Now that we know how to monitor Level 4 - Efficiency, let's see how to use SKU-level data and your computer system to help identify rogue counting processes. We will use our Shrink Report (Review Report email) to highlight variances caused by unknown processes and then interview the person responsible (or consult video) to determine if the count was desired. Once discovered, you can meet with your store manager or owner to ensure enough labor is devoted to the counting process to get a good count, or if the counting is not worth the labor effort, then discontinue the process.

• Not all counting is good and not all counting is bad. Our goal when analyzing these Shrink Report variances is to determine if it was part of an effective counting process or if it was a rogue process. An example of a rogue process is when a person grabs an RF gun and starts counting without pre-sorting bins/hooks, identifying back/top stock and without the benefit of an inventory coordinator performing variance recounts and variance research.

  • Counting in a busy retail environment is hard. Our research shows that nearly 40% of all inventory variances are either an error or a correction. Counting correctly requires four laborious steps, where count quality diminishes as each step is skipped. As count quality diminishes it becomes hurtful to your store's accuracy efforts. 
    
    Steps to obtain a good count

    1. All product needs to be in the right spot; not an easy task when you consider the potential of similar-looking items being commingled in nearby hooks and bins. Additionally, back stock and multiple locations must be easily identifiable. Make sure your shelf maintenance (flossing) is taken care of before engaging in counting activities.
       
       
    2. A good initial count has to be taken. Even when your shelf maintenance is in good order, counting accurately SKU after SKU, hour after hour is difficult and not everyone is suited to this task! 
       
       
    3. Variances should be recounted, ideally with a second (and more experienced) set of eyes. 
       
       
    4. Variances left after recounts should be fully researched to determine the root cause; if not found and resolved then the variance will likely recur.
    Few stores can devote this kind of labor when counting every SKU in their store so steps are skipped. The result is an inaccurate physical inventory count where nearly as many errors are introduced as fixed, any accuracy gained will be temporary because the root cause was not resolved. In fact, the counting process may have masked a serious internal issue.
    
    Widespread counting where each SKU is counted devotes significant labor to low-risk, SKU-dense areas of your store like o-rings, faucet parts, v-belts and plumbing fittings. At the expense perpetually auditing high-risk SKUs such as top-dollar items, theft-for-cash candidates, obsolete SKUs and the other types found on your Count Sheets. 

  If your store has low efficiency then there is a process or two causing a lot of counting, here are some examples of what to look for: 

  • A department manager might take Shooting Outs to mean counting not only SKUs with zeros (holes) but also SKUs with 1, 2 or 3 on hand.
  • A counter may decide to count "around a variance" where nearby and similar-looking items are counted when a variance is found. This is fine as long as the counter is good at sorting similar-looking SKUs and your inventory coordinator has the time to recount subsequent variances.
  • Staff is counting during "red" times: when the sales floor is busy with customers or when a delivery is in process. This can cause poor initial counts because of constant customer interruptions. Counting during delivery windows can cause significant issues unless the counter is in sync with the order's receipt into the computer system and all items have been put away.
  • A department manager routinely counts their department. Sometimes this is desirable, like after a reset or when a deep-sorting of plumbing or electrical fittings has been performed. But a department should not need continuous counting to stay accurate. Accuracy labor is best directed to sorting bins, dealing with top/back stock and cashier testing. 
  SKUs counted while processing Mango's Count Sheets and while Shooting Outs are effective processes because they are dealing with high-risk inventory and counts have a higher probability of being wrong, and variances tend to be more researchable. Other counting processes need to be examined to determine if they are increasing sustained accuracy in your store or are merely contributing to variance noise.

• Our first step is to open the Shrink Report and scroll to the Adjusted section (usually page 2). From there you can write down SKUs who's variance was not caused by processing Count Sheets (CS in the Research Hint column) or Shooting Outs (Zeroed in the Research Hint column).

  

  For example, the Brush-b-gon and the  Fire Starter (1st and 2nd SKUs in the list) were not zeroed out when counted (not part of shooting outs) nor were part of processing Count Sheets (CS is not in the Research Hint column). These variances, along with the other yellow-highlighted rows were likely caused by counting processes other than Count Sheets and Shooting Outs, therefore further examination is in order. 
  
  Let's write down a few of these highlighted SKUs for further investigation: 86057, 73598, 1010073, 601709
  
  Note: because of the way Mango receives data from your point-of-sale system, it cannot determine in every case if a variance was caused by shooting outs. Therefore, some of your variances may have been caused by shooting outs but it will not say "zeroed" in the research hint column. This is okay because these shooting outs zeros are easily seen in the next step.

• Let's take a closer look at the SKUs we've written down or highlighted from our Shrink Report. From Eagle Browser, launch ITR (inventory transaction register).

  

  The Fire Starter example (above) shows that on 9/6 its quantity-on-hand (QOH) was adjusted down by 1 unit from 2 on hand to 1 on hand. This SKU was not zeroed out (Shooting Outs) nor was it on a Count Sheet so we need to understand the process that caused the variance if unknown.  Note: the SKU was also received within two hours of the variance. The user and terminal is logged which makes research simple!
  
  If you are a good inventory coordinator then you should have a burning desire to rule out rogue counting happening while processing a delivery! This may or may not be the case in this particular situation, but it has to be ruled out.  We've seen many cases where a department manager will count his department immediately after "putting truck away" without regard to the status of the delivery's receipt into his quantity-on-hands.

  

  For the Brush-b-gon SKU, we see its QOH was changed to 0 on 9/19. We recognize user 130 is the lawn and garden department head and 9/19 was a Monday, which is when she shoots outs. Although Mango could not pick this particular variance as a zero, it was performed during shooting outs.

  

  Here we see a spraypaint SKU being adjusted downward by one unit from 7 on hand to 6. Again, we see the count was entered a closely after the item was received. This count was likely not a shooting outs process (otherwise it would have been zeroed) and was not on Mango's Count Sheets. So then what process caused the variance and did the process involve all the necessary steps to achieve a good count, or was this a quick change without a recount looking in nearby rows or back stock/top stock check.