How to Implement Data Science for Your Business Part 1: The Data Science Workflow
If you work for a small or medium-sized business or startup, implementing Data Science and predictive analytics could have a huge impact on your organization. Unlocking the insights hidden in your data will undoubtedly lead to better decision making and potentially give you a leg up in your industry.
Unfortunately, figuring out how to do that seems really complicated. Most of what you read online will present Data Science as an overly convoluted field that is completely inaccessible to the average businessperson. It almost feels like you need a Master's in Computer Science just to get in the door!
This is a real shame, however, because Data Science really has to be accessible to the average businessperson in order for it to be effective. When used properly, data answers questions and guides key decision making across an organization. Most key decision makers, it turns out, aren't PhD's in Computer Science with a strong stats background. So the idea that Data Science is only accessible to the experts doesn't make a lot of sense when you really think about it.
The reality is that many small businesses without a big technical staff are still getting a lot out of Data Science and analytics by setting up simple but powerful Data Science workflows that bring insight to the right people.
These workflows typically look different, but they have a few things in common.
OK, it might be a little generous to call that "Data Science," so here's a slightly more complicated example:
More importantly, however, both examples are extremely accessible to any small business, possibly with just a little bit of outside help. A person at your company with the right idea, and a consultant with the right know-how could easily make the second example into a reality in less than a week of working together. If you are reading a blog post on an analytics website then you probably have some bright ideas of how your organization could use Data Science. And that same analytics website just so happens to be for a Data Science consulting company that has the right know-how...
Perhaps the only thing left to do is to reach out to see just how close you really are to making that idea a reality that could have a major positive impact on your business!
Unfortunately, figuring out how to do that seems really complicated. Most of what you read online will present Data Science as an overly convoluted field that is completely inaccessible to the average businessperson. It almost feels like you need a Master's in Computer Science just to get in the door!
This is a real shame, however, because Data Science really has to be accessible to the average businessperson in order for it to be effective. When used properly, data answers questions and guides key decision making across an organization. Most key decision makers, it turns out, aren't PhD's in Computer Science with a strong stats background. So the idea that Data Science is only accessible to the experts doesn't make a lot of sense when you really think about it.
The reality is that many small businesses without a big technical staff are still getting a lot out of Data Science and analytics by setting up simple but powerful Data Science workflows that bring insight to the right people.
These workflows typically look different, but they have a few things in common.
- They begin by gathering data and storing it an logical way
- They apply some type of statistical analysis to that data
- They present the findings of that statistical analysis in a manner that key decision makers can understand
- Key sales data is entered into a spreadsheet and maintained by a single employee
- In that same spreadsheet, a pivot table is used to calculate a few simple metrics for each sales rep such as total sales volume and average deal size
- An email is sent out every month to the sales staff with those metrics as well as a few key findings/takeaways
OK, it might be a little generous to call that "Data Science," so here's a slightly more complicated example:
- Stock price data is scraped on a regular basis and stored in an SQL database
- A DBSCAN clustering algorithm identifies clusters of stocks whose daily prices tend to move in tandem
- A Tableau dashboard is built that compares an individual stock's price performance to it's cluster's price performance, potentially identifying stocks whose price might correct in the short-term to fall back in line with the larger group performance
More importantly, however, both examples are extremely accessible to any small business, possibly with just a little bit of outside help. A person at your company with the right idea, and a consultant with the right know-how could easily make the second example into a reality in less than a week of working together. If you are reading a blog post on an analytics website then you probably have some bright ideas of how your organization could use Data Science. And that same analytics website just so happens to be for a Data Science consulting company that has the right know-how...
Perhaps the only thing left to do is to reach out to see just how close you really are to making that idea a reality that could have a major positive impact on your business!