Let's start with a fact: you need an analytics stack that's customised, fit for purpose, and grows with you. That's the definition of a perfect stack.
If you have data or want data (or contemplate that you will), how this data is captured, stored, organised, analysed, displayed, and distributed should matter to you. Like, a lot. Getting your analytics stack wrong is expensive. It costs time and money to back-engineer.
Here's what else can happen when you get it wrong.
1. You may need to rebuild. Rebuilding in part or the whole stack is time consuming and expensive. It uses not only the time of your dev team (who need to implement) but your strategy team who need to understand what needs to be fixed:
2. You'll build data into data silos. Third party analytics platforms come off-the-shelf, which is super convenient, but they're also not built to integrate. If you get the stack wrong you'll end up with multiple data sources and duplicate data.
3. You'll lose data to leakage. Without the right set up, you'll be losing data points every day that your stack isn't correctly configured. Unfortunately, that's irrecoverable data loss.
4. You won't be able to use the data you have. I've seen this happen - the data is there, but because the platform isn't fit for your purpose, you can't get it in a form you need, which means you need to hire someone to get it out for you. Very frustrating.
The problem, of course, is that we're spoiled for choice. Data is important and interesting so there are literally thousands of tools to choose from that capture, store, organise, analyse, display, and distribute data. With slick websites and free trials, it's easy to be wooed and hard to differentiate.
So how do you know what tool is right for you? Here are four things you need to do to get it right the first time.
Articulate your business objectives in data terms.
Just as in life, so in data, you need to know what you want. Every platform is different in what it tracks and how it tracks it. Consider what you want the data to illuminate for you.
Are you a B2B business that needs to automate some leads research? Are you interested in how visitors use your website or your product? Are you mostly concerned with decreasing cart abandonment or remonetizing existing customers? There are tools to help with all of these.
Also, articulate what you want to do with the data. For example, if you want to use an analytics platform to send out targeted communications, make sure you implement a tool that can track your funnel across email, web, and app.
Finally, avoid data paralysis by resisting the urge to capture everything, but consider how your objectives may evolve.
Do your research.
There are so many options. Which platform is right for you? Now that you have articulated your business objectives in data terms, this step becomes that much easier. Make sure that the platforms you are looking at serve your needs.
Particularly if your data use cases are creative or out-of-the-ordinary, make sure the platforms you are looking at are able to serve you information in a usable form.
For example, if you have decided that Mixpanel is the platform for you, make sure all of the triggers that you would use to send targeted emails, push notifications, or SMS messages are trackable and coded in as attributes rather than events. This means that they persist with the user profile rather than ceasing to be a targetable descriptor after 90 days.
Consider how you'll integrate your data sources.
If you're building your ideal stack, the data sources should be friendly with one another. Insofar as possible, consider how the various platforms integrate. Some tools and platforms play better together than others.
Is your heat maps provider able to send data to your web analytics platform? Can you use your A/B testing platform to personalise your homepage for users who have used your app in a certain way?
Data is better when it's centralised. Synchronicity is not always possible, but choosing tools that are able to talk to one another wherever possible will amplify the benefit you derive from your data.
Get external advice.
There are a lot of options out there and getting your analytics stack right takes time and headspace. If you have doubts, or just can't invest the time you need to to get it right, invest in an expert. Getting the right advice early on will save you a lot of time and money down the track.
Alex is the Head of Growth at DataMuse, a data and growth consultancy that specialises in monetising data and growing data ROI by mining data for insights with impact, developing and implementing robust data strategies, and centralising data silos.