In-Depth
Analysis: BI Tools for Sustainability
Sustainability measurement is coming to BI tools, but slowly. Though a few vendors have products in development, only one has one for sale that can figure intricate carbon footprints across an enterprise or for a single product.
- By Ted Cuzzillo
- 10/15/2008
With all the talk about sustainability at this year's Oracle OpenWorld, held several weeks ago in San Francisco, I thought I might find BI products on the exhibit floor to manage energy consumption and "carbon footprints."
I asked BI vendors if there was anything afoot; if so, no one told the people at the booths. When I asked, most of them replied with some form of "Huh?" One man actually told me it couldn't be done. Collecting the data was too hard, he said.
Though it's good to be green, reducing a carbon footprint -- an organization's net total greenhouse gases produced, either directly or by suppliers -- is more complex than erecting a few windmills. New types of data -- such as energy consumed, manufacturing waste, and emissions -- have to be rolled up into a big picture the way only BI can do it. Without that picture, an organization's stabs at sustainability are like the Web without Google. "Green" remains just trees, not a forest.
Part of the BI-for-sustainability tool I hoped to find would work like any other business intelligence tool: analytics based on trustworthy data to enable a competitive edge. A second part, new to BI, would compare the company's performance to others using public benchmarking framework, such as from the Global Reporting Initiative. Although business people know how to judge financial data, few know how to judge units of carbon dioxide or kilowatt hours, especially as standards change.
BI is a natural solution, I believe, and experts I checked with agree. Dan Esty, co-author of Green to Gold: How Smart Companies Use Environmental Strategy to Innovate, Create Value, and Build Competitive Advantage (2006, Yale University) and the Hillhouse Professor of Environmental Law and Policy at Yale University said that companies are indeed looking to BI tools for an edge in sustainability.
"In this slumping economy," he said, "we've seen 'a number of companies deferring advertising campaigns and factories in favor of investment in BI.'" They've concluded that there is a substantial return on energy efficiency. Companies he's worked with, he said, have reduced energy expense by 10 to 15 percent, with some reaching 60 percent. BI tools can be adapted to help organize the effort.
Several companies -- including Oracle and Hewlett-Packard -- are developing tools to help. So far, though, only SAS has a product that does what I imagined such a product should do, and more, and is for sale right now.
SAS released it in April apparently with the most complex applications in mind. SAS for Sustainability Management, based on the SAS Enterprise Intelligence Platform, is a single platform for collecting, analyzing, and reporting any kind of data to calculate current and future impacts, whether environmental, social, or economic.
"There are lots of low hanging fruit that are no-brainers," said SAS marketing manager Alyssa Farrell, "but about when you get the next level? What to do? How do you spread out that capital expense and figure ROI?"
One potential customer, for example, wants to compare the carbon footprint of two products it manufactures. Each is made from the same materials, but each is produced in a different state and on different equipment. The company expects a major retailer to ask for this information, and if they can provide it before asked, they can increase sales. The information might soon be required by regulation, anyway.
To build a model to provide the comparison, the company must isolate several types of data for each product at each of the two plants and the fabrication machine used to make them. Factors include the power allocation for each machine and emissions from source materials. They also have to factor in each plant's source of electricity, since one state's grid gets power from a nuclear plant and the other's source is a coal plant. For this data, SAS maintains a database based on data from the federal Environmental Protection Agency and International Energy Agency.
One manufacturer of high-tech equipment wanted a better grasp of two key performance indicators: enterprise-wide waste in manufacturing and energy consumption. Estimates of waste help spot problems and spur reuse.
Data is collected manually in online surveys -- a feature of the SAS product -- that are issued to production line managers. "That did take quite a bit of manual data collection," Farrell admitted.
There's that problem again: collecting non-financial data. It's a problem for most companies, according to Will Sarni, CEO of Domani, a consulting firm. Even as companies have done a good job of rolling up their financial data, they've done a poor job of collecting energy data.
"You'd be surprised to see how archaic it is for some companies," he said. "In some ways, we're moving out of the Stone Age to find that data." There's a big scramble now for tools that can be used easily within an organization to collect and understand (and make rational decisions about) how to reduce waste packaging and reduce certain types of material from processing and packaging. What's missing most, he says, is a way to monitor companywide energy consumption in real time. Energy expense is becoming important in smaller and smaller companies. It used to be that an annual expense of less than $1 million was considered unimportant. He said, "Now it's significant."
The threshold for concern hasn't yet hit mid-size organizations, reports a provider of BI tools to that market. Joni Girardi, CEO of DataSelf, said, "You know how mid-size companies monitor energy efficiency? By making sure everyone turns out the lights when they leave."
I predict that one morning they'll turn on the lights and realize they need to pay more attention. If they're big enough to analyze income with a BI tool, they're big enough to analyze the carbon footprint. Nowadays, it all counts.