Why do I need an analytics strategy?
Defining your data strategy in advance lets you make sure data is collected consistently and distributed to where you need it. If you've ever seen Google Tag Manager (GTM) set-up that's grown organically, you'll realise the importance of this!
Not having to ask yourself which platform most accurately actually represents the 'truth' lets you focus on what matters: Building a great product.
With the right tooling you'll get a better insight into customer behaviour on- and offsite, are able to identify what product features work and can appropriately attribute traffic to channels and campaigns.
By sending the right data to marketing integrations like Facebook or Google Ads, you'll improve the quality of your conversion measurements and improve your retargeting ability, beyond just 'You visited our website, look at our logo in your newsfeed'
Plus: You'll probably be able to cut down on custom scripts and code improving your site performance.
What should I do?
Define a tracking plan
First off, you'll want to make sure you set-up a tracking plan. We'll have a more detailed blog post on this in the future. In a nutshell, a tracking plan is the place where you document what you're tracking and why. It ensures you stay on-track and gets all teams on the same page on what gets implemented and how. Your tracking plan should consist of:
- A standardised naming scheme for event names and properties.
- The conditions under which an event triggers.
- The purpose of tracking this event.
- The properties submitted alongside the event and the values they can take.
- A code example of how the event should be implemented.
- Which destinations receive the event data and how it is represented within the destination (e.g. Facebook, Google Analytics)
Generally, you'll want to keep event names as generic as possible and differentiate by using properties. A tracking plan takes some work to set-up, but forces you to think about your data in advance—this is useful whether you're using Segment, GTM or a completely different solution. You can read more about how to create a tracking plan for your business in our detailed blog post!
Set-up your data infrastructure
Next up, you'll want to implement your tracking plan and start routing data to the right destination. My go-to solution for this is Segment (US$120 p.m. / 10,000 unique users), which has a number of advantages over tag management solutions like GTM:
- Track web, mobile and server-side data consistently through the same platform
- Get access to the raw tracking data (Segment syncs to a data warehouse of your choice)
- Adding a new destination to receive all the events from your tracking plan is a single click
GTM is still great if you have very basic tracking needs or really don't want to spend the money—but if there is one SaaS tool I'd invest in, it's Segment. If you stick with GTM, make sure you use data layer events/properties, have good documentation and clear responsibility for who is maintaining the setup.
Note: Most start-up spaces have $10k worth of free Segment credit for you to use.
Set-up event and product analytics
If you're using Segment, this is easy—all the events implemented via Segment will already be available in a platform of your choice. My personal preferences are Amplitude (10m events free p.m.). Amplitude not only gives you a great interface for analysing your event data to surface conversion drivers but also lets you sync the behaviour based audiences you're creating within Amplitude to a number of destinations.
Alternatively, heap.io is also a great choice, especially if you want to reduce the effort needed to implement the events from your tracking plan. Heap automatically tracks everything that's happening on your site and then lets you associated events with those actions retroactively. I believe this is too brittle and still doesn't solve the need to push event properties manually (luckily Heap also accepts Segment events), but can be a great starting point.
Set-up qualitative and behavioural analytics
Finally, you'll want to set-up a solution to capture qualitative data like heat maps and session recordings. These are invaluable to understand how visitors are engaging with your site and can provide insights you wouldn't be able to derive from just looking at the numbers.
Session recordings are especially useful to review for one segment at a time, e.g. 'all users abandoning the cart', 'all users filling in the payment form, but not purchasing'.
A great compromise between cost and features is Lucky Orange. I particularly like that Lucky Orange fully integrates with Segment, which means you're able to filter records based on user traits or events they triggered. Have a user that called up customer service about an error? Review the recording to figure out what happened.
Alternatively, HotJar's free plan is a great starting point if you don't want to spend any money yet. However, HotJar is struggling with overlays (e.g. chat pop-ups and navigation windows) which means you'll need to keep any possibly interface changes in mind as you look at your heat maps.
What about Google Analytics?
Google Analytics is a great tool because it is so widely used. However, it very much comes from a Web 1.0 era, which makes it cumbersome to use for event-tracking and anything that doesn't resemble your classical e-commerce store. Integrating Google Analytics is very straightforward, so you may as well do it—but don't rely solely on Google Analytics for your tracking needs.