Understanding IRA

IRA is a measure of your store's quantity-on-hand accuracy. Unlike Dollar Accuracy or Shrinkage, where an item's cost is taken into account, IRA focuses narrowly on quantity-on-hand (QOH). QOH accuracy is vitally important to day-to-day operation because computer-generated orders and in-stock position are dependent on correct QOH. Confidence is also necessary when researching root quantity-on-hand variances as it is difficult to sniff out theft and errors when we do not trust a SKU's initial QOH. For this reason, we like to see an IRA metric of 80% or better (8 out of 10 QOH's match the computer when counted).

  • If your Inventory Record Accuracy metric is showing "??" this means that your store's Count Sheet Completion metric has dropped below 95%. Note: the Level 0 Completion Metric reported on your Scorecard is a three-month metric but we allow you to make up missed counts from previous months and get credit for the work.

    Low Count Sheet Completion Descending Count Sheet Completion

  • RC Count Type Count Sheets

    IRA is measured by statistical sampling as your store processes its Count Sheets. Each month's Count Sheets contain 55 random samples ( RC Count Type) and the results of these samples are averaged over a three-month timeframe (165 samples).  Your IRA metric is entirely made up by the variances found when counting the RC Count Type so careful counting and variance research is critical to achieving an accurate IRA sample.

    IRA is scored using the "Good Count" method where only "good" (zero variance) counts are considered accurate. For example, if we count 19 foam paint brushes in the bin but the computer shows 20 on hand then count fails because there is a QOH variance of 1, even though the variance is trivial in this case. Because of this, the "Good Count" method of measuring IRA can slightly understate real-world store accuracy because some insignificant variances can count against the scoring. This is why IRA only has an 80% goal threshold. 

  • Inventory Accuracy Scorecard For IRA

    Mango's Inventory Accuracy Scorecard located in your Review Email, Charts.PDF, page 3, is your primary "go to" accuracy report. This report details and trends the results of processing your Count Sheets, measuring IRA and other accuracy metrics.

    The middle-left of this report shows "Count Sheet Results" for the last 3 months     (red bullet 1 above). Bullet 2 shows the total number of Random SKU's ( RC Count Type) counted over this three month period (should be 165 counts if your Level 0 - Completion is 100%). Bullet 3 shows the IRA results of these random counts using the Good Count method. In this example, of the 165 SKUs counted, 124 matched the computer quantity-on-hand without a variance (had a Good Count). This results in an IRA measurement of 75% (124 good counts divided by 165 total counts). This 75% number is what makes up Mango's Level 1 - IRA metric.

    Inventory Accuracy Scorecard IRA % Increase

    In the example above we see a thick red line which makes up the IRA 3-month trend (average). Each month's IRA sample (55 samples per month) is plotted and Mango matches each plot with a validation formula to make sure reported IRA makes sense (Mango looks to various internal store QOH movements that correlate with IRA). If the plot is validated it will show as a solid diamond with a black circle, otherwise Mango will substitute its own internally calculated IRA and the plot will show with an open diamond. Validation protects from an errant or dishonest count. Read more about IRA validation if it applies to your store.

    IRA Inventory Accuracy Scorecard

    Above, we see our example store's 75% Good Count IRA metric reported as the Level 1 - IRA key metric at the bottom of our Inventory Accuracy Scorecard. 

  • The basics of QOH accuracy (IRA) are quite simple: add to QOH when a product comes in the back door and subtract from it when it goes out the front. Assuming QOH is accurate to begin with, and there are no shipping errors, receiving errors, cashiering errors, counting errors, nor theft, then QOH will always be perfectly accurate. But this is rarely the case in the real world, so where do things go wrong?

    Inaccuracy is usually blamed on theft, but a close analysis of QOH variances shows that most errors are caused by our own hands. In fact, a third of all variances are not theft (shrink), but the opposite, called "swell". There are only two conclusions to be drawn by this up/down behavior 1) thieves have guilty consciences and after are stealing thousands of items, are putting them back on our shelves or, 2) our own employees are at fault by causing then correcting errors that makeup of two-thirds of all store variances! So if it is mostly our own fault then what's going on?

    Top causes of inaccuracy
    Stores struggling with IRA are usually focused too much on counting and "fixing" QOH and focused too little on root accuracy processes such as cashier testing, careful counting and variance research, and high-standards department maintenance.  

    #1 Sloppy counting, cashiering and department maintenance

    • Cashiers - without specific training and testing, one-third fail to scan simple basket correctly.
    • Counters - inaccurate counters if left to their sloppy behavior will do more harm than good. Variance research is the only way to determine the good from the bad.
    • Back stock - SKUs with back stock are difficult to count unless clearly marked as such.
    • Departments - if a department is sloppy then so will its counts. Each SKU needs a scan tag and be in its proper home.

    #2 Excessive Counting 
    Abundant cycle counting and/or full physical inventories, the traditional way of obtaining and maintaining accuracy tends to make things worse by introducing significant variance noise that it is impossibly overwhelming to perform careful root-cause research. There is no correlation between a store who counts a lot and its IRA (in fact, there is a slight inverse correlation).

    • Excessive counting masks the root cause of inaccuracies.
    • Excessive counting is not correlated with accuracy and is, therefore, an inefficient use of accuracy labor.

    #3 Theft
    Every store experiences theft, in fact two or three items are stolen each day we are open. Our margins and store operations (processing Count Sheets and Shooting Outs) compensate for this. However, when theft goes beyond a couple of items a day then things can get out of hand quickly and it can cause real business health issues. The risk of professional theft is real (many times it is internal too) so it is imperative to have tight inventory control, confidence in initial QOH, and relentless variance research. But it is difficult to detect theft with low IRA and it is made worse by a lot of sloppy cashiering, counting, departments, and variance noise caused by overabundant counting.

    • Too many errors mask true theft.
    • Too much counting and "fixing" create too much variance noise and masks theft.

    Bottom line: if your store has a root accuracy issue like inaccurate cashiers or sloppy departments then counting a SKU once or twice a year will not be sufficient to maintain its accuracy or discover procedural or personnel issues and detect theft. However, when you tighten up your procedures and work with your staff on accuracy, the error rate will be cut considerably and great IRA can be sustained!

    Mango's Level 1 Checklist will guide you through establishing these tight procedures to get your store in shape for accuracy! The first step is to assign someone to head up the challenge. This person is called the Inventory Coordinator (IC) and will be responsible for guiding your store through the checklist. The IC does not have to be a full-time position, rather it is a role assigned to an assistant manager or another employee who exhibits good inventory skills.