Data Basics

Digital Marketing - Study Notes:

What is data?

Before we can begin to understand data, we need to define it. Essentially, data is information in raw or unorganized form such as letters, numbers, symbols that refer to, or represent conditions, ideas, or objects.

Challenges facing analysts

Data is potentially limitless and it is important to recognize the challenges this poses. Here are some of the core challenges related to data.

• The first challenge is amount of data
• The second challenge is that of interpretation. With the abundance of data, we can try and manipulate it the way that we want, and try to get it to tell the story that we want as well. This can lead to falsehoods, which is one of the big issues with analyzing data.
• The third challenge is security. This is becoming a fundamental issue in terms of privacy. There have been plenty of examples where data has been compromised, and was leaked on to the internet, so it is crucial to think about the way we store and secure data.
• The fourth challenge concerns relevance. Data, quite frankly, is unequivocally linked to time. With so much data, it can become irrelevant quite quickly. Today, real-time data is actually becoming something that people, particularly consumers, expect, so we need to react very quickly.
• The accuracy of the data is another key challenge. How accurate do we make the data? And is it capturing all the factors and variables that we want it to?
• Coming back to the point about relevance, data is time-bound. It is important to make sure that you use the data quickly, almost in real time, as that is becoming an essential factor in business today. Largely speaking, organization systems often make it difficult to do that.

Common misconceptions

Let’s think about some of the more common key misconceptions when thinking about your data.

Misconception 1: You need lots and lots of data to do analysis.

This comes back to statistics. A larger dataset means more meaningful results. Sometimes having a massive pool of datasets actually can be more complex and confusing, and the common phrase ‘analysis paralysis’ can prove true. So think about that and how you might not necessarily need to focus on massive datasets.

Misconception 2: Time will tell

Historical data may prove to be less valuable upon further reflection. Relying on old data may not be the best predictor of the way things are going to happen in the future. Be wary of this when evaluating your data.

Misconception 3: Correlation equals causation

Correlation basically means that, when one thing happens, another thing also tends to happen. So the two things are correlated together. The problem with that is that it can be difficult to discern why the second thing happened. For example, if you make a change and customers respond, are they responding because of the change you made or because of something else in the customers’ psyche. When analyzing data, make sure you don’t make assumptions about causation.

Misconception 4: You can ignore the wider context

This occurs when you try to look at data through too narrow a lens. Understanding context is essential to make sure you’ve got the wider, holistic picture around what the data is saying.

Misconception 5: Percentage and percentage points are the same

In fact, they are completely different. Be careful about that because you can experience different results if you just look at percentage against percentage points.

Misconception 6: The average trap

This can occur when you have large sets of data, and people try to understand what is the average of what that data is telling you. The problem with average or median is that you don’t truly understand what’s happening in the various data pools so you might miss out on vital insights or you might miss out on what the data is really telling you. However, if the standard deviation from the median is very, very narrow, then the average becomes more reliable.

These are just a few key misconceptions to watch out for to make sure that you don’t fall into data traps when analyzing your data. Standard deviation is a great way to overcome some of those core biases that you might come across when analyzing data.