Analytics Tools: How to select an Automated Analysis tool
This is part 5 of our series on Selecting Analytics Tools, previous segments are available in our archives.
Today we’ll cover the last link on the analytics value chain: Analysis tools.
This is the newest link in the value chain, only becoming possible in the recent past with the rise of cloud computing and advances in machine learning / artificial intelligence. Analysis tools tackle the problem of information overload by looking through all of your data, all the metrics and dimensions, and reporting to you when unexpected changes are happening. In doing so they help you ask questions you didn’t even know to ask, ensuring you have no blind spots.
Because this category is so new it can be hard to compare options since they are all very different in their approach and capabilities. Still, there are some important things to look for when selecting your tools.
Great tools will be…
- High Fidelity. These tools are not useful if they give you hundreds of insights every day as it would take all day to look through them. Look for a product that reports only the most important insights.
- Low Effort. The more personalization and customization these tools require, the less likely they are to tell you things you did not already know. The best tools can process all of your data and find insights with minimal setup.
- Repurposed Tools. Many tools that were originally built for monitoring manufacturing processes or devices have recently been “repurposed” for business analysis automation. They are referred to as “Anomaly Detection” tools or similar names, but they were not built for business data and as such provide a low quality experience.
- Faux Automation. Many data visualization tools have tried to increase their value (and prices) by claiming the ability to automatically analyze your data. In reality, they still require a person to explore and find insights, it does not happen automatically. This might be right for you but is a very different use case than automated analysis.
As this is such a young category you will see many more options in the coming years. However, within 5 years I would be surprised if it is not in use at almost all companies.
One more thing…
This week we have broken down the tools in the analytics value chain and how to select them. However, it is also important to consider that there are platforms that do collection, storage and visualization (and sometimes analysis) in one package.. How do you choose between assembling your own chain or buying one of these all-in-one platform?
You should choose an all-in-one platform if…
- You are resource constrained and lack the time or expertise to evaluate, select and implement the tools to build your own value chain.
- You only need basic metrics that are generic. All-in-one platforms provide default metrics, but they typically lack much customization.
- You don’t expect your needs to change in the coming years.
You should assemble your own value chain if…
- You need to assemble metrics and analytics that are across many different functions of your business. It’s rare that an all-in-one platform will effectively help you, for example, connect your CRM data with your website traffic and your customer support data.
- You want the flexibility to answer new questions in the future that you can’t even imagine right now.
- You are an expert (or aspire to be an expert) and need as much power in your analytics tools as possible.
In Summary: The choice of your tools is a critical step in building the analytics capabilities of your company. Choose carefully, as it is harder to change these tools than it might seem. Hopefully the criteria and discussion from this week equips you to be a savvy shopper. If you have any questions about tools just let me know as I’m happy to help!
Quote of the Day: “Buy! buy! Says the sign in the shop window. Why? why? Says the junk in the yard.” – Junk by The Beatles / Paul McCartney