A Recipe for Successful Data Analysis

Data is everywhere. Big data, micro-trend data, real time data, social-media and web tracking data - everyone is talking about it, and I get why.

The power of data is immense, and the desire to learn from your data is strong, but when faced with the mountain of data now available, it can sometimes be overwhelming. If you don’t know how to work with data you can feel like your sitting on the side lines of the big game, wishing you could play but not quite sure what to do.

To me, analyzing data is a lot like cooking a gourmet meal. It requires allot of prep work, and planning, you have to keep in mind your audience’s tastes, you need to test your ingredients and at the end of the day, its hard work. But like any good meal, the output is delicious, and better yet – no dishes to clean! So here is my step my step guide to gourmet data analysis.

1) Get to know your customers

Before you even plan the menu (i.e. analysis) – get to know your audience. You have to know their wants, their needs, and their challenges to plan an analysis that works for them. Just like cooking, if you serve someone something they are allergic to, or just plain don’t like, they will not take the time to really taste and consider your work.

2) Plan your output

This is one of the most important phases in the process. Taking the time before you even touch a data set to think about what you are trying to understand, and what problem you are trying to solve will save you when your head is stuck in a million rows of numbers. The plan will serve as your compass to help guide you when you feel like everything is getting very complicated by keeping you grounded in your objective and driving towards success.

3) Find for the right ingredients

Like any really good meal, the analysis you are doing may need a fair amount of “prep work”. In most businesses it is very rare that the data you need to get a full picture of the situation will sit with in one person’s, one department or even one countries data warehouses. Asking the right questions and chasing all the people you can to find what you need to will help differentiate your analysis and bring it from nice to know to need to know. Think of truffles – not exactly easy to find, but they add the perfect flavor to many gourmet meals.

4) Clean your tools

A grubby pan and a mealy tomato will make a very poor bolognaise sauce. The same goes for your data. Looking at your data and understanding your variables and how you need to structure your data sets will be vital to creating valid results. For example, if you have monthly data but need to look at overall y-o-y trends, it may be worth taking an average of the months and collapsing them into one variable so that you can do a range of comparisons of this variable with other non time-related variables. However, if you are looking for monthly changes, you may not want to do this.

5) Cook with calm

If you have ever seen a kitchen on Christmas day, you know, cooking is stressful. Well, so is analysis. Sometimes things just don’t work and instead of scrapping your whole menu, you need to adapt and remember the end goal. Get creative think of other analysis methods, or variables that can help you get the answers you need. No matter what – you will get something, and if you keep your calm, you are more likely to get there quickly and effectively.

6) Keep it simple

There is a reason salmon truffle soufflé with lavender ice-cream is not a popular dish. It’s just too much. The same goes for data analysis, when you overcomplicate the equation you end up with a confused and not relevant story. Always remember your plan and take the simplest, most effective route to get there.

7) Plate it beautifully

A little carrot flower has certainly never hurt anyone. The more you think about the visualization of your analysis, the presentation and the communication style you the better. To put it simply we are visual creatures and the more attractive and intuitive a presentation, the more receptive the audience. Even though you have found the ‘answer’ the only thing that matters is if your audience finds it too!