Created: 1993.09.03
I've had a bit of experience in inventory count, and I've though more about it - yes I think about weird things! Here are some thoughts that might help you if you are doing inventory counts (or should be.)
I first did an inventory count when I was 17 years old and I was 'seasonal department manager' at Woodwards department store. Sometime after, they offered me a job as a full manager trainee, I turned them down preferring to do a 4 year Management BComm stint at the University of Calgary. But I learned a lot of extremely good lessons from 'my' manager at Woodwards. Some of my thoughts in this post come from my experience there, some from the great ideas they had for doing inventory count and some of the weaknesses I saw in how they operated.I hope the comments here will help some companies. The ideas here are not in any great order - but many of the ideas are great ideas! Whichones are important for you will depend on your inventory reality.
Do your count when the quantities are as low as possible. The higher the number of items to count, the more likely the count will be inaccurate, and the higher the cost of doing the count. Doing the count just after you got all your Christmas stock in, and there are boxes of stock piled in the isle because the warehouse shelving isn't sufficient to contain all the stock is a terrible time to count.
Some items are better to weigh than count. For example, if you stock 1000's of several kinds of nuts and bolts, you would be better to weigh a quantity - perhaps 100 very carefully counted bolts, then weigh the stock. Even better, weigh 3 or 4 samples of 100 carefully counted each - and see how close the weight is. Make sure you are using a scale that can handle the level of accuracy - for example, many electronic scales may only be accurate to 5 or 10 grams, but if your items weigh about 1 gram each, then with 100 you could be off by 10% just because of the scale error. But if your items weigh 20 grams each and your scale is accurate to 1 gram, then you likely have a very accurate measurement even with 1000's of items.
Decide how many people are going to do the count. Also consider their skills, someone who is detailed oriented is preferred to someone who is known to be sloppy.
Decide who to do the count. If you have a lot of inventory, it may make sense to train a team that do this as their full time job, and find people with the right level of attitude to do it. If you have a small amount, you may pick people who have blocks of time with nothing really to do, and task them to do inventory counts during down times.
Decide how much of the stock to count at a time. All of it? Some of it each day for days, weeks or even months?
You should have 2 people do each count - at slightly different times - working independent of each other, so they don't make the 'same' mistakes.
Coordinate between your 'regular' staff and your 'counting' staff, to avoid friction, errors, confusion and getting in each other's way.
Document your process
Mark items (samples etc..,) that should not be counted
Mark items that may be hard to count (such as things that come in packages of 100 but are sold in singles - you don't want the package of 100 being counted as '1'.
Make sure the staff doing the counting has the right equipment, space and time to do the job.
Determine how you are going to count hard to count items - such as 'yards of gravel' and try to be consistent from count to count.
If appropriate, provide maps showing clearly what is and is not to be counted and where to find what needs to be counted.
Order more tags that you think you will need so you can note what has been counted and what hasn't. Even if you are using a computer to record, you will want to tag everything that has been counted so that, when you are done, you can check for any product that wasn't tagged (wasn't counted) and take corrective action.
Clean up your data. Make sure all the inventory is in the system, that everything 'looks good', everything has part numbers.
If the product is missing tags, tag it all before the count starts.
If working with an independent inventory service, review their instructions and your agreement with them.
Prepare the counting location
Clean up, go through and make sure that boxes of product contain the proper items. Clean up loose bins - making sure the correct product is in the correct bins, this can be especially important in a retail environment where customers can put things back in the wrong bins, but I ran into a similar problem in a aviation repair facility with 1000's of parts where one of the staff members was, well, I guess we would say sloppy.
Get rid of all the obsolete and defective product and disposed (or moved to a 'discount area')
Make sure expert staff is available to help when the counters run into difficulties and questions.
Anything that can be 'counted and sealed' ahead should be, and depending on how you are recording, the sealed containers can have the counts. If you are ENTERING the counts, then tag them as 'counted', otherwise just seal them with the count marked for the counting stage. Note that if you typically sell 10 of something a day and you have stock in the 1000's - you could count, box up and seal all but around 100 ahead of time, reducing the time that the count is inconveniencing your operations (the stuff that makes you money.)
Make sure all paperwork is entered into your system (computer or paper - whatever you have).
If you have outside storage locations (overstock), count and seal them ahead of time
Let your 'normal' staff know all of this is going to happen, tell them to plan ahead - this may mean getting stock out of the outside storage locations so they don't run out and cause headaches either for customers or for the counters during the count.
Communicate with everyone who will be affected. Reminds me of one of favorite Dilbert jokes: '"You're not understanding what I communicated to you."
Stop shipping, receiving or work using the parts during the count of the parts. Sometimes this can't be done, but it will make the process more difficult and more error prone.
Ideally you would count all the 'overstock' ahead of time, and seal the boxes or area so that those items can't be moved/used during the count. This makes the total amount that has to be counted 'in a hurry' the smallest and easiest.
If possible, doing the count 'overnight' when nothing is moving is ideal. This may mean scheduling the count by sections over several nights. Obviously if you run a 24/7 mail order operation, this won't work unless you can 'lock' section by section
Don't show the counters the current count estimate. They will be tempted to be sloppy if the count is 'close' - and they may overlook that box of 10,000 parts on the top shelf.
Have the 2 counters work independent - so they aren't tempted to make the same mistakes and/or try to have their counts be the same - human nature, not an intent to deceive.
Consider counting 'slow moving' items first, partly as 'practice' if your counters don't do it as their full time job, and because with slow moving items you may be able to lock them down until the count is over without causing undue hardship.
Allow the counters to make notes. So for example, they might note: Large box on top shelf contained 50% part A and 50% part B. Encourage them to make notes of EXCEPTIONAL items, not notes on the 'normal' things, this will make recounts and accuracy checking easier.
Have a few 'random' counts done by an 'expert' to ensure the counts are progressing reasonably - pick some simple ones and some complex ones. This is especially important if your counters are newbies.
Items should be counted by 2 people independently, without seeing each other's counts, and the values compared and a level of consistency - you might decide that if they are within 1% it is not worth doing a third count. You may be tempted if the counts are extremely different to have a third count done and, if it matches one of the 2 previous counts - go with that count. But this can be a bad idea if the 3rd count matches the lower of the 2 counts, because the error may be due to missing a box, and it can be a bad idea if the 3rd count matches the higher of the 2 counts because the error may be due to a mis-labeled box being counted that the person with the lower count verified was not the product in question. You may be better off sending both counters back and recounting together all the bad values - and see if they can agree on the recount, especially if they can agree on why the one count was bad.
When the initial count is done - go through and see if everything is tagged. Anything not tagged was missed in the count process.
Track your results, find areas where results are consistently wrong and see what action can be taken. This is called 'metrics' and 'exception reporting' and 'simply good management'.
Keep records that may be necessary for a CRA (IRS) audit if your processes may be called into question.
Not everyone has the mental and physical skills to do a job like this. Some consider it tedious and may get sloppy. It doesn't mean the person shouldn't be working in your company - just that they aren't cut out for this job in the future.
When you are content the count and recounts are done - clean up all the labels, seals etc.., so your business can return to what it does to make money. (If you are a inventory count company - cleaning up means you are more likely to be hired again.(
It is often called 'shrinkage' because most of the time the count shows that stuff is missing, not that you have more than expected.
When selling rope or chain, you undoubtedly have seen a salesman sell you 3 metres - but they give you 3 metres and a few extra centimeters 'just to make sure' that they don't cheat you. Repeat this 1000 times and you will be missing 30 metres of rope when you do the count. This is generally considered a worthwhile cost for customer satisfaction. But the result is that your stock level in the computer doesn't match your actual stock, and if the discrepancy gets too big - you will run out of stock when you think you have lots left.
When selling a product like grain - the shrinkage can be literal - if you buy by weight or by volume, it can dry out and shrink in both weight and volume even if you never touch the stock.
Mislabelling: a box might have the wrong stock thrown in it. This doesn't actually mean the stock is gone - it just means that you think you have more of one stock and less of another item.
Similar to the above, if you have a bunch of little bins that customers can remove (and return) stock to, you can be pretty sure that some bins will have the wrong product in them.
And, Similar to the 2 above, this time the fault of someone stocking shelving, recently, at Aikenhead's (A major hardware chain here in Canada), I was buying a bunch of one product. The product was all missing labels (it was a slightly oily metal drywall product so I guess the labels didn't stick), and it was in the wrong bin. When the price checker came back he told the cashier a price that was way too low - when I mentioned this, he went back and verified that someone had put the stock all in the wrong bin. Now ... from an inventory count situation, this meant that they were going to sell me 100 of one product but they were going to remove it in the live stock levels from the stock count of a different product. The result would have been that both would have been out (and had I not mentioned the mistake, they would also have lost a lot of money.) Update 2017.11.30: Aikenhead's went out of business in 1994, so maybe this practice was one of the reasons - who knows.
Theft by customers.
I don't know the sources, but I've read several times that most companies will have a lot more theft by staff than from external. A fun story I heard (probably made up): Every night in a factory in Russia a man took a wheelbarrow of sawdust home 'for his garden' he told everyone. The guard at the entrance ran his hands through the sawdust every day - he was sure the guy was stealingsomething. After months of this he pleaded with the guy saying 'you are driving me crazy - I know you are stealing something - just tell me what you are stealing and I promise I'll let you keep stealing it". The guy answered simply 'wheelbarrows'. Update 2011: I had a construction company, I was losing 20 pair of deerskin gloves a month. When I fired one person (unrelated reasons) I stopped losing deerskin gloves. He was stealing 1 pair a day - and I'm sure it was on purpose.
Evaluate your process - interview your 'regular' staff and 'counting' staff, change the process documentation for future counts.
Evaluate the time it took as part of this - this will let you know which teams are more efficient, which processes are more efficient, and from that you can learn. The more efficient teams may be able to teach the less efficient teams how to improve their work - but don't just go based on speed, fast but wrong is not the better team.
The higher the frequency of counts, the more accurate they will be.
As you gain experience, consider setting count priorities based on past experience of errors. So if you find that one type of stock has a great deal of shrinkage (or increase) on a regular basis, and another type the counts almost always match the values calculated based on usage, then you would want to count the problem ones more often and the ones you have no problems on less frequently.
Also consider setting count priorities based on dollar value - where are you 'losing' the most.
Use the count results as 'training'. For example, you may be ok with 5% shrinkage in your rope and chain sales, but if you are seeing 15% - your people may be giving too much excess away, and training can help fix this.
Evaluate shrinkage and theft
Making sure you have physical stock when and where you need it. If your computer says you have 1000 and you only have 2, you may lose sales when customers are waiting for you to fix your problem. When you have multiple locations for stock, being able to track what is where at what quantities will let you make adjustments to lower your overall inventory and make sure product is where you need it when you need it more often.
Managing stock levels, Sometimes shrinkage is just part of the cost of doing business. But knowing what you have and where, and how much shrinkage to anticipate can help make better decisions.
Insurance, if you know what your levels are and can show your count history, you are more likely to get a fair settlement in case of disaster. Talk to your insurance company - ask them what they want to see in your records. Also make sure your records are kept in a place where one disaster won't wipe out both the physical product and the records (with computers, you should be using offsite backups routinely.)
Accounting and Taxes. Think about the points in time that these are needed. Counting in July when the tax department and your accountant want to know your counts in December, might be the wrong time to do the count, or, maybe it is fine. But make sure you have thought about this.
I hope a few of these, or even just one, has helped you run your business more smoothly.
If you have suggestions for improving this, please let me know. If you want credit for your idea, I'll be happy to do so. Perhaps reciprocal web links? Update 2017.11.30: Back when this was written originally, reciprocal links were very valuable. Not so much as the SEs have gotten smarter.