Big Data

10 ways to optimize business analytics in your organization

Business analytics enables companies to make better decisions and become more competitive -- but only if it's approached the right way.

Companies believe that business analytics will deliver competitive advantage -- but what analytics delivers will in large part depend upon how insightful organizations are in using BA tools. Here is a list of 10 best practices to help you optimize business analytics in your organization.

1: Collaborate to identify the "hard" questions you want answered for your business with analytics

If you are going to profit from using analytics in your business, you have to find ways to learn about your customers and the market that exceed those of your competitors. This means formulating insightful questions that analytics can bring answers to. One way to approach this is by getting a strong interdisciplinary team together to review the business and to lay out all the questions that have traditionally been hard or elusive to answer. If you know the questions you want to ask before you talk to analytics vendors, you will be in a good position to see whether vendor solutions can offer you answers that your traditional reporting tools haven't been able to provide.

2: Avoid the initial temptation of just converting your top 20 reports when you move to a business analytics solution

You'd be amazed at how many analytics vendors report that the first thing new customers want them to do is to convert the top 20 existing company reports into reports that run on the new analytics solution! Instead, the goal should be to create new reports that analyze information in different ways that will bring your company greater competitive and operational advantage. The tough sell here is getting people to let go of old reports that they have become comfortable with over the years.

3: Plan for the impact that running real-time data analytics is likely to have on your data center

It's no longer business as usual for the data center once you start running business analytics. If you're a retailer and you want to respond with instant promotions based on what your customers are buying online now, you're going to need real-time analytics that gives you up-to-the-minute shopping results. This means that IT resource usage and operating procedures have to be modified so both online transaction processing and real-time batch processing (in the form of analytics) can be run concurrently and at high priority. Traditionally, IT has run transaction processing at high priority during the day, and has waited until evening to batch process. This strategy doesn't work with real-time analytics.

4: Make sure you have the in-house talent to design and run your business analytics

This is a major problem for most companies, because you can't do much with a business analytics solution if you don't know which questions to ask. Add to this the reality that the people who have traditionally designed batch reports often find that business analytics demands more critical thinking than just summarizing the month's sales results and comparing them to the same month last year. This situation has created high demand in the IT job market for business analytics specialists. It is also why many business analytics vendors offer consulting and training to help organizations develop business analytics skills. Be sure to ask your vendor candidates about their capabilities in these areas.

5: Don't forget to revise your company operations to incorporate real-time analytics

Financial companies have done this for years. If an unusual buying pattern emerges for a credit card, analytics software tells staff to put a freeze on the card and immediately notify the cardholder. This same automated intervention of analytics can work in other company areas -- but it means revising workflows in operational areas of the company so they can take advantage of the analytics. These operational revisions (and potentially employee retraining) should be tasks included in any business analytics implementation plan.

6: Take advantage of your vendor's best practices

Virtually every vendor has a set of best practices for its solutions. These best practices have been gleaned from experience the vendor already has with enterprises using its products. Take advantage of these best practices -- and incorporate the ones that make good sense for your business in your own practices.

7: Leverage Excel spreadsheets

I have yet to talk to a business analytics vendor that doesn't incorporate Excel spreadsheets as an option in its report-building process. Vendors recognize that these spreadsheets are a de facto reporting standard in most companies and that little training is required for staff to use them effectively.

8: Design analytics that deliver both summary and drill-down capability

Great business analytics functions on two levels: It gives a summary or dashboard view of activity that C-level and other top managers can look at, and it presents drill-down capability for those who want to take a closer look at the data and what is behind it. BA reports should contain both top-level and drill-down capabilities when you build them.

9: Eliminate junk data

We all are familiar with the slogan "Garbage in, garbage out." Business analytics doesn't change this. The best approach before implementing business analytics is to ensure the accuracy and integrity of the data that those analytics will be operating on. This can be done by running a "data cleaning" project that gets rid of inaccurate, duplicative, and incomplete data before you bring in BA.

10: Don't let your analytics do all the thinking for you

Business analytics is a tool that enables companies to "get smarter" about how they do business so they can avoid seat-of-the-pants decisions. The power of analytics rests in its ability to process reams of data that normal computing and human beings just can't get to. But data (and analytics) also have their limits. There will be times when instinct is the purest form of intelligence.


Mary E. Shacklett is president of Transworld Data, a technology research and market development firm. Prior to founding the company, Mary was Senior Vice President of Marketing and Technology at TCCU, Inc., a financial services firm; Vice President o...


It may seem that many of these are common sense applications to BA, however in my experience anyway, it seems that management doesn't think this way. I have seen some very intelligent peopole who are very capable managers operate with the mindset regarding item 2 on the list- in many cases even though it may not be the best solution. I've thought about this exact issue and have tried to understand the thought process which allows me to provide persuasive, logical and tangible reasons as to modify reporting, and what I think may be the greatest driver of managers not wanting to explore other reports or look at different metrics is because they are used to looking at certain information and they simply do not have time to familiarize themselves with new reporting. When it comes to analytics, it is my opinion that number one is likely one of the most important issues - which ties into number 9 - and I couldn't have stated it better. I think one of the most important aspects of Business Analysis is to determine what information is important, and I appreciate the way the author explains it in the article. Then, once the questions about what it is the organization is trying to get from the uncountable Terrabytes of data that is available and having it organized and presented in an understandable way, number 10 on the list is also key as understanding why the results are what they are and sometimes that means you have to turn off the cruise control and find out what is driving certain trends, what may have caused the deviation of the trend or anything that is out of the ordinary. These are just the top 10, I'd be interested in seeing a follow-up post to find out what the Author thinks is the next 10. I noticed that Excel is mentioned, and I agree. But I also use Access almost as much because Access can do much more than Excel - and when used together properly, I have seen exceptional analytical outcomes. Thank you Ms. Shacklett for the great information.

Editor's Picks