Jain and Sharma's BADIR™ Framework

Extracting Information From Data, Intelligently

Jain and Sharma's BADIR™ Framework - Extracting Information From Data, Intelligently

© GettyImages

Turn your data into neat packages of information for better problem solving.

Do you have a nagging feeling that you should be using data analytics more, but you don't know where to start? Maybe you're suffering from a case of analysis paralysis, and you're bewildered by the challenge of making data work for you.

These days, it seems like data is everywhere. There's the Internet of Things, email, photo keywords, Google search, Fitbit, self-driving cars, smartphones, news feeds, social media channels, digitized company records of every conceivable kind… The list goes on. And on. Data is now more plentiful and pervasive than ever before, giving rise to the term "Big Data."

The challenge is knowing how to use such superabundant digital information, and that's where this article comes in.

Why Look to Data for Answers?

More and more of what we do leaves a digital trace (or data) and, essentially, Big Data is a method of harvesting that information on a scale that traditional approaches to data processing cannot even attempt. You can use this data to improve future experiences, spot trends, and enhance your business operations.

The way that Big Data works is usually explained using three Vs:

  • A very high Volume of data…
  • With lots of Variety…
  • Flowing at a high Velocity.

But size isn't everything in the world of data. Whether you're dealing with Big Data or just a small amount, the key to unlocking its potential is to capture and manage the information that's going to be most useful to you, rather than just gathering together a vast jumble of facts and figures.

Obtaining and managing the information is one thing, but picking the most relevant needles from the haystack – and using them correctly – is quite another.

If you get it right, data's potential is enormous. Airlines are using it to improve flight ETAs. Retailers are tapping into customers' buying histories to create personalized offers. Smaller companies and decision makers at all levels are using it to make smarter business decisions. Data is helping to enhance product quality, to fine-tune marketing operations, and to boost performance in countless other ways.

The BADIR™ Framework's Five Steps to Data Success

BADIR is a data analysis framework developed by the authors of the 2014 book, "Behind Every Good Decision," Piyanka Jain and Puneet Sharma. It allows you to convert raw data into decisions and insights by analyzing it in a structured manner, and it has five sequential steps:

  1. Business question.
  2. Analysis plan.
  3. Data collection.
  4. Insights.
  5. Recommendations.

Terms reproduced by kind permission of Piyanka Jain of aryng.com. Further information and an online quiz about this framework can be accessed here.

BADIR combines technical skills with soft skills, and is applicable to any type of business. And the best news is that you don't need to be a data analyst to use it. Whether you're a CFO or an office junior, if you can use an Excel spreadsheet then you can work with BADIR.

Let's take a more detailed look at the framework's five steps:

Step 1: Ask the right Business questions

Although BADIR is a process for analyzing data, it doesn't begin with gathering data! Instead, your first step is to articulate what you want to know, and to decide what business question(s) will provide you with the answers that you need. Pinpointing the specific issues that you wish to address before you roll up your sleeves and delve into the data ensures that the information you're going to collect will be useful.

Ask yourself the usual problem-solving questions: what, who, where, when, why, and how. For example, an HR manager who is looking to build a retention program that reduces staff turnover might ask who stays with the organization for the shortest time, and why they leave.

Also, consider who can work with you to achieve this, the schedule that you need to work to, and other good project management practices. The more questions you ask at this stage – and the better those questions are – the better the information you'll mine from the data. Break big questions down into more manageable parts if you think this will help.

Step 2: Create an Analysis plan

Once you know what your business questions are, set goals that you want to achieve during your analysis. This will help you to focus on answering your questions. You can set SMART goals to ensure that they are clear and achievable.

Next, you need to come up with hypotheses, and the criteria to prove or disprove each one. This will give your analysis some initial direction. A hypothesis is an informed guess about what's causing the issue that you're trying to resolve. You'll be able to test your hypotheses once you've gathered your data. (Be patient, that comes soon!)

With your hypotheses in place, you can choose the methodology for testing them. Aggregate analysis, for example, allows you to describe and compare different groupings. Correlation analysis is a way to identify relationships between different groups, and trends analysis examines performance and growth over time.

You also need to create a data specification. To do this, identify the type of data that you'll need to gather and determine the granularity of it. For example, if you have to analyze sales data to answer your business questions, you'll need to determine whether it should be at a weekly, monthly or yearly level.

In our example, our HR manager might set a goal to "identify segments of those who leave the company and the factors that drive resignations," and hypothesize that it's the youngest, female staff who leave soonest, so that they can earn higher salaries. He or she may choose an aggregate analysis methodology and identify the company's personnel and exit interview records as the data required. The HR manager has now started to compile a project plan.

Finding This Article Useful?

You can learn another 46 problem-solving skills, like this, by joining the Mind Tools Club.

Join the Mind Tools Club Today!

Step 3: Collect and validate the Data

And now – drum roll, please – you gather the data.

If you've successfully applied Steps 1 and 2, you should now have a timeframe for answering your questions, you'll have a project team, and you'll have decided what data you want to gather. There are specialized software programs available that can make gathering data relatively easy.

To complete this stage, you should collect – or "pull" – your data and validate it to ensure its integrity. This means checking that the data is clean and free of input errors, and that you've gathered it in a usable format.

There are various ways to validate data. You could, for example, perform a range check, where you ensure that your data sits within a specific range (say, 1-50 or between certain letters of the alphabet), or a type check, where you assess data to make sure that it is the correct type (numbers or letters, for example).

At this stage, our HR manager would tabulate his company's staff tenure and exit interview data, and check that there are no obvious issues, such as data being duplicated (which can occur, say, if you have used more than one system for collecting it).

Step 4: Search for Insights

You're now ready to assess the valuable and accurate data that you gathered in Step 3, using the hypotheses and methodology that you decided on in Step 2.

Here, you aim to look for and examine patterns, to prove or disprove each one of your hypotheses. You may be able to eliminate some of them and refocus your energies on others. Lastly, you should present your findings in a coherent fashion.

Our HR manager might, for example, see a pattern proving that women are indeed likely to leave the company more quickly than men, and that earning a higher salary is the main reason. However, it's actually those aged between 30 and 40 who are most likely to go.

When you've found your insights, you can move on to the fifth and final step.

Step 5: Make Recommendations

In many ways, this is the most important step in the BADIR framework. If you don't put what you've learned into action, you won't get the full benefits of doing your data analysis.

To do this, you need to present your recommendations to your audience in an actionable and engaging way, and drive it toward taking actions that resolve the business problem(s) you identified in Step 1. Short executive summaries that present your key findings, supported by detailed slides, are an effective way to do this. Remember that your goal is to show what your analysis has revealed, and to get the relevant key stakeholders' support for your plan.

So, our HR manager will likely recommend that his company reviews its salary payments, reconsiders its approach to work-life balance to help struggling employees, and investigates how female staff members could engage more with the company.

Big Data Versus Business Analytics?

The term "Big Data" is often applied as a catch-all phrase for how we both collect and use information.

Arguably, the terms "business analytics" or "business analysis" describe what we think of as Big Data more accurately.

The meaning of Big Data is evolving all the time and, as a general term for the whole process, it's just fine – semantics aside, people will know what you're talking about. But if you want to impress people with your insider knowledge, think of Big Data as a complex lock constructed from facts and figures, and business analytics as the key that unlocks its potential.

Key Points

Data pulls together huge amounts of facts and figures relevant to your business. When you analyze it in the right way, you can use this information to overcome problems or enhance your business performance.

Jain and Sharma's BADIR data analysis framework gives you a structured way of turning data into actionable decisions to improve your organization.

To get the best results from the business analysis element of the data process, follow the framework's five steps:

  1. Ask the right Business questions.
  2. Create an Analysis plan.
  3. Collect and validate the Data.
  4. Search for Insights.
  5. Make Recommendations.


Rate this resource