In the 2004 treatise Moneyball, The Art of Winning an Unfair Game, Michael Lewis tells the the story of General Manager Billy Beane and the Oakland Athletics rise to the top through the use of an Apple II based database stats package. Although bigger market teams had all the advantages - bigger payrolls, glamorous ballparks, historic baseball rich towns, Oakland was able to re-write the record books, winning more games with lower payrolls than any other team in the majors.
Baseball is a notoriously statistics driven industry - yet Oakland realized most traditional baseball stats were designed measure potential to be great - as in Runs Batted In (RBI’s), while other, less common metrics like On-Base Percentage (OBP) and Slugging Percentage (SB) measured actual player performance. While other teams spent heavily on established stars with great RBI numbers, Oakland acquired up and coming minor league SB and OBP leaders - and revolutionized baseball in the process.
Other teams followed, the Boston Red Sox, after years of hiring the best free-agents available (which all teams try to accomplish) finally hired the best analytical talent. The New England Patriots bring a strong analytical approach to a record four Super Bowl appearances in the NFL.
And how does this relate to your business?
You basically need only three things to compete with analytics - accurate measurement statistics, quantitative analysis and fact-based management.
Or more simply put, 1) data access, 2) reporting and 3) analytics.
In a small business, the owner or Primary Manager can be handle all three tasks: 1) He’s a front line source of data - merely observing the business as it happens, 2) Reporting - either by compiling what he’s seen because he has first hand knowledge of business process, observed time and time again, and 3) Analytics - analyzing and determining better ways of operating.
As businesses grow, homegrown solutions such as reports and spreadsheets allow various managers to replicate what the Primary Manager accomplished. The fault of this model is it’s inability to scale. Growth of departments and specialization means different managers are compartmentalized, with different measurements, varied goals and over time, the dreaded information silos are created - where purchasing has different reports than sales, and manufacturing doesn’t know what inventory has - and if anyone wanted to do some real analysis on the numbers, the first question is, who’s numbers and the second question is how do we know those particular datasets are correct?
In a larger business, ERP (Enterprise Resource Planning) and BI (Business Intelligence) serve these functions. ERP deploys information gathering to multiple departments and eliminates information silos. For example, if manufacturing records completion of 25 pallets of finished goods, the warehouse manager has to record 25 pallets being transferred into inventory - if there’s a discrepancy, it’s noticed immediately and corrected at the source. Information silos are destroyed. The company operates a single set of metrics. And what’s more, we can now see the effect of interdepartmental decision making - sure we can eliminate stock outs by over-ordering inventory, only now, we’ll see the cost increase in finance immediately - instead of waiting for accounting data some weeks and months down the line.
BI or Business Intelligence
There’s a Harvard Business School Press book called Competing on Analytics, by Thomas Davenport and Jeanne Harris that studies how companies create sustainable competitive advantages by crunching numbers - not suprisingly, they find that 65% of the high performing companies have real-time analytical capabilities - and actually rely on that data for decision support.
Pricing, customer support, logistics, manufacturing - all can be improved with the application of real time data. In some companies, for example, Netflix, the customers who rent the fewest movies earn the quickest mail replacement service - why? With flat fee pricing, the customer who spends $25 a month to keep 5 movies at home and only watches 3 a month is far more profitable than the customer who pays the same $25 fee, but watches 5 movies every weekend for a total of 20 a month. Counter-intuitive, but quickly apparent when looking at the analytics.
In our practice, food manufacturers face a difficult task. Food costs (raw materials) fluctuate on everything from the pork bellies futures market to the costs of corn escalating due to ethanol production. Then transportation costs factor delivery sizes, distances, frequencies of delivery, not to mention skyrocketing fuel pricing. Food distribution involves brokers, broker agents, commissions, shelf allotment fees. There’s spoilage and shelf-life issues, then charge-offs and returns.
Given this complexity, most food manufacturers know that they make money at the end of the month - but that often means losing money on 5-10% of the contracts or customers they serve, and sometimes losing money on certain products they manufacture.
With our established ERP customers, we often demo our Business Intelligence modules (EPM - for Enterprise Performance Management) by loading the software the night before the demo, extracting data and populating the BI data warehouse with live information. A couple of years back, we opened a demo at a midwest food manufacturing company by bringing on screen a dashboard showing their most profitable customers, their most profitable products. It seemed to be going well until the CEO rose and asked us to stop, and please leave the room for a few minutes. We were quickly ushered out while the CEO and management team examined the data, realized three production runs that were scheduled to be produced that week would lose money on a customer who’s return rates were legendarily high and was one of those customers that everyone in the company dreaded dealing with. As soon as the shop floor schedule was changed, with those particular runs cancelled, we were allowed to return and complete the demo.
But what we find, more often than not - is that using BI we will identify 5-10% of a company’s production that, if not actually money losing, falls far below established revenue targets. For a growing company to realize they can outsource these problem clients to their competitors (by firing the customer) and then realize without spending another dime on equipment, labor, or raw materials, the can have another 5-10% of plant capacity overnight - the impact can be huge.
Try it - take your annual revenue derived from manufacturing X 5% - and realize that's the number you can add to your bottom line without adding labor, materials or equipment. Makes BI look pretty attractive.
You can contact Gene via GH@GeneHammons.com.
Trends Fads Bubbles and Bashing. Take a couple hundred Geeks and give them typewriters, you'll never get Shakespeare, but what you will get is the Technology Press - here's where we sort it all out to see if there's any 'there' there.