First Thoughts
  By Dan Gilmore - Editor-in-Chief  
  March 12, 2009  
  Is The Perfect Order the Perfect Supply Chain Metric?  

It’s hard to believe, but in the 5 years since we launched Supply Chain Digest, I don’t believe I have ever done a piece on the “Perfect Order.” We've done several news stories and even a Videocast on the subject (view Achieving the Perfect Order on-demand), but never a First Thoughts column.

The term Perfect Order came out of the grocery industry many years ago, and while generally recognized by most supply chain professionals, it can mean different things to different companies and supply chains.

The most common definition, I believe, is that an order is “perfect” if it is:

  • On-time
  • Complete
  • Damage Free
  • Correctly “documented”

There are several caveats or gray areas, however.

Gilmore Says:

One key question obviously is: just how perfect do you want to be? There is a cost to be perfect, and above a given level of perfectness, it actually starts to reduce profit. Where do you draw the line?

Click Here to See
Reader Feedback

First, many companies use other definitions. Several years when I was in the supply chain software world, we did a survey of our customers and found something like 11 different metrics in total that companies used to calculate Perfect Order. I believe the Grocery Manufacturers Association has suggested recently the use of seven different attributes.

There are some important questions. For example, some use whether the order was shipped on-time rather than delivered on-time in their calculations. While we could say this is an out-of-date view of things, the reality is that there are many things often outside of the shipper's control – like not being able to secure a appointment at the retailer’s DC consistent with the stated delivery date – that can really mess up the metric.

At one point, Hershey Foods only considered an order “perfect” (in addition to the other attributes) if it was received/entered into its order management system once and then not touched again. I am not sure if they are still looking at Perfect Order that way, but it played havoc on the achievement scores, as you can imagine.

I have also heard many companies that have problems with their “on-time” metric due to delays in delivery notifications from carriers. So, if you look at Perfect Order performance weekly, as some do, last week’s numbers often look not so good because confirmation still hasn’t been received from the carrier. Sometimes they have to spend resources tracking it down to avoid an inaccurate Perfect Order “miss.”

“Documentation” is also an issue. First, documentation can mean many things, including invoices, shipping manifests, bar code labeling, carton contents labeling, advanced ship notices, and even now RFID tagging, I suppose. Of course, customer requirements for these and other “documents” vary dramatically between industries and individual customers.

I did some work a few years back with one of the major beer companies which was trying to benchmark its EDI success with advance ship notices. It wanted to compare how well it matched up with others in terms of having ASNs that: (1) were actually sent by suppliers; (2) showed up before the shipment itself arrived (many of its suppliers were very close by physically); and (3) which matched the shipment and invoice exactly.

My memory is their number just for this one element of a supplier perfect order was about 50% and, in a sense, they had a rather simple scenario - many of their shipments were full truckloads of just cans, for example, making the ASN easier to construct than it would be for many others. Later, they found that many of the issues with the ASN not being received before the shipment were due to batch processing and other delays within its own IT system. So, as a supplier, how do you deal with that in measuring your perfect order performance?

The great challenge, of course, in achieving the Perfect Order is its binary nature. In theory, if you were supposed to ship 10,000 widgets and you can only ship 9,999, the order isn’t perfect – so you get a big Zero for that particular shipment.

So, the chances of actually achieving a Perfect Order can be very small. Our blogger and well known supply chain performance management guru Kate Vitasek of Supply Chain Visions and our friend Dr. Karl Manrodt of Georgia Southern University last year literally analyzed thousands of orders using real data from the “compliance tracking systems” for three major retailers. The results were put together in an excellent report, in which they calculate something they call a “Perfect Order Index.” The POI is calculated by multiplying the percent of orders that were on-time by the percent complete by the percent documented correctly. Undamaged was calculated at 100% (obviously not reality) because that data couldn’t really be obtained. Report is available from SCDigest: Perfect Order Report.

Even using that, the Perfect Order Index for these three retailers and their suppliers came in at just 27%. The good news is that was up from 19% in the same study two years before that. The actual true “Perfect Order” percentage you would have to assume was much lower than the index, given the binary nature across all four categories.

Perfect Order has a bit of a consumer goods-to-retail connotation, though many other companies and industries (such as pharma) use it as well. In some industries, the related term “On-Time and In-Full” (OTIF) is used, though it generally does not have the same scope as Perfect Order (labeling, EDI, etc.). Several years ago, for example, industrial giant 3M launched a big program to improve OTIF company-wide after an analysis showed a very high percentage of backordered product. The improvement program was linked to the company’s aggressive Lean Six Sigma efforts.

So, what’s the point of the Perfect Order?

Many companies certainly use the metric. One key question obviously is: just how perfect do you want to be? There is a cost to be perfect, and above a given level of perfectness, it actually starts to reduce profit. Where do you draw the line? That’s part of the real rub. You certainly could (and some do) model those trade-offs (including estimates of the revenue and margin lift from increasing perfectness of service). Some companies simply strive for higher levels of service than financially optimal in the short term in the hope of long-term customer satisfaction and market share.

Several years ago, AMR Research put Perfect Order as one of the top three measures in its hierarchy of supply chain metrics, along with forecast accuracy and supply chain costs. That probably stemmed from research AMR did a few years ago that found a 3 percent improvement in perfect order fulfillment translated to a 1 percent increase in profit margin, while a 10 percent increase meant an additional 50 cents in earnings per share.

Is that right? I am not so sure. As we all know, association does not equal causation. A subtle but important point is that it could just be that companies with excellent supply chains both enjoy better financials and higher perfect order scores.

I also think that from a financial perspective, the “on-time, in-full” view is probably a bigger driver of financial results than a broader definition of Perfect Order that includes labeling, ASNs, etc. Not that those things aren’t important, especially to some customers (e.g., retailers), but, in other cases, may be more ancillary in considering overall supply chain excellence.

Recently, Forrester Research also did a take on “Perfect Order” from a whole other perspective – how the customer views the experience in this multi-channel retail and distribution world. That was very interesting too.

I am out of space, but we’ll pick this back up again in a few weeks. Until then, would love to know your thoughts on Perfect Order as a metric and what you use to calculate it.

How important is the “Perfect Order” or OTIF to your supply chain? What do you use or think the definition should be? How “perfect” should you really strive to be?

  Send an Email