A Better Backspace Button

When I was listing some of the various factors that affect one's APH in "The Stats You Keep", one factor that I probably should've mentioned but didn't was the quality of the equipment being used when performing a count. In regards to the equipment, one factor has to be the machine itself, the device actually used to count with. When I read the entire "History of Inventory" article by Jack Henry, one of the most interesting parts of the article for me was the description of how companies counted inventories back in the early days of external services. According to the article back in the late 50's and early 60's RGIS were counting inventories by entering counts into "fifty pound rotary calculators" mounted on shopping carts. Naturally these calculators required an electrical source, so they were constantly plugged in using 100-ft extension cords.


Well needless to say, inventory counting technology has come a long way since then, and has grow along side the technology advancements in other walks of life. I've had the opportunity to use several different machines over the years, below is a list of the ones that I've used, in order of which ones that I have liked the most.

DC-5

There are actually a number of different machines in the "DC" series, I've only had the chance to use the DC-5 machine. The numeric keys always felt good when I used this machine, the keyboard desgin basically followed older machines like the ICAL where you had your basic 10-key design, with a multiplication key to the left and the equals key to the right, that allowed the input of various quantities under the same price. One could also utilitize chain multiplication, where you can enter in quantities multiplied together. (For instance once could type in 1.59 X 12 X 30, instead of doing 1.59 and then hitting "12 =" 30 times in a row.) Above the numeric keys where function keys that performed various tasks. One was the search function that allowed one to review their keystrokes on their machines. One could also delete erroneous entries. There were also function keys to separate totals by sections and categories. The most interesting and useful feature of this machine was that it allowed you to print off reports straight from the machine itself. By simply attaching the DC-5 to a printer, one could print off totals at a given point in time, or print off keystoke detail for any section or category that one wanted to review. For the keystroke details a report could be done for any section or category, so one could zero in on a certain sections of the inventory that might call for counts to be reviewed. Also the DC-5 could print from any computer that had the proper ports, so if your printer went down, there was the possiblity that you could plug the DC-5 into the store's printer and still print off reports. The machine's ability to essentially generate reports without a laptop and to print from multitude of printers, that made it one of the more interesting machines I've ever used. Other features include a LCD screen that could light up, when counting in the dark, and an off switch that was tucked away between the numeric keys, making it hard to turn off by accident. And even if you did all the data would still be retained. The only way to delete the data would be to choose this option on the menu options, and the chances of doing this by accident were pretty slim. The machine did malfunction on me once in a store, and unforuntely I lost my inventory data. However, I did close to 250 stores with this machine counting stores by myself, and this only happened once. So the machine was still pretty durable. The power mangement for this machine was very good as well, it ran on 4 "C" batteries that would last for weeks, and consider that I was counting 5 stores a week by myself, in some stores I would be using that machine for 6 to 7 hours a day. There really wasn't anything about this machine that I could truly hate. The machine also had a port for scanning capabilities (which I never needed to use) and there currently is a wireless version of the DC-5, which based on the pictures seem to look exactly like the non-wireless version. This is a good machine.


RGIS' "A" Series

These were the machines, RGIS were using when I first starting counting inventories, and they kept using them for another 8 or 9 years after that. These were fairly well designed machines. The numberic keys felt great for handkeying, and the machine also included ports for scanners and for transporting data to the computer. These machines seemed more designed for collecting data, then for multiplying things out like a calculator. Chain multiplication was not possible when in a store program. Store programs would require entries for several different menu fields, the ones most often used were F5 (usually for price) and F6 (quantities). Other functions would normally be F1 for areas, and F3 or F4 for barcodes when doing scanning inventories, F2 in most cases were used for subcategories like sections or departments. The use of the F2, F3, and F4 though was generally not very constant. The search functionality of this machine was for the most part equal to that of the DC-5, but of course you could only search through counts that were still in your machine. The one odd thing about this machine was the presense of the "6=" and "6K" keys. Other machines essesntially had a "6K" key which kept the price constant allowing you to keep inputing quantities. The "6=" key did not keep the prices constant, it looped back to the previous menu and required you to input the next price. Also this key was located where the 6K key would've been on other machines like the DC-5 or the ICAL, so using this machines after time spent with other machines was a little bit disconcerning because of those keys. Some machines also included a section for the entry of alpha characters, and with these machines you could almost do any type of inventory. The fact that these machines could be used for a variety of different inventories. weather to capture dollars, barcodes, or alpha-numeric codes was what I liked most about these machines. The only real downside to these was that they required the use of 6 AA bateries, and batteries did have to be replaced often. Also data could only be transferred by the use of cables, this included times when store downloads had to be sent to the machines. Downloading a whole bunch of machines only with cables was time consuming for certain clients. Another weird thing about these machines was that when you transmit data to a computer, you literally "send" it, it was not retained in the original machine after it was sent. All in all these were good machines, the few weaknesses it had could still be worked around effectively enough, to were they weren't really an issue at all.


Quantum's QIC

Like most other machines, this machine has the general 10-key design, with your mulitiplication key to the left and the equals key to the right. But one of the things I noticed when I first started using this machine was that the function keys on the side of the numberic keys were literally right on the side of the numeric keys with no space in between. With the "A" Series machines and DC-5, there was some space between the multiplication key and the "1" key, so I found the lack of space between these keys a little disconcerning, but I was able to adjust to that quickly, so it really wasn't much of an issue. Like the DC-5, one could do chain multiplication. The machine also had function keys at the top, but for financial counts they never got used, what was weird though was that the equals key was used for both entering quantities and for entering section numbers, to breakout counts. One was required to hit the clear key (or I think it was actually the "No" key) to back up to the previous menu and enter in a different section. Hitting the "No" key accidently may mean instead of inputing $1.29 X 3 in your machine, you may end up counting merchandise in section 1293, and hitting the multiplication key wouldn't stop you from doing this. If you're lucky enough to catch this early enough you could correct it fairly easily, if not you may get to the end of the inventory, and wind up trying to find where you counted something at $1.29. I prefer the design of the DC-5 and the "A" series where you have one key to input the section and another key to input the quantities, plus those machines will alarm at you if you hit the wrong function key as well. The search utility on the QIC though is one of the machine's best features, instead of hitting a button to bring up the search utility, your keystrokes are automatically generated as you count on the LCD screen, so you can look at them at any time. Not only that but the LCD screen for the QIC is far bigger than the other machines, in fact the LCD screen on the "A" series or DC-5 are so small that usually you can only see one entry at a time, even when you do search through your keystokes, although one downside to the search function is that one can only look at keystrokes in batches. Meaning that one can see the keystokes for section 1000 and then for section 1005, and then for section 1010 separately, one does not get a sense of the order in which these areas where done in. The keystoke detail for the DC-5 works the same way, but for the RGIS machines one can see keystokes as one long continuous string, so you not only see all the prices and quantities, but also when you inputted your section numbers, department numbers and so forth, you can also see the order in which you did your sections. When dealing with the mysterious section 1293, the RGIS model would be far easier to use, with section 1293 the issue isn't going to be the individual keystokes, the issue is finding out which area it should go into.

However, what I hated about these machines the most was the power management feature. Instead of running on batteries, the QIC ran on an internal power source. The problem with this was that the machines had to recharged daily, and sometimes you may even have to charge the machine up between your first and second stores, just so you would have enough juice to last through the second store. And if you did run out of power, your choices were either to insert batteries into the machine (which for some reason never really worked all that well for me), or to stop counting and spend time charging your machine up to where you can count some more. There were times during an inventory, where I would have to stop counting, go out to my car and charge my machine up using my car charger, because it was the only charger I had. I also remember several inventories were I would be counting in the back office at the end of the day, and I would literally have to count with the machine plugged into the wall in order to have enough power to finish. Having to charge these machines up all the time was to say the least very annoying. I would much rather use a DC-5 where I could toss in a few batteries and not worry about my power for weeks. The QIC also wouldn't rank high on the durability scale either, I knew a guy in another market, who went through 3 different QIC's in a span of 2 months because they would keep malfunctioning on him. When the machines worked and where fully charged they were good machines, but the power management and their durability are enough to keep them from being my favorite.


RGIS' RM-1

These are the machines that RGIS have used since they stopped using the "A" series. If you look at the pictures of all the other machines, you'll see one main similarity among them all, and that is they all have the same basic 10-key design for the numberic keys with the same function keys to the left and right of the main 10-key pad. The RM-1 is first machine I've seen that has messed with this design. They have essentially taken all the function keys that used to be off to the side and shoved them under the numberic keys. On the RM-1, the "multiplication" key is actually under the numberic keys instead of to the left of the "1" key, furthermore the quantities or "6K" key is under the numberic keys as well. This allows the machine to have a thinner design, which I suppose makes it easier for holding in your hand. But this benefit is slight when compared to the keyboard design which is far more troublesome for handkeying. A few co-workers of mine at RGIS have flat out said that you can't handkey with these machines, I have to agree with them on this. The keys just don't feel right on this machine. The machine has the button for entering quantities at the top of the keyboard, essentially the "6=" key, but using this just feels too unnatural, and reaching for keys below the "1" or the "3" still doesn't quite feel right either. The button at the top is also used for entering sections, departments and other breakout numbers, but unlike the QIC, hitting the multiplication key when you're supposed to hit the quantity key will cause the machine to alarm, so there's little risk of a section 1293. As for the other features, the search utility is slighty better than it was on the "A" series, and you still can't do chain multiplication. The best improvements over the "A" series are that these machines have built in lasers for scanning which can be triggered using buttons on the side, but of course RGIS encourages it's counters to use finger lasers that need to be plugged into the machines, just like they were in the past. The other big improvement is the wireless capability. These machines can transmit the data to the main computer wirelessly, which allows reports to be generated far quicker then in the days when data had to be "physically" transported to the laptop. The wireless feature is pretty valuable for RGIS when you consider that they do a lot of big stores, that require a large number of counters, but did this wireless capability also have to include unnecessarily toying with the main keyboard design? Seeing how there's a wireless DC-5 machine with the exact same keyboard design as the orginial, I'm thinking "No". The power mangement is also better than the "A" series, they now use built in batteries that can get recharged, in fact this may be the best power management system I've seen yet, as for the durability however these machines do have the tendency to freeze up ever once in awhile, but for the most part work fairly well. One annoyance though is that sometimes the machine has trouble sending data wirelessly. If you're standing too far away from the antenna, it's almost impossible to transmit, and this either requiers transmitting to a flash drive, or walking a little closer. There have been numerous stores where I literally have to walk towards the antenna in order to transmit. There are still some kinks to work out with the wireless functionality. But still the handkeying aspect is what bothers me the most. Thankfully most of the stores that RGIS does these days requires scanning. In an auto-scan environment the RM-1's aren't really that annoying, but in a financial count that requires nothing but handkeying, this is the last machine I would ever want to use.


ICAL

Oh, yes the good old ICAL, the perverbial dinosaur of inventory counting machines. In the brocheure that I have for the ICAL it shows a picture of the ICAL 150, attached to a computer, that looks exactly like the one from the second season of the TV show "Lost" when they had to press the button every 108 minutes. I also love the "floppy" disk in the photo which gives you an indication of how old this machine is. But what can I say about the 100R, well it has the same 10-key design that has been used for other machines (Except the RM-1, of course), it allows you to store your running totals into sections for breaking out totals. It also allows you to use chain multiplication. The machine runs on 2 9-Volt batteries that actually will last a long time. The machine utlilizes a delibrate keying sequence to turn on and off, (I've forgotten what that sequence is). Of course if by chance you turn it off then you've lost all your data. It won't retain the totals like other machines will. Also the search utility only allows you to see the last 4 extensions that you've keyed in, but nothing beyond that. Another thing of note is that the ICAL-100R is a self-contained unit. There are no ports on this machine to communicate to a computer or even a printer. If you want the totals transferred from the ICAL to a computer, you have to do it by hand. Essetinally this is just a large high grade calculator. Granted it's not a bad machine to handkey with once you get used to it, personally I'd rather handkey with this then with the RM-1. But the real weaknesses here is in the search utility, only being able to look at the last 4 extensions, is not very useful at all. If you make a mistake on an ICAL, generally your only option is to do a recount. With no way to transmit data to a computer, and no real way to view one's keystokes, there simply isn't any real alternative. There's less of a safety net if you screw up with an ICAL, so for this reason, I'll rank it last.


It seems that each machines has it's own little quirks that requires certain behaviors when one uses them. With the "A" series it was trying not to use the "6=" key. With the QIC, I would have to review all my sections periodically to make sure I wasn't inputing any wrong sections. With the RM-1 I find myself actually holding the machine in my hand more often, with it's more narrow design and a built in laser, it felt more natural to hold than other machines, where I would just let them hang from my belt. With the ICAL I find myself working a little more meticulously, and being more vigilent of possible keying mistakes. This leads me to think, does using certain machines help mold a certain counting style? What effect do all these various features have on how a counter goes about performing their job?

It's amazing to consider that there are still inventory companies today that still use ICALS to count stores. I spent time working for one called AIS. When I think back about my expereince working for them, I must admit the way they did things was far more unique than any other company I ever worked. For one working for them required a lot more paperwork. Every total you compiled on your ICAL had to be recorded on a worksheet, and that total had to be recorded to another piece of paper where it would be compared to the totals from the previous inventory. Everything was hand recorded, there were no computers, or printers involved at all. This was a sharp contrast to Quantum who never really did anything on paper. With Quantum all the counts got inputted into the QIC, which got transmitted to a laptop, which did the comparisions electronically. The QIC even stored an electronic version of the store map. Also there was a well defined process of how things were run. Counters for this company never "roamed free", they were given assignments by their supervisor, and when they were done, they reported back to their supervisor to report totals. Current to previous comparisions were very important here, because there were few other options in determining the quality of the count. If a manager thought a count might have been off, then recounts were done. Another aspect of this process is the notion of an "audit trail". When doing financial counts one would write the running total for that section on a tag and place it at the end of the shelf to mark how much we had counted up to that point. This would allow someone to see not only the cadence of how one counted a grocery aisle side, but it allowed one to break totals down into smaller numbers, so you could see how much money was on one particular shelf or natural break. Of course, for this to work one has to tag well, and of all the inventory services I've seen, this company is the best at tagging, there's no question about it. There was one client where we scan their inventory using their own scanners, even in a scanning inventory tagging was important to AIS. At one scanning store I remember my supervisor, pulling me aside and letting me know that I wasn't tagging correctly, he then proceeded to tell me how I was supposed to do it. Let me make this clear, 'he was telling me how to tag properly', this is something that would never happen at RGIS, absolutely positively never. Granted like all other inventory services, AIS did care about our overall APH, and they did want people to be productive, but productivity wasn't the only thing that they cared about. And they seemed less inclined to toss other aspects of inventory counting aside for the 'high APH at any cost' approach. People with this company weren't afraid to pull someone aside and show them the proper way to do their job, because of this I have a healthy dose of respect for AIS, these types of moments probably don't occur enough in the world of inventory counting. That being said the methodology, despite some strong points, still felt antiquated and frankly old-fashioned. Counters who are more used to 'modern' methodologies would have a difficult time working for AIS. But one of the more difficult aspects has to be the ICAL. This is not a machine you want to make a lot of mistakes with (as if there is one you do). A mistake here means a recount, nothing else. In some ways using the ICAL forces you to handkey better and more accurately than another machine would. Despite being somewhat outdated the ICAL can still force habits that actually might benefit one's ability to count. It seems that technology (or in some cases the lack of it) can play a role in developing one's style and ability. When we did those scanning audits for AIS, we used handheld devices where one simply entered in a quantity for each barcode that they scanned. The trick with these machines however was that there was no '6K' button, so you couldn't just go 3 "6K", repeatedly over and over again until you counted everything for that barcode. Without the "6K" key, I found myself actually physically counting the merchandise more, I also spent more time counting to far bigger numbers in my head then I ever would have done with RGIS. To this day I'll still find myself doing this, and I have these machines to blame for that.


To further consider the effect of technology on human behavior consider the notion of typing. The art of typing as seen a lot of changes in technology from the days of manual typewrites utilizing typebars to personal computers. One of the biggest changes has been in the correction methods used. Older methods involved using rubber erasers and correction fluid. In more recent times other methods have included correction paper, or correction ribbons used on electronic typewriters. Correcting mistakes on a manual typewriter could at times be problematic and difficult. Erasers would leave debris from both the eraser and the paper that needed to be carefully swept away, one also had to be careful not to create smudge marks. Correction fluid however could leave a visible patch of fluid when held up to the light, and also required time to dry to the paper before it could be retyped. Also both methods would probably require removing the paper from the typewriter and reloading it, so as to not get debris inside the typewriter. With these methods it seems as if the best way to deal with mistakes was to not make any. Accuracy in typing is very important when using manual typewriters, and in the mid-1900s accuracy was valued as much as speed was. Methods for correction became easier and less troublesome with electronic typewriters, until the advent of the personal computer. With the PC paper and ink have essentially been replaced by pixels of light emitting from a computer screen, and this has probably to some extent made the use of liquid paper and typewriter erasers almost obsolete. The PC has changed typing so much that today we have a different name for it 'keyboarding'. In the world of keyboarding to correct mistakes we have the Backspace Button. This allows one to just wipe a letter off the screen as if it never existed and to key in something else, making it look like we typed in it correctly all along. The Backspace Button has made correcting mistakes so ridiculously easy and effortless, one has to ask what effect this has had on the notion of accuracy. You could argue that in the era of 'keyboarding' the importance of accuracy has decreased sharply. Mainly because the cost of making a typing mistake has been reduced to almost nothing. One could make all types of mistakes, and be able to fix them with a few additional keystrokes. Mistakes are so easy to fix that prevention is no longer the best cure. Speed on the other is still desired as much as ever. It fact people may feel more willing to aim for speed knowing that they have a backspace button to save them in regards to accuracy.


I've seen this concept played out as an inventory counter. When I was still relatively new to counting I remember doing a store for RGIS that required detailed printouts on every area. The store staff would then follow behind us and review the printouts for accuracy and then inform the crew leader of corrections that needed to be made. At this store there were a couple of instances where I would fat finger a quantity, for example I would hit '56' instead of '5'. The store manager became aware that I had done this about 2 or 3 times, and instructed the crew leader to talk to me about this. When the crew leader came up to me however, she told me to keep counting the way that I was counting. She didn't want me to slow down in order to count more accurately, her attitude was that if I made a mistake like that she could fix it, and it really wasn't a problem for her. Granted the crew leader was probably a little more vigilant about spotting mistakes prior to posting the detail printouts, but other than that no adjustment needed to be made. It didn't matter that I, or anyone else for that matter, were making mistakes, because RGIS had a system that could effectively deal with these mistakes, and to correct them to make it look as if we counted it correctly all along.


Every technological advance in inventory counting from 50-pound rotary calculators to machines that can transmit data wirelessly, has had one objective in mind to increase productivity, while maintaining accuracy. But the concept of accuracy seems to take on two different forms, on one hand there's "process accuracy", how accurate one is in actual counting the merchandise initially, and there's "product accuracy", how accurate a count is at the conclusion of the count. I remember an auditor once saying to me that "It's not a mistake until you leave the store", in this case they were referring to "product accuracy". Audit companies seem to focus more on the final product than on the process one takes to get there. Along with this are technological advances that seem to have enabled a sloppier counting process, much like 'keyboarding' has allowed us to type in a very error prone manner, something we could never get away with using more rudimentary equipment. It's possible to achieve a high level of "product accuracy" even when there's some suspect "process accuracy" present in a count. As a result the notion of "process accuracy" doesn't get stressed very much, and really why would an inventory service stress it at all? Speed is viewed as a more important facet during the process of counting anyway. Focusing on "process accuracy" can mean counting in a more time consuming manner. Wheather this could save more time at the end of the count is debatable.


As long as the magnitude of mistakes can be kept at a manageable level, and a service can deal with them in an efficient manner, they probably won't mind mistakes if it means that counters are productive. With more modern methodologies there's no such thing as too fast, especially when mistakes can be dealt with. Of course, in regards to mistakes, speed and accuracy are not necessarily adversarial concepts. Sometimes process accuracy is speed. Counting something once is faster than counting it twice. Nothing brings that point home more than the ICAL. If the ICAL has one redeeming value it's that it teaches you that sometimes the fastest way to count something is to do it correctly once and be done with it. With the ICAL "process accuracy" and "product accuracy" practically become one in the same, and essentially the process becomes everything.