One of the most important questions with regards to Big Data is, of course, whether companies are seeing a return on big data investments. As we mentioned earlier, one of our primary research goals was to begin to assess the return on investment on Big Data.
We asked survey respondents1 to tell us what return they expected in 2012 on their Big Data investments2. To boost the number of people who would respond to this question and increase the chances that they calculated ROI the same way, we provided a range of returns with bands (e.g., “26% to 50%”), and furnished a formula for calculating percentage ROI:
Percentage ROI = (Gain from investment – Cost of investment )* 100/ Cost of investment
Here’s what we found on the ROI front: Only 8% said they expected a negative return, and 16% said they didn’t know what the return would be. That could be read to mean that in about one-quarter of the companies, the return on Big Data was either negative or questionable. A larger percentage of respondents (33%) said the return would be low (between 0% and 25%). However, 20% expected the ROI to be between 26% and 50%. And nearly one-quarter (23%) projected a 2012 return of more than 50%. (See Exhibit II-11)
Exhibit II-11: Across Regions – The ROI on Big Data
Q17: Company’s Expected Return in 2012 on Big Data Investments
Using the above data , we then calculated the mean ROI overall across all regions of the world that we surveyed, as well as for each region.3 Of those companies that reported ROI numbers (even if they were negative):
- The average ROI was 46% across all four regions
- Companies in Asia-Pacific (71% ROI) and Latin America (64% ROI) reported greater ROI than their counterparts in the U.S. (37%) and Europe (43%)
So what would an average 46% ROI be on, say, a $2 million investment in Big Data? It would mean that a company had a $2.9 million increase in revenue (or decrease in cost), but a $900,000 return after subtracting the cost of the investment.4 That appears to be a decent return, and one likely to exceed many companies’ internal ’hurdle’ rates for investments. In addition, based on our 12 interviews with best practice companies, this 46% mean ROI actually appears to be a very modest return. We heard about ROIs on Big Data of 25 to 1 or even more from senior executives.
For example, an analytics director in one large Internet company indicated his group has helped increase firm revenue by hundreds of millions of dollars over the last five years. An executive at a major telecommunications company (which spends hundreds of millions of dollars annually on marketing) said analytics have boosted sales on promotional programs from 2% to 10%. After using analytics to determine what offer it should make to customers whose contracts were expiring, the company reduced the churn it could control by 20-30 basis points, a significant amount. And the analytics work also boosted sales of one product line to consumers by 10% to 15%, this executive estimated.
In a $2 billion Internet company, an analytics director says the ROI has been at least tenfold. “Big data has changed the way we do business fundamentally. We have a very big member base, which buys products that we create for them,” he told us. “Analytics have helped us decide what products to build to serve our base. It’s enabled us to segment them and understand what they need.” The company appreciates the potential of the analytics team, which has grown from one person to 30 in two years. “Our CEO and the executive team understand the value we bring to the table,” the executive said.
Exhibit II-12: Mean Projected ROI on Big Data by Regions of the World
Q17: Mean Percentage of Expected Return in 2012 on Big Data Investments by Region
For a detailed breakout of ROI by region and by the ’bands’ of ROI we used, see Exhibit II-13.
Exhibit II-13: Detailed Breakdown of Mean ROI on Big Data by Regions of the World
Q17: Companies’ Expected Return in 2012 on their Big Data Investments – By Region
« Previous Post
What Kinds of Digital Data are Companies Using?
- For this calculation of Big Data ROI across a company, we asked the survey respondents who were either in the IT group or an analytics group, believing that since they supported the larger organization (and not just one business function), they would have a better view on overall enterprise ROI. When we present the data in our section on business functions, the managers of those functions calculated ROI numbers for their function. [↩]
- Note: Since we began fielding the survey in December 2012, we consider these ROI ’expectations’ more reliable indicators of real returns than if we had fielded the survey at the beginning of 2012 and asked companies to project their ROIs for 2012. [↩]
- We then calculated mean ROI for each region by using the halfway points in our ranges as the multiplier. [↩]
- This data is self-reported, meaning that we haven’t verified the ROI. [↩]