery: data is uploaded via Firebase and this functionality is available only to customers with the Blaze plan. That is, the uploaded data may be unsampled, but you’ll have to pay extra for data processing depending on the volume.

As for the uploading time, data on events for the current day is loaded into the intermediate table every 15 minutes (stream export). A full data export occurs once a day to a separate table, after which the intermediate table is deleted.

At the moment, the upload structure for App + Web is identical to the upload structure for Firebase. In this regard, it’s likely that the data in the tables is processed according to the Firebase logic. Judging by the description of the fields, when data is uploaded, the source, channel, and campaign that first brought the user to the application (or website) are recorded in the BigQuery table. For attribution, the Last-Click (Multiple Channels) model is used in the traffic_source field.

Having conducted a test upload, we noticed that the source, medium, and campaign are also transferred with the event parameters, but not with all events (we haven’t found any official information on this yet, but we assume that they’re transmitted with pageview and/or conversion events). Which attribution model is used is not yet apparent.

Attribution nuances in Firebase:

• Attribution data is available only for conversion events.
• The (not set) mark in sources appears when there’s no data on the source, medium, campaign, ad network type, or creative.
• The conversion window in Firebase cannot be changed: it’s 30 days for install and 180 days for re-engagement (in-app conversions).

Not only the evaluation of advertising channels but also the implementation of the sales plan depends on which attribution model you use. Ideally, the attribution model should take into account the objective contribution of each channel in the chain before ordering, combine online and offline data, and have transparent calculation logic. OWOX BI Attribution does all of this.

With OWOX BI Attribution, you can evaluate the contribution of your campaigns to sales on your site and at your physical stores and efficiently distribute your advertising budget.

With App + Web, Google has changed the approach to building reports in Google Analytics. This new property uses a data model based on events and their parameters. Accordingly, the data format for uploading to Google BigQuery has changed.

In general, the functionality of App + Web is still crude. But the product is in beta, and the Google team is planning changes and improvements. Many reports, settings, and features familiar to users of standard Google Analytics are not yet available in this new property type.

So should you switch to App + Web now? It’s definitely not worth stopping your collection of data in the standard Google Analytics. It’s better to try both methods and compare their data models and reports.

Companies who want to analyze data for their application and website in a single interface will like App + Web. It lets you find out which advertising channels and which platforms best attract new users and bring more conversions.

Since data is uploaded to Google BigQuery from App + Web in a different, simpler structure, there are fewer nested fields for product analysts to worry about. Accordingly, it’s more convenient to work with SQL queries.

With this update, Google has simplified the process of collecting data and delivering it from Google Analytics to Google BigQuery for projects that are ready to significantly change the structure of their reports and their metrics system on the site

• Since App + Web does not yet have standard reports on traffic sources and cost analysis, it will be difficult for marketers to use it to evaluate the effectiveness of advertising campaigns – to find out how many sessions a particular source has brought, compare the ROAS of all channels, etc.
• Businesses who care about getting real-time, raw, unsampled data from Google BigQuery 
• and companies that transfer a lot of custom parameters from their site to Google Analytics and want to build Google BigQuery reports on these parameters will also not benefit.
• In addition, App + Web isn’t for projects that want to upload to BigQuery and use transaction data in reports or for companies that want to take into account each step of the user in the funnel and honestly evaluate advertising channels without giving all the value for conversions to the last channel before ordering.

However, not everything is so sad. You can still accomplish all of these things using OWOX BI or the BigQuery Export function, which is available in the paid version of Google Analytics. You can read more about the features, pros, and cons of these two methods in our article on how OWOX BI differs from Google Analytics 360.

If you need help setting up web analytics and creating an individual metric system for your business, write us at contact@owox.com or fill out the contact form on our site.