Data Analysis Tips To Power Your Business

What’s going on with your business?

Let nothing be a surprise – that’s a motto of ours.                               

One of the critical components of management involves monitoring. Things change with your employees, customers or clients, and your community (geographic or web-based) stakeholders.

The simpler it is to know what’s happening—the good, the bad, and the ugly—the quicker you can respond and keep your operation smooth.

Think about some questions which might affect your business. Examples:

•  Who typically buys my service? Why not other demographics or psychographics? Should I expand my target market or focus on meeting my current demographic better?
• How long does it take someone on your website to make a purchase? Should I keep them longer so they will act or make their time shorter and simpler?
• What proportion of email subscribers click links? Are my emails even worth my effort?
• How can I keep more employees and reduce turnover? Do long-term employees prefer higher wages, better benefits, or more exciting perks?

Rarely will professional experience, fine-tuned instincts, age-old sage wisdom, or educated expert opinions be better at telling you the future than data trends, when available. An increasing body of research since before the 1950s suggests that random chance is often better than even the greatest minds. Here we will discuss the fundamentals of gathering and using data.

“Data” means information. That’s it! “Data analysis” is just looking at information and can be as simple as reading this blog. What’s crucial to analysis is that you can trust the source of the info.

Whom are you looking to know more about?

Let’s say you send a survey to customers to gauge how simple your software is for their use. Forty-five respond and rave excellent reviews!

Wait… what’s this? You only got responses from women!

If you know that you regularly have an equal share of male and female customers, then your sampled pool of participants does not represent the whole. You might find that only women like your product! This demonstrates the power of demographics (and psychographics when appropriate) so you can check with the different subset groups of your audience.

What good is knowing that your online store is financially performing well when it’s not a significant part of your overall fiscal income (as with Amazon, a little-known fact)?

Here’s what to take away: you don’t need to know everything about your business. You need to know the right things about your business.

So, make some calls. The men you emailed might not have had access to your survey. They might feel too cold about your product to waste time with the survey. Or, another fact is that men just don’t respond to surveys at the same level as women, so you may actually need to oversample, or you’ll always have to weight each man’s response heavier.

“Data analysis” isn’t as scary as it sounds. 

Data (plural) are a collection of information. Each datum (singular) is a unit of information.

Data analysis is summarizing lots of information in a simplified, usable format. Statistics dive deep into data to understand trends and probabilities of future (uncertain) outcomes.

A data analyst simplifies information (like Patti and Dr. Jack). At the same time, a statistician may go more detailed with the consequences of the information (most definitely Dr. Jack).

Each time you survey a group of people, you collect information from a sample of people to roughly gauge the sentiments of the entire population. Data analysis uses pieces of information—data—to summarize that information for better decision-making.

Management must consider many factors to make well-informed decisions that affect the wellbeing and longevity of their business. Fortunately, data analysis does not require an advanced degree. You probably learned all you need to know to analyze data from your middle or high school math classes (but you use it or lose it, amIright?).

Likelihood and Central Tendency

Let’s look at measures of central tendency. This is like looking at the middle of the graph. To somewhat oversimplify, central tendency helps you gauge what you can expect most of the time.

The arithmetic mean (A.K.A. the average) is found by adding the elements of a set (ex: adding all the ratings in a 5-star rating system) and dividing by the number of elements (ex: the number of people who rated the service). In other words, mean = (sum of parts) / (# of parts).

In the set {1, 2, 2, 4, 5}, “2.8” would be the mean because (1+2+2+4+5) / 5 [because there are five data points) = 2.8. In this case, the average star rating on a five-point scale would be 2.8 stars.

The median (the “middle”) is found by laying out all the data in order and then finding the middle point. In the set {1, 2, 2, 4, 5}, “2” would be the middle because it is the middle point (the third data point is smack in the middle of five data points). The median is beneficial here because you know at least half (3 of 5 in this case) of ratings would be two stars or lower.

So far, we can tell that the mean skews the average rating to 2.8 stars—looking like you have a better chance of enjoying the product. However, if 50% or more of people rated two stars, you have a less likely chance of enjoying the product.

The mode is the data point used most frequently. In the data set {1, 2, 2, 4, 5}, “2” would be the mode because there are two “2”s and only one of the other ratings.

The mode can offer some information but is only sometimes useful empirically. Knowing that more people chose a rating of “2” does not mean you will if you use the product, but it can tell you that more people rated it lower. The mode cannot typically be used to predict the future but has its uses.

You probably have what you need to get started.

A great place to start is familiarizing yourself with spreadsheet software. Create some tables of information and play around with the formulas and functions. Think Microsoft Excel, Google Sheets, or Apple Numbers.

From here, search online! You’d be surprised at what spreadsheets and data analysis can do. We like to think that every tedious process can be automated… at least in some small way. The more you ask Google, the more you will find.

Best of luck, and when you need a data nerd to help with the bigger problems—you know where to find us!