.jpg)
Every business, independent of size, has the same basic questions: how much am I expected to grow next year, how much is my sales volume this month, and how can I optimise my processes?
The answer is relatively simple: data. If you are not gathering data on all aspects of your business, you should start as soon as possible. The more detailed, the better because capturing all the details will enable you to discover underlying patterns you never thought you would see.
Here’s a practical example: imagine you sell t-shirts, and all the colours have the same monthly sales value. However, because you gathered customer details through data analysis, you discover that 75% of your brown shirt buyers are middle-aged men who live in the suburbs.
You have now gained some ground to make a few decisions. You can cross-check with your ad settings and determine if you have a problem with targeting or if middle-aged men simply like brown shirts. From here, you can either reinforce the ad investment for this slice of the market or aggressively try to expand your targeting to a different audience and find a new niche.
A simple insight empowered this entire decision-making process. Insights are valuable pieces of information that will support decision-making processes.
Statistics and plotting (graphical representations to illustrate statistical or numerical information) are powerful tools that enable you to understand your business if you use them correctly and to your advantage. If you are analysing your sales and have a small value range, an average of those values is a good measure because it will represent a typical sales value for the specified conditions with a high level of accuracy. Suppose you have an enormous range indicating significant outliers (an unusual or significantly different data point compared to the other values in a data set). In that case, a median is a much better choice.
One of the most common errors is looking at a data table and assuming something like, “This column’s average seems to be X”. You’d be amazed at how often people get the value wrong and don’t even realise it. We usually refer to these critical values as KPIs or key performance indicators.
These simple values should be driving forces behind your business, so it’s crucial for you to identify the KPIs that are useful to its growth. A common mistake is to have multiple and very complex KPIs: problems should be addressed at an atomic level. Don’t expect to maximise all your KPIs at once.
Usually, the most meaningful and vital KPIs have simple metrics like average invoice value or daily average sales. Overcomplicated KPIs will not make them any better than the more simple and direct ones, so try your best not to overcomplicate them.
You don’t need to analyse endless rows of data to understand your business pulse or detect patterns. As mentioned earlier, plots are your best friends, and there are plenty of them for you to use.
The most common ones, for good reasons, are bar plots and time series. Bar plots allow you to immediately see your best (or worst) performers through bar height visualisation.
You can put them in an order that allows you to see this quickly or not order them to understand if any pattern shows up. As the name suggests, time series are very good at showing a variable’s evolution through time and are suitable for identifying trends and seasonality. There are also some more precise plots, like box plots, that are amazing analysis tools. However, be careful when presenting them because they require deep issue knowledge to interpret and are often confusing without context.
There are also less helpful plots that are often used, but that should rarely be used. Looking at you, pie charts. Pie charts have their uses; they are suitable for illustrating a variable like sales and visually understanding which product category is responsible for the most sales. However, that completely disappears if you have various product categories or the slices represent similar values.
They can be useful if you want to show that, for example, four product categories have a similar weight but are terrible if you want to showcase the best performer. You can add labels, of course, but at that point, the plot’s primary goal is already lost. A table with ordered rows would perform better in this case.
Another essential aspect of plots is colour. The colours should relate to the data you are plotting and immediately relate to a good or bad outcome. For example, creating a sales bar plot where all bars are red is an excellent way to scare everyone before they properly analyse the information it contains. Choose neutral colours if you don’t have a comparison value or set colours according to the performance against a metric like the average. For example, if the value is above average, show it as green; if it is lower, show it as red.
Nowadays, with relatively simple tools, you can learn a lot about your business and make informed decisions based on KPIs and Insights. Data is your friend; learn to explore it, and you’ll love how helpful it can be.
.svg)
