At the foundation of any data-informed organization is a solid analytics stack. Without the right solution in place, it doesn’t matter how much data you collect -- you’ll have no means to get useful insights from it.
How do you go about choosing the right analytics stack for your business? Unfortunately, the big data space is incredibly crowded, with everyone claiming to offer the best possible solution for your needs:
Before jumping straight into evaluating vendors, you should first take a step back and ask yourself a basic question: how do you want to deploy your analytics solution?
3 ways to deploy your analytics stack
There are three main ways to deploy your analytics stack:
- In-house: You have your own custom-built analytics stack for your business. It might be your gut instinct to build everything yourself in the beginning (especially if you’re an early-stage startup), but be sure not to ignore the other costs of building in-house. Do you have the time and engineering talent to build and support your own custom analytics stack? How will this affect product development?
- On-prem: You purchase an analytics solution and deploy it “on-premise” -- that is, on your own infrastructure.
- SaaS: You purchase a SaaS analytics solution, which means you send your data directly to your vendor of choice who then centrally hosts all of your data.
So how do you figure out which one is right for you?
The one thing your analytics needs to accomplish
We’ve talked a lot about the importance of a data-informed organization; the foundation for building a data-informed organization starts with your choice of analytics.
The one thing your analytics should accomplish is to make it simple for everyone to discover data insights. Regardless of how much your company (and your data volume) grows, analytics end-users should be able to remain curious about their data. Every individual in every department should be able to answer questions about the entities that interact with their products with both speed and flexibility.
It’s the ability to remain curious about data that fosters a data-informed culture.
Whether you’re planning on in-house, on-prem, or SaaS, each of these options has its own advantages and disadvantages, which we won’t get into here (that’s another post). The most important thing to consider is whether your choice provides value for the price you’re paying -- and accomplishes what it should.
The problem with SaaS analytics
At first glance, a SaaS analytics solution sounds like an easy, straightforward choice. You can begin tracking user actions relatively quickly and make use of several out-of-the-box functions.
But, at the end of the day, the dirty secret of SaaS analytics is that you’re not paying for data insights. You’re paying for storage.
Keep in mind that data insights aren't comprised of canned metrics. They are an intimate look into how users, devices, and other actors interact with a product. Data insights come from the ability to ask any kind of question of your data on the fly, without having to worry beforehand about the complexity of the question or how long it will take for you to answer it. It comes from making sure your analytics is customized to your needs and that your data is fully under your watch.
When choosing an analytics vendor, you need to make sure you’re choosing one that will help you discover value in your data; that means, you need to make sure that you’re paying for insights, not storage. SaaS vendors almost never satisfy this basic requirement due to the very nature of their business -- they charge you for the event volume that you send them, because it costs them to store that data centrally.
Now, this might not be a problem if you’re a young company and your event volume is low. But as your organization scales in number of users and data volume, this cost quickly becomes prohibitive. And worse, you may be unable to move to another vendor without facing substantial lock-in costs. The last thing you want is to be stuck with a third-party service that doesn’t want to give you access to your own data.
How to evaluate analytics solutions
As you evaluate which type of analytics solution best suits your needs, make sure the vendor will be able to give you actual insights into your data as you scale, not just store more data. Ask yourself:
- How does the vendor’s pricing scale as the number of events you track increases? As your product has more users? As the number of analytics end-users increases?
- How does the vendor’s pricing scale on data insights? Do they offer behavioral analytics features or other tools that will help end-users ask questions faster and/or share insights better? What types of solutions are offered at each pricing tier?
- Can your vendor handle different data from different sources (e.g. web, mobile, IoT devices, infrastructure usage)?
- What does the end-user license agreement look like? What are the terms for vendor lock-in? Can you get your data migrated off their systems if you decide to move on?
As you decide on the right type of solution for your business, stay focused on what analytics should help you achieve: data insights for your whole company. Make sure your vendor offers features and tools that allow you to scale on insight, not just data volume. And lastly, ensure that your organization has control of its data, regardless of whether it’s a SaaS, on-prem, or in-house solution.
We're curious to know how you're thinking about deploying your analytics. Are you choosing to go with a SaaS solution? Why or why not? Let us know in the comments below!