Do You Need All That Data?

Just in case we get lost in over-analyzing everything, including customer data, Ron Ashkenas suggested that we step back and think about what is really useful in a post for the Harvard Business Review.

Organizations love data: numbers, reports, trend lines, graphs, spreadsheets — the more the better. And, as a result, many organizations have a substantial internal factory that churns out data on a regular basis, as well as external resources on call that produce data for onetime studies and questions. But what’s the evidence that all of this data is worth the cost and indeed leads to better business decisions? Is some amount of data collection unnecessary, perhaps even damaging by creating complexity and confusion?

For many years the CEO of a premier consumer products company insisted on a monthly business review process that was highly data-intensive. At its core was a “book” that contained cost and sales data for every product sold in the company, broken down by business unit, channel, geography, and consumer segment. This book (available electronically but always printed by the executive team) was several inches thick. It was produced each month by many hundreds of finance, product management, and information technology people who spent thousands of hours collecting, assessing, analyzing, reconciling, and sorting the data.

Since this was the CEO’s way of running the business, no one really questioned whether all of this activity really was worth it, although many complained about the time required. When a new CEO came on the scene a he decided that the business would do just fine with quarterly reviews and exception-only reporting. Suddenly the entire data-production industry of this company was reduced substantially — and the company didn’t miss a beat.

Obviously different CEO’s have different needs for data. Some want their decisions to be based on as much hard data as possible; others want just enough data to either reinforce or challenge their intuition; and still others may prefer a combination of hard, analytical data with anecdotal and qualitative input. These preferences at the top of the company often influence the “data culture” that is created. In all cases, though, managers would do well to ask themselves four questions about their data process as a way of improving the return on what is often a substantial (but not always visible) investment:

  1. Are we asking the right questions? Many companies collect the data that is available, rather than the data that is needed to help make decisions and run the business. So the starting point for simplifying and improving data processes is to be clear about a limited number of key questions that you want the data to help you answer — and then focus the data collection around those rather than everything else that is possible.
  2. Does our data tell a story? Most data comes in fragments. To be useful, these individual bits of information need to be put together into a coherent explanation of the business situation, which means integrating data into a “story”. While “enterprise data systems” have been useful in driving consistent data definitions so that things can be added and compared, they don’t automatically create the story. Instead, managers should consider in advance what data is needed to convey the story that they will be required to tell.
  3. Does our data help us look ahead rather than behind? Most of the data that is collected in companies tells managers how they performed in a past period — but is less effective in predicting future performance. Therefore it is important to ask what data, at what time frames, will help us get ahead of the curve instead of just reacting.
  4. Do we have a good mix of quantitative and qualitative data? Neither quantitative nor qualitative data tells the whole story. For example, to make good product and pricing decisions, we need to know not only what is being sold to whom, but also why some products are selling more than others.

Clearly business data and its analysis are critical for organizations to succeed — which is underscored by the fact that companies like IBM are investing billions of dollars in acquisitions in the business intelligence and analytics space. But even the best automated tools won’t be effective unless managers are clear about the questions raised above.

What’s your assessment of data in your company? Is there anything we can do to help you make sense of what you have?

Leave a Reply

Your email address will not be published. Required fields are marked *