Universal Analytics: How to Make the Switch to it Without losing Data.
Now is the time to begin the migration to Universal Analytics from any existing Google Analytics set-up. The new code is not only the ongoing standard for Google, but it also gives you access to additional features. Universal Analytics officially left the Beta period in April and it's now compatible with all analytics features, including Remarketing, DoubleClick and more. It's time to migrate to Universal Analytics! We will give you some reasons to do so and advise on the migration process without losing any of your data.
Why migrate to Universal Analytics?
The main benefits of migrating to Universal Analytics include the ability to connect multiple devices, sessions and engagement data with a unique user ID, and the possibility to track any digital device with the three new data collection methods:
- Android and iOS devices use the Mobile SDK V2.x and higher
- Other devices, such as games consoles use the Measurement Protocol.
Universal Analytics also makes it easier to measure across multiple domains and sub-domains. Plus you'll be able to configure much more within your Google Analytics account rather than having to adjust your tracking code, including organizing Organic Search Sources, custom session and Campaign timeout length, Referral Exclusions (such as from a 3rd party shopping cart), and Search Term Exclusion (for instance brand name searches will appear as direct traffic if excluded from search terms). And you'll be able to use custom dimensions and metrics to analyze data not tracked ordinarily. For example, product details or authors of content pages.
If those aren't enough reasons, there's also the knowledge that although the older methods of data collection will continue to work in Google Analytics for some time, non-Universal Analytics data will be deprecated and no longer work in around two years. So it's best to get started with the new standard and ensure that everything is implemented correctly now.
Luckily the basic process for transferring is relatively simple. You'll need to transfer each individual property within Google Analytics and implement the new tracking code. It should take between 24-48 hours for Google Analytics to complete the transfer. It is also worth considering using Google Tag Manager to complete implementing the new tracking code and to avoid having to change code in the future. By starting the process now, you could potentially run old and new tracking in parallel to check for any problems.
Not Losing Your Data
Upgrading should still let you access and analyze all of your historical data using the same reports and tools you currently access, but you'll also get all the additional new features.
So you're unlikely to lose your raw data, but what may happen is that if you have customization, such as session and campaign timeouts in your classic Google Analytics tracking code and you don't configure these to match in Universal Analytics, everything will return to default and you can end up with corrupt or incomplete data. Before upgrading, check any customization made to the tracking code currently in place and make sure you know the correct way to ensure this is continued via Universal Analytics.
Create an implementation plan that includes all of the Google Analytics features you use, like ecommerce or event tracking. When you implement the Universal Analytics tracking code, you'll need to update the tracking code tags for features that require additional set up.
For eCommerce tracking, for example, you'll need to add an extra line of Java Script to your thank you page to load a special eCommerce tracking library, in addition to the standard page tag for every page of your site:
ga('require', 'ecommerce', 'ecommerce.js');
You'll then need to add transaction details:
id: '00001', //
affiliation: 'Dan's T-Shirts.', //
revenue: '10.00', // Grand Total
shipping: '0' , // Shipping cost
tax: '2.00 ' }); // Tax.
id: '00001', //
sku: 'WR10101', //
name: 'White cotton T-shirt, //
category: 'T-Shirts', //
price: '10.00', //
quantity: '1'}); //
To check your code upgrades, visit the Real-Time reports in Google Analytics and check hits are being sent from your site or app. If they're appearing in Real-Time but not in Standard Reports, then don't panic as the data is just not processed yet. If there's nothing reported as expected in Real-Time then you'll need to check your implementation and make sure it's correct.
Run Universal Analytics in Parallel
One very good option for implementing Universal Analytics would be to run it in parallel with your existing reporting to check your setup before upgrading your original account.
There are no issues with using Universal alongside regular analytics. If you set up a new property for this in Google Analytics, you could then copy all customization across when you upgrade your original property, and simply remove the test property.
Alternatively, you can create a new account in Google Tag Manager and create a new container for your website. Place the tracking code immediately after the opening tag throughout your site.
Set up a new Property in Google Analytics using Universal Analytics and select the UA ID. Create a new tag in Tag Manager for Universal Analytics and inset the UA-ID into the Tracking ID field. Now add the appropriate Firing Rule, Save, Create a Version and Publish.
When you're happy that everything is customized correctly and reporting properly, you can remove all old Analytics tracking code, saving the UA-ID. Once that's done, simply put the new UA-ID into your Tag Manager code, Save, Create a Version, Publish and then check that your legacy account is now reporting via the Universal Analytics code.
Regardless of whether you begin the migration to Universal Analytics immediately, it's important that you start preparing and learning how to replicate and improve any unique customization as soon as possible. Although Google has said that it will be at least two years before older code is deprecated, there are never any guarantees. It's best to be prepared, particularly if you want to test your Universal Analytics implementation before rolling it out as your main measurement tool.