A few days ago I met an old acquaintance of mine, who had been working in the marketing department of an IT giant. As we got to talking, I could see that he was a data skeptic — someone who did not believe in the power of data. Data Science is not for those who fall on the business side of the company, he said.
Quite surprisingly, it is not just him who considers the myth that only techies and analysts need Data Science skills for their work. There are many with the same belief. And to state the obvious, they are wrong.
It is true that Data Scientists employ advanced methodologies and require training in programming languages to do their jobs. This is why people think they cannot take advantage of Data Science as they lack such skills. However, Data Science can be bifurcated into Data Literacy (DL) personas based upon a person’s knowledge and interest in the subject. The personas are:
Data-driven Executive: Data-driven Executives understand the power of analytics, which is the discovery and interpretation of meaningful patterns in data and their application for effective decision making. Data plays an integral role in the decision-making process. They hold their team accountable for their work and can understand when analytics has been done in the right manner or not.
Data Scientist: Data Scientists are well versed with advanced analytics methodologies. They can solve almost 100% of business problems using analytics. They are adept in the use of advanced tools like R, Python, SAS to manipulate data and build a model. These Data Scientists can align stakeholders toward an actionable solution and excel at data-driven decision making.
Citizen Analyst: Citizen Analysts are data-driven employees who can solve 80% of their business problems using a structured approach to analytics using simple methodologies and in the process align the stakeholders. They can get to an actionable solution and move the key metrics for the company. Some citizen Analysts can also be taught advanced analytics.
Data Literate: Data Literates understand the analytics landscape and can be an active participant in discussions involving data. They are willing to hone up their analytics skills. A good recipe-based analytics program with hands-on practice on the most employed analytics techniques can take them and the company a long way toward being data-driven.
Data Enthusiast: Data Enthusiasts believe in the power of data. They are eager to learn more about how to use data and interpret it in their work. A good data literacy program could usher them and the company to new heights toward being data-driven.
Data Skeptic: Data Skeptics don’t believe in the value and power of analytics. They see analytics and data as ‘burden’ to their work. They can derail any data literacy project unless they are carefully nurtured into becoming data enthusiasts. A good data awareness program is imperative to turn the skeptics into Data Enthusiasts.
From the above-mentioned categories, one can clearly see that the profile of Citizen Analysts is suitable for those who fall on the business side, such as my acquaintance. These are marketers, managers, and business folks who need to leverage data in order to drive and make decisions. For eg, a marketer needs to know who he/she should target for a particular campaign, what segmentation strategy should he/she follow, or what kind of offers should be sent to whom. Data Science plays a pivotal role in making these decisions.
To summarize it, Citizen Analysts are those who employ simple tools and methodologies such as MS Excel and Correlation Analysis to solve 80% of their problems. About 20–30% of their work involves doing analysis. On the other hand, 100% of Data Scientists’ work revolves around the analysis and deriving insights.
Now, let us delve into the essentials statistics for Citizen Analysts to help understand the said role. A Citizen Analyst needs to be familiar with some tools in order to master their role. They are:
- Microsoft Excel: Most Citizen Analysts use Microsoft Excel for their works revolving around the analysis.
- Business Intelligence (BI) tools: These tools are used to access data, and software like Microsoft Power BI, Tableau, and Qlik.
In addition, Citizen Analysts should have command over a few methodologies such as:
- Aggregate Analysis: This helps in analyzing single variable distributions. It answers questions like what does our customer look like — definition based on single variables like gender, age, or product usage, etc. So typically Profiling a customer base or product usage uses Aggregate Analysis.
- Correlation Analysis: When two or more things are considered in relation, it refers to Correlation Analysis. It helps in answering questions like why has revenue slowed down, what segments responded well to what campaigns, among others.
- Trend Analysis: It is basically like Correlation Analysis but the difference is that Trend Analysis occurs over a period of time. In other words, whenever a time component is involved and two or more things are considered, it is called Trend Analysis. It helps in answering questions like why has growth slowed down this past quarter, which search terms are peaking for different customer groups.
- Sizing/Estimation: Technically, this is not an analytics methodology as it not based on historical data, but it is helpful in building a business case for an impending decision where data isn’t available. For eg, A company is relocating their headquarters to the Boston area and they need to estimate the talent available in the area in order to not run into hiring issues. The Estimation uses a structured method, data, assumptions, and facts.
A Citizen Analyst needs to be well versed with the above-mentioned technologies. Moreover, he/she should be aware (not fluent) of advanced methodologies like
- Predictive Analysis: Linear Regression, Logistics Regression, Decision Tree, etc.
- Cohort Analysis
- Time Series
To reiterate, a Citizen Analysis needs to know these methodologies but hands-on experience is not needed. He/she should, however, understand what methodologies should be used for what kind of problems.
Further, there are a few statistical concepts a Citizen Analyst should know. These are:
- Averages: This includes concepts of Mean, Median, and Mode
- Standard Deviation, and Variance
- Error and Confidence Interval
- Correlation Analysis
These are the essential statistics that one should know for becoming a proficient Citizen Analyst.