Analysis as a journey
1. Analysis as a journey
Let's expand from exploratory, explanatory, and predictive to recognizing common data-driven objectives. However, we need to address change management first.2. Know what you know and learn what you don't
It's obvious to say "know what you know" and "learn what you don't". Here, it's about knowing your analytical limitations. This may mean getting really good at spreadsheets and performing excellent exploratory data analysis. Then, adding skills to spot quantitative decisions, graduating to explanatory and even data science methods for predictions. Often individuals think they need a cutting-edge model to predict a complex phenomena like a stock price, yet have no technical experience. That person identified an opportunity but without the proper knowledge they won't be able to execute. Start slow, build upon what you can do to avoid frustration while learning.3. The same is true for organizations
Organizations also must recognize gaps to a data-driven decision-making culture. It could be a lack of data, no technical personnel or a culture of HIPPOs.. that is the Highest Paid Person's Opinion.4. Maturing a data-driven culture - Descriptive
Assess where you or the organization are on this maturity diagram. Many emerging data organizations can only describe their operations with data, asking "what happened?"5. Maturing a data-driven culture - Diagnostic
After a little more experience, organizations graduate to diagnosis. Here, organizations start to understand the root cause of a phenomena.6. Maturing a data-driven culture - Predictive
Further, a smaller number of organizations have the technical fluency for predictive analytics. Predictive, using data to understand what will happen.7. Maturing a data-driven culture - Prescriptive
Lastly, the most data literate and driven organizations employ prescriptive analytics. Prescriptive analytics entails using data to define strategies and courses of action based on the previous insights. Make sure you know where you or your organization are within this journey.8. Cost vs. benefit
As you or your organization scale analytical prowess, now is the time to identify objectives. One common objective includes cost versus benefit. For most decisions there is a cost; time, money, personnel, opportunity to pursue something else. That cost must be weighed against the benefit, such as increased productivity. A data-driven analysis helps quantify the cost of an action for comparison to its benefit. For example, a cost is incurred for purchasing new software. However, there is a benefit, the software let's your factory build additional widgets. Analysis helps decide if the software cost is worth the benefit.9. Risk vs. reward
Another data-based objective exposes risk versus reward. Similar to cost versus benefit, a risk-reward objective is trying to quantify the risk against its potential benefit. Often, you're using data to understand the probability of an event, or a failure, then weighing that probability against a data-based expected value that is returned. Suppose you have $100 to invest, you could put it in the bank. This is low risk since you can get it almost anytime. However, that also brings a low interest or reward. You can invest $100 in a startup to power the Internet with mouse treadmills. If that works, you'll be rich... lots of reward there, but purchasing hamster treadmills is highly risky, you probably won't get your $100 back.10. Supply vs. demand
Lastly, many problems boil down to supply versus demand. Many scenarios have a demand placed upon a system, machine, or people that must be matched to a corresponding available supply. When you call a customer service line and wait annoyed on the phone, you are a demand on a system and there isn't enough of the supply of agents. Supply versus demand projects often forecast a demand of material, or personnel, then determine the optimal supply at any given time. Forecasting the number of calls when matching the number of agents needed is a supply versus demand data-driven decision.11. Putting it all together
In the end, we are left with this 3-by-3 matrix. You have common objectives; cost vs. benefit, risk vs. reward and supply vs. demand. And you have the methods; exploratory, explanatory, and predictive analysis. This 3-by-3 may not cover everything but it helps identify where data can be applied effectively.12. Examples
Here is that same 3-by-3 matrix filled in with even more examples. As you encounter opportunities, this framework should guide you to the type analysis and a succinct objective for your projects. I suggest pausing the video and reading these examples to learn more.13. Let's go!
Let's apply these concepts!Create Your Free Account
or
By continuing, you accept our Terms of Use, our Privacy Policy and that your data is stored in the USA.