The 1% Rule

Years ago when I was just starting out as an inventory counter with RGIS, I somehow ended up sitting shotgun to my district manager on a way back from a inventory with a van full of counters sitting behind us. The trip home was a long hour and 45 minutes trip mainly caused by the distance. During this trip my DM got to talking about all the different clients he's had to put up with in all these various stores. All of his accounts were extremely one sided tales, containing some level of contempt for the people he described "You've done 20 inventories, well that's a good week for me" was one memorable line. In response to one client who questioned the accuracy of his crew's work, he said as an aside to me and to anyone else who was listening "You know as long as it's within 1%, it's okay".

Eight and half years later, I was in a meeting with my boss and several other co-workers
from Quantum. The topic of counting candy came up, my boss mentioned several techinques for counting it, none of which included actually taking the candy out of the box and physically counting it, and at some point someone brought up the point that as long as it's within 1% it okay. This of course was followed up with a rather obvious condition, "but you don't ever say that to the client". To the question of what divides unauthorized from authorized estimation, here exists an unwritten standard to seperate the two concepts, or perhaps to better define what one of these concepts is. The 1% rule, as the above stories may hint at, seems used more to identify what is accepted than what is isn't. There are some questions to consider with the 1% rule.

1% of what would be the first obvious question. Whenever someone says "within 1%", it's not always clear what the 1% is being taken of. 1% of the total inventory? I've been to grocery stores that carry over $1,000,000 in inventory, 1% of that would be $10,000. For a c-store that has $150,000 in total inventory, 1% of their total inventory is $1,500, which could represent the same total as a small gondola. Being off by less than 1% for the total inventory can offer a lot of wiggle room of counting errors. A total inventory count for a grocery that's off by by $9,000 can still be within 1% for the entire store, a c-store count that's off by $1,400 can still be within 1% of the store's total inventory. There probably exists a lot of store managers who would find these variances too large in spite of what the overall percentage is. Perhaps the 1% rule gets applies to individual sections of the store, instead of the entire store. When applied to smaller portions of the store the notion of 1% can become far smaller. For a box that contains 36 candy bars, 1% is roughly a third of a candy bar. If a shelf had 10 boxes containing 36 candy bars and all were the same price, then counting that shelf within 1% would allow us to be off by 3 candy bars total. And of course that variance might be smaller if some of those boxes were not completely full. If a whole side of the gondola had let's say eight shelves like this, than for the entire side being off by less than 1%, will, conservatively speaking, mean being off by no more than 24 bars, still less than a full box. Of course how the the variance for sections gets used in relation to the whole inventory is still interesting to consider. I could be off by more then 1% for one section, but if I nail enough sections elsewhere I could still be within 1% for the entire store. Being within 1% for the entire store doesn't guarantee that every section carries that same level of accuracy. I've also thought that 1% is in reference to the store's level of shrink, but a counter in most instances would have no idea about the resultant shrink percenatge a certain count produces, so this wouldn't make much sense at all. I've also considered the notion of 1% of the stated book value. But once again a counter usually won't (or better yet shouldn't) know what the book values are, so this wouldn't make much sense either. Generally the 1% rule seems to be in reference to the actual value of the inventory that truly exists in the store. How this gets used, or how deeply it gets enforced is still not quite clear.

Another question would be why 1%? Why not 2%? or 3%? or even 5%? There probably isn't a satisfiable answer to these questions. But if the notion of a perfect 100% accurate count is too lofty of an ideal to actually achieve, than a line definitely has to be drawn somewhere, and it has to be drawn in order to deliver a high level of accuracy, while still allowing room for "ethical estimation". In an essay called "What causes Inventory Auditing Fraud?" written by Carl Jackson and Jack Henry that can be found on the NAAIS website under the "History of Inventory Services" link (an interesting read if I don't say so myself) they briefly refer to the concept of ethical estimating, saying "these types of estimates are controlled and do not affect accuracy". They proceed to give 2 concrete examples of when it's okay to estimate but offer no such examples of when it shouldn't be done. Addding that such measures wouldn't affect the "tolerance of accuracy required for a good count". Okay fine, counters don't have to be perfect, but how imperfect are they allowed to be? Jackson/ Henry don't mention anything about being within a certain percentage, but from my experience levels of tolerance can change from client to client, and from manager to manager. Some store managers will have a pretty high tolerance of accuracy (or should it be inaccuracy), and then there are some that have a fairly low tolerance. These types of managers usually get complained about alot within inventory counter circles. One of the examples that Jackson/ Henry cited was estimating a partially full box of bubble gum, when you know that it contains 480 pieces when full. From this perhaps comes the notion that the degree of difficulty comes into play when determining when estimation should be used. The problem with that is the notion of "difficult to count" can be somewhat subjective, different counters will have different conceptions on what is too difficult to count, and thus maybe different views on what can be ethically estimated. As I counter myself I'm probably more hardcore when it comes to actually counting merchandise than the average counter, I try to count as much as I can. I'll take the candy off the shelf, I'll dump the merchandise out of a dump bin, count it and throw it back in. I probably have a lower tolerance of accuracy than most people (probably to the chagrin of the companies I work for), and If I have a lower tolerance of accuracy than a store manager I know I won't have any problems with them. Some counters may look at a dump bin, say to themselves that it's too difficult to 'actually count', eyeball it and come up with a roundabout figure, justifying the whole process by saying that's probably close enough. The problem with the concept of "ethical estimating" is that it sends a message to counters that it's okay to cut some corners here and there during an inventory, and this can be a very slippery slope to get on. Adding the notion that their counts only need to be within 1% only further encourages such behavior. I don't neccesarily agree with the phrase that such practices "don't affect accuracy". Estimating does affect accuracy, maybe at times it doesn't affect it to the point where it creates any serious problems, but let's face it, estimating is not the same thing as actually counting it. Anybody can eyeball a box of bubble gum and say it has 240 pieces in it, however there is no certainly to this method. We have no way of knowing wheather this is right or not, all one can do with estimation is to claim that it's probably close and move on. If such practices don't affect accuracy then a counter could adopt a reasonable line of thinking that these practices could used more often without affecting accuracy, and that may not necessarily be true. The 1% rule is in some ways a license to cut corners when counting, to adopt a counting style that has a relaxed tolerance of inaccuracy. It would probably be inappropriate for inventory company supervisors to instruct their counters that it's okay to guess, but they apparently don't need to, all they have to do is remind us that as long as we're within 1% we're fine.

Of course how does anyone even know when a particular count is within 1%. Let's say for the sake of argument that we're trying to get within 1% of every section in the store. For a given section let's say I count, I dunno $527. Is this within 1% of the true actual amount of money in that section? How would we know? Wouldn't we need to know the true total to make such a claim? Should recounts be done to see if these types of claims are true? Maybe the 1% rule, strangely isn't about any actual numbers at all, but rather sustaining a belief that the way someone counts is acceptable and good enough. The 1% rule exists to offer validation for all the estimation techniques that get used during a count. The danger exists in wheather the 1% rule helps mold certain attitudes and counting styles with inventory counting. Eight years ago, I saw a brief glimpse into how this can affect a counter's approach to the job. Working in a book store for RGIS, we were capturing barcodes for all the books and recording a piece count for each area. For each area we were required to count back afterwards to verify that our count was correct. At one point a counter near me uttered a question I'll never forget "Do we have to verify at 100%?". Unfortunately for the counter who posed that question, the unwritten rules of estimating don't always apply.