Over the years, John Thielman has managed dozens of data warehouse and business intelligence projects for a variety of companies, from manufacturers to major retail chains. While he's measured plenty of soft benefits around the various business intelligence (BI) projects he's been involved in, he's been hard-pressed to pull together a formal return on investment (ROI) calculation.
His situation is no aberration. "I challenge anyone who talks about a hard and fast ROI on anything to do with business intelligence," said Thielman, currently an independent data warehouse project manager in the Dallas area.
"ROI from BI doesn't come from producing the reports quicker or producing the information quicker," he continued. "The real value of BI in manufacturing is arrived at with the common proactive decision-making based on information that is uniformly defined and calculated across the organization. It's an intangible value that you can't always estimate."
Struggle to calculate BI ROI metrics
According to experts in the data warehouse arena, Thielman has lots of company in his struggle to build a formal ROI case for BI projects. BI, which refers to a broad category of tools to help enterprise users make better business decisions, encompasses everything from traditional query and reporting, forecasting, decision support and data mining applications to newer offerings such as dashboards and portals.
Typically used as part of a data warehouse initiative or in conjunction with operational enterprise platforms like ERP or CRM, these BI tools organize and present data about the business, often in real time, to users who can apply it in their day-to-day decision-making.
Since BI is typically deployed for long-range strategy and competitive trends and planning, nailing down formal ROI metrics has proved elusive. Despite that lack of financial validation, companies continued to eagerly fund BI implementations because they realized there was intrinsic value in the solution.
"How do you put a monetary value on better decision-making?" asked Claudia Imhoff, president and founder of Intelligent Solutions, a data warehouse consultancy. "It's so intuitively obvious that you need this capability, [companies] forget about ROI."
However, with BI tools increasingly being tapped for more tactical applications and with the recession putting pressure on companies to reduce IT expenditures, companies can no longer afford the luxury of intuitively knowing they need BI.
In its report 2009 State of the Market: Mid-Year Insights, market research firm Aberdeen Group found that more companies are now actively measuring the ROI on BI investments. The report refers to a survey in which 42% of respondents selected sales performance metrics as their top target, followed by cost savings metrics, customer performance metrics and process efficiency metrics.
Identify business goals before implementing BI in manufacturing
But before putting those targets in sight, there is work to do to ensure that the ROI exercise is fruitful. According to Imhoff, the first steps are to define the business goals of a BI implementation and to have some measure of the current state prior to deployment so there is a benchmark for comparison.
It also may make sense, she adds, to tie the BI rollout to a corporate-wide initiative -- for example, reducing customer churn -- so that there is a clear connection between the use of BI and any progress the company makes toward its goals.
Tallying up BI costs is relatively simple -- the usual expenses around hardware, software licenses, dedicated personnel resources, training, etc. It's much harder to pinpoint the benefits. Hence the need to start with small, targeted BI projects that can show clear results.
"To get funded, look for things that can get you over the hurdle of minimal ROI," said William McKnight, president of McKnight Consulting Group, which specializes in data warehousing. BI projects that can prove out some sort of expense reduction for the organization -- reducing fraud, cutting back shipping or printing costs, or decreasing the number of customer returns -- are a good place to start, he said, rather than trying to establish value on a massive BI deployment.
"Identify something of high importance and keep your affinity to that outcome throughout the duration of the project," McKnight said. "Otherwise, BI can get lost in the shuffle."
User surveys aid in calculating ROI of business intelligence
Opportunity cost is typically the most problematic ROI metric, but it is often the area where BI has the most impact. Access to previously unavailable or hard-to-interpret data can be the linchpin for giving a company a competitive edge in seeking out new markets or reaching customers in a different way.
To identify these and other hard-to-prove BI ROI benefits, it's critical to connect with the operational areas of the business and ask users -- through surveys or one-on-one interviews -- about their problems and where data can help, said Mark Madsen, president of Third Nature, a data warehouse and BI consulting company.
"Find out what problems could have been prevented if they had information or what could have been done differently," Madsen said. "Those things are key elements to getting at the underlying ROI of BI."
About the author: Beth Stackpole is a freelance writer who has covered manufacturing techniques and manufacturing technology extensively.