Having more than one Google Analytics property will allow you to divide your data into different categories. There are numerous benefits to this, and you will likely have your own reasons for wanting to do so.
Whereas it is still possible to double tag your website via code, this has never been a recommended practice, as it can lead to inaccurate data collection, processing, and reporting. It is also more difficult to maintain, as you will have separate code to block each property. A better way to set up multiple properties is with Google Tag Manager.
Setting Up a Second Property
Create a second property
In your Google Analytics account, head to “Property” under “Admin.” Click the drop-down menu and choose “Create new property.”
Create a Constant String Variable
Under “Variables,” choose “User Defined” and then “New.” In the “Choose Type” field, type “Constant.” Give your variable a relevant name. Add the tracking ID to the “Configure Variable” box and hit “Save.”
Add the GA Pageview tag
Under “Configure Tag,” delete the current tracking ID. Click the icon at the side of the box and choose the variable you created earlier. Save your changes.
Preview the tag
Check both the original and new Pageview tags are working correctly. If there are no problems, go ahead and publish.
Create more tags and properties
Repeat the steps above for all your current tags to send all the relevant data to the new property. You can also use the instructions as many times as you want to create multiple properties.
Benefits of Multiple Properties
Complete and sampled data
Large sites (those receiving more than 500,000 visits per month) that rely on just a single property only receive sampled data for their reports. When you create multiple properties, on the other hand, you can receive complete data.
Although sampled data is often sufficient, it can lead to inaccuracies. This is especially true if you want to look at specific cases, you need exact numbers, or you want to deal with a particular tracking issue. The problem worsens if the sample rate is less than about 75 percent of your data. Even creating subsets of data using views will only help standard reports — they are of no use to advanced segments and customized reports.
Data collection limitations
Google states that it has a limit of 10 million hits per month. If you have more than this for a single property, Google may not process the data.
Data refresh delays
If you receive a little over 200,000 visits per day, multiple properties will help you receive your data sooner. When visits exceed this number, Google only processes the data only once per day, leading to a delay in refreshes. An even larger number of visits can mean that custom variable values in standard reports see no update for up to three days.
Separate web properties for different personas
Most websites are split into several entities, such as a main site, blog, support center, and landing pages. There is no need to join these under a single property, as users visiting one section are likely unrelated to those visiting another. For instance, if you have a blog with a relatively high readership, you can expect that many of these visitors will never become customers nor visit another area of your site.
By dividing your site into different properties, you can measure the activities of your different audiences. This will give you a more accurate depiction of data for metrics like bounce rate and conversion rate. For the best results, split your site into personas, such as blog readers, potential clients, and registered clients, free and paying customers, and traffic from online marketing campaigns and organic traffic.
From blog readers, you would expect a high bounce rate and number of unique visitors and visits but a low visit duration and conversion rate. From paying customers, however, you would expect the exact opposite. By separating properties, you can check the analytics of each.
It is also possible to divide personas using advanced segmentation or separated views. However, these are poor alternatives to multiple properties.
Advanced segmentation, for instance, is ineffective for the long term. You will need to choose the segment again each time you look at the data and there is no way to only look at SEO traffic. Plus, you will only be able to use sampled data — and we have already established why that is limited.
As for separated views, these are difficult to set up for cross domain tracking, especially if you have a large site. Worse still, there is no way to reverse changes in the filter and data may be inaccurate. Plus, there is the issue that separated views only support standard reports.
If you want accurate data that you can utilize to improve site performance, multiple properties are ideal. Whereas you may be able to function with a single property if you have a small site, any large or complex site needs multiple properties.