The Data Analysis Framework

Wouldn’t it be great if you would know exactly what to do when it comes to data?
This is a step-by-step framework on how to translate data into actionable insights and business decisions.

Are you drowning in data?

One struggle I hear from marketers, again and again, is that they are drowning in data. You are seeing walls of endless figures and don’t know where to start.

Of course, you want to finally start translating data into cost-saving and money-creating decisions for the business. But you just don’t have a clue where to begin. And spend too much time playing around with it without gaining any learnings.

You know you should be able to translate that data into actionable insights. But you don’t know how. It seems there is a missing clear path from data to insights to the next steps for the business. And without business implications, data is worthless and a waste of time.

What if you would know exactly what to do when it comes to data?

Wouldn’t it be great you could self-efficiently and autonomously work with data?

If you could render the endless figures into actionable insights to allow you to become a better, more effective marketer?

Imagine you had the power to look at your companies’ data and translate it into the next steps for the business.

If you would really understand and use your data you could not only drive your marketing more precisely but you could easily create marketing campaigns with predictable, repeatable success. You would take the uncertainty out of marketing.

The problem is people tend to start with the data.

And from there they try to figure out what to do with it, how to analyze it, and what it means for the business. However, this usually means getting lost in the data and no actionable insights at all.

A much better approach is to work backward.

Start with the business challenge, define the key questions you need to address to solve the challenge, and from there decide which data you need to answer them, how to analyze the data, and what actions to take based on the outcome of the analysis.

The Data Analysis Framework

The Data Analysis Framework

This is so much easier as you know exactly what part of the data to focus on (vs. the endless figures). And by starting with the desired business recommendations you ensure that your analysis will be actionable.

And this is exactly why you need a data analysis framework. It’s the link between the data and the business world. It guides you on a clear path between data, what data analysis methodologies to use, and actionable recommendations for the business. Without wasting time playing around with the data.

Step1: Define the business challenge

Yes, data can be key to understanding a problem. However, being able to define a problem is 50% of a solution.

Instead of starting with the data start with a clearly defined business problem. This will cut down the type of data you have to look at and the potential analysis technique to use very fast. With fewer options, you will easily avoid being overwhelmed by data and make sure your data analysis is actionable.

So let’s have a look at how you can define a Business Challenge and good Key Questions based on that challenge which you can answer with data.

The Business Challenge

Defining the general Business Challenge is the easy first step you can take to come up with Key Questions.

It’s just the general situation the company is currently in and the complications to that situation that are coming up, e.g.:

  • Industry trends
  • Competitors
  • What change for you, the market, or the customers

Basically, anything that will help you come up with some key business questions that you want to answer. It doesn’t need to be too long. 2-3 sentences can be enough (but it can be longer if necessary).

Here is an example:

In 2020 even though OrangeCo’s customer base is growing their average customer lifetime value is dropping. Furthermore, new players are moving into the industry. The profitability of their marketing efforts is dropping (-5% YoY) while they are losing their share of voice (-10% YoY).

Key Questions

Based on your Business Challenge you can define Key Questions you can solve with data and which answers will help you solve your initial overall challenges.

Define your Key Questions as SMART questions.

SMART Key Questions

SMART Key Questions

Be specific to keep it simple and on point to one facet of the challenge. Make sure it is measurable by adding KPIs or metrics. Make sure it is action-oriented and relevant so you can act on the answer and business decisions can be implemented afterward. And make it time-bound by having a specific time horizon.

Example questions:

  • Why has OrangeCo’s marketing channel profitability decreased by 5% YoY?
  • Why has OrangeCo’s average lifetime revenue dropped significantly in 2020?
  • How will OrangeCo’s marketing channel profitability develop in the next year if they continue like this?

Step 2: Look for potential solutions

In the last part, we talked about the Business Challenge and how to define SMART Key Questions to make sure your data analysis will be actionable and lead to business decisions.

In this section, we will shift towards solutions. However, not the actual answer to your Key Questions but rather core issues, reasons for what is happening, or just the general approaches on how to answer a particular key question.

If you have ever heard of hypothesis testing –> this is pretty much it!

Specify which data you need

Specify which data you need

Let’s say for example your Key Question is the following:

Why has my company’s marketing channel profitability decreased by 5% YoY?

Two reasons for that could be:

  1. One of the advertising channel’s ROI decreased
  2. Average order values decreased

Since we precisely defined two possible issues it is now A LOT easier to focus on the right data. You don’t need to look anymore at the overwhelming entirety of data but only at that data that is relevant to your possible reasons.

So for the ROI issue, you would just get the ads data from your corresponding ad platforms (e.g. Google Ads or Facebook Ads) and for issue #2 you would pull your transactions data from your CRM.

This is much better than just looking at all your data and waiting (and possibly wasting time) for an actionable insight to improve your business, right? 😀

How many solutions should you have?

It can be overwhelming to try to find all possible solutions to a Key Question. In theory, solutions should be mutually exclusive so they don’t overlap but collectively exhaustive so you theoretically have ALL possible solutions covered.
In reality, though you don’t have to find ALL possible solutions. Use the good old Pareto rule instead. Find the 20% of solutions that inflict 80% of the effects. Use logic to weed out improbable solutions to focus only on those which most probably make sense.

Step 3: Decide how to analyze the data

Now it’s finally time to choose a technique to analyze your data to get some actionable insights for your business!

When people talk about data analysis this is what they usually refer to: The applications of calculations, models, and techniques to translate data into insights.

Analyze data and find insights

Analyze data and find insights

So, how do we know which methodology to use? Unfortunately, it’s not that easy to answer. Especially if you are new to data analysis.

The good news first:

At the end of this section, I’ll give you a tool that you can use to easily find the exact data analysis technique you have to use for your specific Key Questions.

However, let’s talk about the four different areas of data analysis first.

The different types of data analysis

Descriptive Analysis

This is the foundation of all data insight. It is the simplest and most common use of data in business today.

Descriptive analysis answers the “what happened” by summarizing past data.

Diagnostic Analysis

After asking the main question of “what happened”, the next step is to dive deeper and ask why did it happen? This is where diagnostic analysis comes in.

Diagnostic analysis takes the insights found from descriptive analytics and drills down to find the causes of those outcomes.

Predictive Analysis

Predictive analysis attempts to answer the question “what is likely to happen”. This type of analytics utilizes previous data to make predictions about future outcomes.

This type of analysis is another step up from the descriptive and diagnostic analyses.

Prescriptive Analysis

The final type of data analysis is the most sought after, but few organizations are truly equipped to perform it.

Prescriptive analysis combines the insight from all previous analyses to determine the course of action to take in a current problem or decision. Prescriptive analysis utilizes state-of-the-art technology such as machine learning.

Know exactly which data analysis method to use

As we have shown, each of these types of data analysis is connected and relies on each other to a certain degree. They each serve a different purpose and provide varying insights. Generally speaking, moving from descriptive analysis towards predictive and prescriptive analysis requires much more technical ability, but also unlocks more insight for a business.

As you can see those categories are also just merely parent categories for all the different applied statistical calculations and models in each of them.

And oftentimes you can even use different approaches to solve the same problem. That’s why it is so difficult to know which methodology to use.

Wouldn’t it be great to just know exactly which data analysis method to use for your specific challenge?

To give you some practical support here, you’ll find a tool below which lists some of the most common Business Challenges and Key Questions and which data analysis method I would use to solve them!

The Data Analysis Guidepost

Use the tool by filtering the Business Challenge and Key Question columns to find the data analysis method you’ll need!

Step 4: Putting it into action

With the above, you already worked through a lot: You defined your Business Challenge and Key Questions, decided which data to use and where to get it from, and chose an appropriate data analysis method to get some actionable insights.

However, one significant last step is missing: Putting your insights and recommendations into action.

The full Data Analysis Framework

The full Data Analysis Framework

I experienced it so often that data was analyzed, fancy calculations and statistical models were applied and then… nothing happened.

That is why it is so important to follow the data analysis framework. It has one goal: to make your data analysis and your translation from data to insights as actionable and relevant as possible.

And that’s the only thing that counts: All analysis is worthless if no business action is triggered by it in the end.

So always have a plan in place on how to implement actions based on your data insights.

When implementing an action based on data insights you usually have two options:

You are the decision-maker yourself and can initiate the implied changes yourself.

  1. Make sure you have clear todos or a strategy based on your insights in place to make sure you follow through with the action.
  2. Identify potential stakeholders you have to work with and which are affected by your actions.

Somebody else is the decision-maker.

  1. Identify the decision-makers and stakeholders.
  2. Communicate your business recommendations.
  3. Have a clear roadmap for them on how to implement your recommendation to make the insights actionable.
  4. Ideally also include the impact those recommendations will have. This will make it a lot easier to convince any stakeholders.

That’s it! It is now time for action! You have a data analysis plan in place and know what to do with your data to help you with your business challenges.