Earlier this month, a member of the iSixSigma Discussion Forum asked about the best methods to measure financial performance among business units within the same company, but in different regions. The answers that came back ranged from simple bottom-line comparison using accounting department data to more complex metrics, such as profit per item sold and cost per unit. Regardless of the methods implemented, the most common advice was to understand the context of the data and to factor in what the practitioner wants to do with the data.

Sept. 7, 2010

michsigma: I am looking to compare the financial performance of a dozen or more business units across the country. Fundamentally, they are all selling the same products… I would like to know if there is really any difference between the performance of the units. I am trying to move the needle here from raw numbers and “gut feel” to data feel. Has anyone set up a template for something like this or can offer some recommendations on how to proceed?

Sept. 8, 2010

Jsev607: At times when I’ve done this, I start with the bottom line, using a multiple comparison procedure, and then work my way backward through specific areas (comparison of sales, costs, warranty claims, etc.) to explain the differences in the bottom-line performance. It is important to partner with accounting and your leadership early in the process so you can understand just what your organizations definition of “financial performance” is, though.

Also keep in mind that numbers and hypothesis tests won’t give you the whole story. Data has no meaning apart from its context, so while two sites may be selling product A, if they are selling in completely different socioeconomic environments, then you cannot reasonably expect them to sell at the same rate.

nmudd: There are so many factors that could differ from region to region or even city to city, such as city/state taxes, market size, affluence of population, employee wages (a.k.a. overhead) and transportation costs. My advice would be to find a common factor or factors that could be used to normalize the data in some way so you can come as close as possible to comparing apples to apples.

One possible metric would be “profit per item sold or produced.” While this might be far from perfect, it does attempt to capture multiple factors in a single metric (sales volume, raw material costs, warranty claims, etc.). This should help you pick out the top-performing unit on paper, but then you have to dig further to understand why they are performing so well and what that means for the whole company.

[For example], one unit might be the most profitable in the company, but does that matter if they only account for 1 percent of total revenue? Most importantly (and just as Jsev607 said), you must consider the context of the data as well. While there are some out there who may not agree with him, I think that Don Wheeler frames a pretty good argument in his book “Understanding Variation.” Whether you agree or disagree, it is definitely worth the read to understand his arguments. Then you can draw your own conclusions.

Sept. 9, 2010

CFB: My question would first be: What do you intend to do with the data once you have it? It would appear that you want to determine how well a given strategic business unit is performing relative to their peer group. If this is the case, why couldn’t you simply pull the last three to five years’ worth of financials and conduct a fairly straightforward ratio analysis to look at things like profitability, liquidity, asset utilization, etc.? And why doesn’t your finance/accounting area already have this information???

A simple ratio analysis to normalize your finanicals (e.g., dividing through by “sales” on gross, operational and net income to create unitless margins) allows you to compare each organization against itself over time, to industry leaders and against each other – to include trending. Again, you need to define what you need to know and understand how that information is going to be actionable to you before you can decide what kind of data collection and subsequent analysis/modeling to engage in.

Personally, I would use [your finance department] to help identify the one organization that management thinks is the biggest underperformer and then work with [finance] to build a model on what drives value for this organization – sales growth, capital management, price strength, operational effectiveness, etc. – and then identify where the continuous process improvement opportunities are to generate wealth.

[For example], pick a target business, gather the last five years’ worth of financials, identify key drivers of financial performance based on the magnitude and direction of their effects on cash flows, establish these key drivers as primary dashboard metrics, and then start building a pipeline of actionable improvement efforts that will move these needles. This theory of knowledge would be a tremendous help to your management team and make Lean Six Sigma project assessment and selection much more effective.

gurucom2006: Whatever factors you consider to compare the performance, please don’t use money as a factor. …I say this [because] the rate of currency arises from time to time. Carry out the analysis in a segment approach, which gives you the real result.

Daniel Santillanes: If all sites have the same products, a very good comparison is to measure cost per unit. [This] is a measure that can help you compare sites and analyze the elements that mainly affect this measure [across] the different sites. Things like absorption, volume, materials cost, etc., will help you compare…who has the best suppliers for key raw materials, who has the least taxes, who has the best capacity utilization, etc.?

This will help your company compare itself between sites [and] understand their differences, but also decide where opportunities [are] for assigning more volume. One of the main goals of every organization in to simplify the supply chain and minimize its complexity. This comparison between sites will help them focus on the appropriate things to work [on] to remain competitive within their own network and to create partnerships between sites.

It’s not about fairness or justifying particular requirements of each site. …If a site has more requirements than others and it’s cost per unit is greater, [then it] has to do more than everyone else to remain competitive.

To add your two cents to this or any other online discussion, please visit iSixSigma’s Discussion Forum.