Article

How to Make the Most out of Data Analytics

October 14, 2019

Data, data, and more data. We’ve got all this data that comes from our connected devices, how do we make sense of it? We all know measurement is important and that there is value in data. There are hundreds, possibly thousands, of analytics products in the field and all of the vendors have a different angle.
Don’t panic; it seems insanely complicated; it’s not. Once the data is generated you’ve got to get it into an analytics system, clean it up, analyze it and deliver it. Some key questions to guide you along the way include:

  • Who needs to use the data, and why do they need it? What are the metrics that matter to the organization and your customers?
  • How does the data need to be delivered? Examples include email notifications, dashboards, printed reports, etc.
  • Is the data high volume, low volume, or somewhere in-between?
  • Is the data real-time or can it be batched?

In general, before using the data you should have a good understanding of what questions you want to answer.

There are typically three types of data an organization can leverage:

  1. Easy-to-access data – This data is the low hanging fruit, data points that are easy to access. It is worthy to note that just because this data is easily accessible doesn’t mean it is valuable.
  2. Hard-to-access data – Data that you don’t even know you are collecting or that you don’t even know you have. This data is more difficult to get at. Sometimes the data that is the hardest to get to is super valuable.
  3. Undiscovered data – In IoT, there is a large amount of this type of data. This is data you could be collecting that you aren’t mostly because you probably haven’t thought about or identified how it could be helpful.
Data Analytics: Leveraging “Easy” Data and “Difficult” Data

If you’re starting an analytics practice at your organization, here are some suggestions to get you started.

  • If you have easy data and it’s super-valuable, totally use it, that’s a no brainer.
  • If you have easy data and its low value, well maybe ignore it or use it for training purposes. But don’t spend a lot of time on it. Look at your data and figure out what you need.
  • If you have hard data and its low value, ignore it, don’t even go there. Sometimes it is difficult for some engineers to pass by because it’s an interesting engineering challenge.
  • If you do have hard data, that’s high value, do it. Don’t ignore it. It is easy for executives to push the other way, you should be able to communicate; if we do this, we can charge X amount more per unit or product.
Data Analytics: Undiscovered Data

The undiscovered data is really interesting and is where some of the really interesting business opportunities are going to emerge in the next couple of years. Organizations should be thinking about what can we learn now that differentiates us from our competitors or gives us an edge that no one ever really thought of because it was never possible to collect this data.

Final Thoughts

In closing, if you can’t connect the data back to the business value, then why are you storing it? Don’t make analytics an afterthought. Don’t assume that the device has the data you need.

Want to learn more about IoT and leveraging data analytics for innovation and business success? Join SpinDance for an IoT Bootcamp on January 22, 2020 in Holland, MI.

 

 

 

Brian Tol, VP of Engineering at SpinDance
As a ‘full stack IoT’ engineer and business leader, Brian contributes daily to the strategy, development, and support of connected systems. When he’s not creating the future, he enjoys cooking, reading, and traveling.