Google Analytics for SaaS Products

Google Analytics has and is highly adopted in the analytics world, but when it comes to SaaS, there are a few things to keep in mind and adjust to get the full benefits available from Google Analytics.

Being a SaaS startup or existing product, you will have to look at Google Analytics in some things from a different perspective than how a regular website or ecommerce store, such as a SaaS need to account for users and recurring revenue instead of only lead form fills or ecommerce purchase transactions.

Although Google Analytics was not built for tracking individual users for products, with some changes in its’ implementation, it can make that possible to do to help with optimizing your marketing and product analytics initiatives. Other analytics tools like Amplitude, Heap, and Mixpanel are great alternatives for tracking how your users are interacting with your product but come at a high cost for small SaaS startups after a certain usage.

Here, we’ll assume that you have successfully installed Google Analytics and Google Tag Manager on your website and application and go over the main things to do for a complete Google Analytics implementation for a SaaS product.

Goals and Funnels

As a SaaS product and business, your primary goal is subscriptions, so some important goals would probably be a purchase and secondary goal might be a registration or a lead subscription. Keep in mind Google Analytics has a limit of 20 goals for the free version. See some examples that would make sense for a SaaS company to track:

• Sign In
• Sign Up
• Paid Subscription
• Cancel Subscription
• Change Plan
• Invite User
• Create a Report/Invoice/Campaign/etc.
• Upload Image/Document/File
• View Report
• Download a whitepaper
• Lead form fill

So, create a goal and/or create a funnel for that goal so you can see the conversion rate and in the funnel the drop-off or abandonment rate as well. This will all help with being able to see the issues, A/B test, improve your funnels, and in the end increase user sign-ups.

Ecommerce Tracking and Recurring Revenue

I’ll start with saying that Google Analytics isn't exactly well-suited analytics solution for SaaS businesses and let’s see why and some workarounds to make it work.

For a general process of events SaaS companies must account for and track:

1. When a visitor signs up for an account (registration)
2. When a user adds a paid subscription and provides their payment information (free/trial to paid subscription)
3. When the payment transaction is processed via Stripe/Chargebee/PayPal/etc. (purchase)

Although these are all possible to track and report back to Google Analytics as event goals, but the difficult part is tracking what comes after with the payments processed and that the events listed above usually happen over a span of multiple days/sessions. Because Google Analytics tracks everything by sessions, these events would have separate sessions and thus not track everything together correctly. This would be especially helpful to track the LTV and CAC from your marketing campaigns, channels, etc.

A workaround we've found has been to use User ID’s as a custom variable for each user subscription and get the data via the Google Analytics API. Then we can connect the GA data and the application database using the User ID as the mapping key and be able to connect the data and report correctly. Finally, to track recurring revenue, Google Analytics Measurement Protocol would be used to send payment data back to Google Analytics from the application transaction data.

To reiterate, Google Analytics isn't best suited for tracking recurring revenue, and although a great ecommerce analytics tool, when it comes to recurring revenue it gets difficult and with workarounds to try to make it work.

Google Analytics Views

Since you’ll be receiving traffic from multiple sources and different types of visitors, seeing all the mix traffic in one place might mess with the data, so it’s best to segment the data so later we can analyze the data for each segment separately.

This will allow you to see and compare how your users are interacting with your application, compared to how visitors of your website are behaving landing on the website.

You should have the following three types of views configured:

• App View (with User ID) – For https://app.website.com (see Cross Device Tracking below)
• Website View – For https://www.website.com
• All Website Data (default) – To track all sessions and users (unfiltered and raw data)

Custom Dimensions to Track

Using Custom Dimensions, you’ll be able to define a segment of users and within the Cohort Analysis, help in filtering out the specific segment to analyze their retention rates, activity, source, and much more.

As the name implies, custom dimensions will be custom and vary for every organization, so some strategic planning must go into what dimensions will help marketing and user initiatives for growth.

Some examples of custom dimensions a SaaS product might use:

• Subscription Plan Type
• Member Since
• User’s Company Vertical/Industry
• User Occupation
• User’s Company Size
• User ID

AARRR Model (Pirate Metrics)

Like many other SaaS products, you want to be able to attract and retain new users and tracking everything around them is especially important. The AARRR metrics (also known as Pirate Metrics) are:

1. Acquisition: Where visitors come from through various channels
2. Activation: Visitors that convert to registered paying users
3. Retention: The users come back and stay using your product
4. Referral: Users that refer your SaaS to other people
5. Revenue: Users that pay to use your SaaS product

Google Analytics is a great analytics platform when implemented correctly to help you answer questions involved in the AARRR model to find meaningful insights from your results.

Cross Device Tracking with User ID

From the start, Google Analytics by default uses ID’s to track every visitor on the website, and this is called a Client ID. The thing to keep in mind here is that Client ID’s give a separate Client ID to each platform, for example opening the same website with four separate browsers would record as four separate Client ID’s and users.

User ID tracking is set up using Universal Analytics and Google Tag Manager and you user authentication system to expose the User ID and be able to be sent to Google Analytics. This User ID could be the User ID identifier for the user, but should never be any PII such as email, username, etc.

A great benefit of setting up the User ID is you’ll see and understand the relationship between multiple devices and sessions. Note that your user metrics will decrease which is because every device/browser is treated as a separate user. By using the User ID feature, several devices will merge into one actual user.

EU Visitors and the GDPR

Even if you’re not operating out of the European Union, you may have heard of GDPR and not be fully aware of what it all means. To get the full explanation, take a look here.

In short, if you have traffic from the EU and you want to setup or already have setup cross-domain User ID tracking (as discussed above), then you are not GDPR-compliant and by law require consent from each user.

Here is a basic overview of the steps that usually happen to be compliant with GDPR:

1. A person visits the landing page
2. A banner displays and asks for consent to track user
3. A person can agree or disagree to have them tracked and their data processed by you
4. The consent can then be stored in a cookie

For more detailed information see Google’s policy and the implementation of this can vary greatly by application and development team.

Closing Thoughts

If you’ve followed all of what was above, you’ll be getting some of the same tracking as large scale and advance product analytics implementations that large companies have for free (with data sampling though).

This is only the first step to having a full view of your analytics, where Google Analytics is great for web and marketing analytics, you’ll also need to go a step further and implement product analytics with platforms like Amplitude, Heap, Mixpanel and many others.