Let’s Talk Retail Excellence:
How do you measure on-shelf availability?
Full video transcript
Chris Field (CF) - Welcome to Let's Talk Retail Excellence, the video series from Retail Insight. I'm Chris field, a retail analyst, and editor of Retail Connections. Today, we're going to talk about on-shelf availability in grocery, which is, I'm sure you'll know, continues to be a problem or, as I hope my guests will put me straight on, an opportunity, depending on which way you look at it. And on that note, we have very luckily today, Colin Peacock from the ECR Retail Loss Group, and that's a group which exists to enable the retail sector to sell more and to lose less. And I like that, very concisely put, thank you, Colin. So Colin, tell me a bit about yourself first of all. And then I'm going to ask you about your role at ECR and just how long you've been looking at this knotty problem of on-shelf availability.
Colin Peacock (CP) - Yes, well, not hard enough to have solved it, I don't think. But I suppose I've been around this area for 35, 40 years. My career started at Gillette as a sales guy on the road, where my primary responsibility was to make sure that product was on the shelf. And that's really carried me through 30 years at Gillette and Procter and Gamble, through to sort of when I retired from Procter and Gamble and sort of moved into sort of the space around ECR, which is a sort of passion of mine, the ECR Retail Loss Group. We do research, we create benchmarking opportunities and we can become a platform for collaboration around the sort of, as you said before, sell more and lose less. So I suppose I've been nearly 40 years thinking about how to solve for on-shelf availability.
CF - And the ECR group, any relation to that event that we all used to go to many years ago, Efficient Consumer Response. I don't know if you remember those in the early days.
CP - No, yes, I do remember those in the early days, Berlin, Barcelona, some of those great meetings with 3,000 or 4,000 people. Absolutely, the ECR Retail Loss Group was a part of that. We were funded by that. And so we started in 1999 as a group and we got a research grant of 250,000 a year to go away and study the problems around on-shelf availability, barriers that get in the way and problems linked to on-shelf availability, like shrink and unknown loss. And so those pesky, non-paying customers who come in and take the products without adjusting the book stocks.
CP- That's been part of it, but also as we've learned, there's a very broad palette of reasons why a product might not be on the shelf.
CF- Okay, well, I'm going to ask you to talk about on-shelf availability in the context specifically of grocery retail. And that's really setting you up to talk about what is it that the shopper means, which could be about product in the wrong location, out of reach, locked up. Some of the things you started to talk about. So if we can talk about what we actually mean by on-shelf availability, if it's not obvious, which I don't think it is necessarily.
CP - Yeah, it isn't obvious, and it's becoming increasingly relevant in both the online and the offline world to think about what it means. And certainly in my experience, the best way to think about on-shelf availability is, is best defined by what the shopper sees. So it's what the shopper sees is the product where they expect to in the right condition at the right time. And probably the parallel for the online world is, is it on the internet or for the online picker, picking in store, is it on the shelf when they find it. So that's really the best way I've found of thinking about it. Of course, there's many ways in which you can then start to measure it. You can look at audits and they've evolved over the years from hiring specialists agencies like Nielsen to go in and take a spot check of a hundred stores for a particular skew and find out how many were there. And that tends to come out at sort of 8% of stores would be able out of stock with that particular skew. And it's moved and evolved to crowdsourcing and where you can actually get your own shopper to help you identify gaps. So there's that sort of audit methods. People are starting to use mechanical methods. So they're using the sort of online picking accuracy. So were we able to pick those items in that many stores, looking at other mechanical methods, such as was the distribution centre able to send the product to the store in a reliable way? And we know that can be sometimes the service levels from retail or distribution centres can be as low as 90% and gets as we've gone through Brexit and other things that can get lower, as the product is not supplied from the distribution centre and it's not supplied by the vendor. But there's other ways that people are starting to look at it, using data, to understand EPASS and understanding unusually low sales movements, and to use that as a proxy for the products not being available on the shelf, because normally you sell 10 bananas an hour. "Hey, why didn't you do so one in that hour?" Statistically, that would be very unusual, might not be down to the shelf being out of stock. It could just be that it's damaged, but that's been an incredibly powerful way that retailers have been able to use their data, to create a proxy for on-shelf availability and the major retailers like Walmart and others are using that to again, just put their business and to think about quality consistently across. So that's another way. And then people are starting to use robots and cameras as we'll probably go on to later to start to explore the gaps on the shelf by the use of cameras.
CF- But the point is there's no single perfect solution. It's a number of ways in order to build up a picture.
CP - Exactly, it does suffer from that as a problem because, you know, when you have something that's where you can't have a metric that's inarguable. You will inevitably have doubts within a CPG organisation or within a retail organisation. Is that a real number? And I don't think it's a real number. It doesn't really make sense to me and you therefore get trouble sort of aligning an organisation behind that number. And that's where the problem can begin. Unlike many other lines in the retailer P&L, let's take shrink or let's take labour, the clear number and you've got a target to go for. And there's normally someone responsible for workforce planning or for shrink, but from shelf availability, it lacks, you know, a single owner. And that's partly because of something that a CEO of a retailer told me that it's something it's everyone's job until it becomes a crisis immense. That's when I think sometimes you do get some leadership, but when it's everyone's job, it kind of is no one's job, or it can't be anyway.
CF- Is it possible to describe the scale of the problem? I mean that's a rotten question, of course, because it depends on which grocer we're talking about in which country, but I just, I just want to communicate to our listeners that it is a big problem. And I think later you get into this knotty again, a big problem of a grocer thinking, well, it looks like I've got that stock, but actually it appears that they don't, whether that's because it simply isn't really isn't there or is in the wrong place, not available, et cetera. So can we talk to you about the scale of the problem? 'cause clearly there's a lot of focus on it and has been for many years.
CP - Yeah and I think as people look to that sort of certainty in terms of, well, how big is it? You know, the industry tends to gravitate to some very old studies by Carlston and Gruen, who point out that they did a study of 55 studies in about 2004 I think it was where the number was 8.3%. And the industry tends to use that number and they tend to then sort of use the other data within that. Well, what happens if the shopper doesn't find the item? Surely they just buy something else. Well, they looked at sort of some analysis in terms of switching, you go switch to the brands or switch to the other retailer. But the conclusion they came to was that roughly 4.3% or half of that out of stock turned into a real loss sales. But the truth is, we've seen huge variation across stores in terms of on-shelf availability within any one retailer. They can be a store that's running at, 95% of the stores that are running at 80% and that often can be down to the circumstances of that store. It could be that one, the first or the last on the route. Then you see variations across categories. So some categories have much higher levels of out of stocks than others. Ice-cream and other items are very seasonal. So suddenly it's a hot day and they all go out of stock or burgers or, so it's a problem. I see it as a quality measure, possibly the ultimate quality measure, for a supermarket operation to have an on-shelf availability number and a metric. And again, you'll find huge variations across stores and within any one store across departments, but we have 8.3% number. Other people have used numbers such as a third of shoppers don't get what they came in for. That came from a very old study that Marshall Fisher and others have been able to find in this world of consulting. I think CapGemini were responsible for that number.
CF- I'd like to add some context as well, which is one of the many reasons that we're having this conversation now, which is the context of multichannel retail or whatever we want to call it, which is that a grocer is dealing with third party pickers, particularly in the US and indeed its own shelf picking for online orders, which is adding complexity in all sorts of areas. So hopefully we can sort of get into that later because on-shelfability for, you know, which is the customer, 'cause there's almost three customers. There is you and I, if you like, and then there's the store pickers. And then there's these third parties that come in and do the store picking, do the picking as well. So that's the kind of context I'm interested in, because it sounds like a big problem is only going to become bigger.
CP- I think that as we've seen retailers have had to ramp up their online availability, thanks to the pandemic. So we've seen huge increases in capability that they've been delivering. I think Tesco's famously built an online business, the size of Ocado within six weeks with their Click+Collect, to respond to the crisis. Inevitably there are sort of problems that do come from that. You've designed a planigram for a catchment area and a volume that was X. And suddenly now you've got click and collect, which means your catchment area is expanded at twice. And so you now got a very much bigger potential of volume, products that disproportionately go onto online, such as nappies, diapers as they say in America. Suddenly the shelf allocation was X, but now the demand at certainly at 3 o'clock in the morning is exponentially higher. So the shoppers come into the store in the morning, and it's all been picked for the overnight shop for the online shopper, by the store. In the case of third parties, Instacart and others, they'll come in and they've got a very specific brief and fresh products. They're looking for the products with the longest shelf life. So those customers who are looking to buy steak, who have made the journey to the store themselves, will now be faced with possibly an out of stock, but then possibly very short shelf life, because the Instacart shopper has been in there and they've picked the longest shelf life for the customer who sat at home in the armchair. So it's a really big problem.
CF- As you'd expect.
CP- And some retailers stores are doing, well I know some stores, where 50% of the sales from that particular box are now are going out through online sales. So it's huge. On average, sometimes it can be around 20%. So 20% of your volume going out through either your own people picking it or through a third party picking it. And I think it really comes back to a central quality issue that retailers have around inventory record and accuracy. And this is probably at the heart of the problem. And the biggest challenge for retailers is to get those inventory records accurate. So at ECR, we-
CF- This was my next question for you. That was my next question for you.
CP- See, you're ahead of the gate.
CF- Yes, well, it is a tough question, who has the true, can you get a true inventory position, because presumably it comes down to, well, who do you trust for that figure? Given, as you were saying earlier, that there are so many inputs.
CP- Yeah. Great question. Let's say you have RFID on everything and then you can trust the technology, but then do you really trust the technology to be able to read every single tag? But no, I think that people tend to rely on external independent counts as the sort of truth. And people might increase the frequency of those counts by external auditors or internal specialists in order to sort of bring back to what they believe should be a true inventory position. There's certainly in the industry, a feeling that the more store counts, the more you corrupt the accuracy of the inventory. So I think people sort of rely on their specialists, the external auditors, like RGIS or WIS or whatever to be the arbiters of truth. We've certainly seen some retailers who've increased the frequency of their accounting from maybe once a year to six times a year. And actually they've got a very nice sales lift by doing that, because every time you true up the inventory records, you'll get a sales lift because the inventory is where the store thinks they're sitting on 12. Physically there's nothing, you then true it all up and suddenly you trigger a replenishment order and that item then starts selling, because it's on the shelf again. But you've really zoomed in, I think on one of the biggest causes of the problem of shelf out of stocks, which is the system thinks there's inventory there, but actually physically the inventory isn't there, but because there's no sales going through, that inventory position will remain static until it's adjusted. And then that adjustment takes time for someone to spot that there's an inventory problem. And then there's approval processes for it to be adjusted. And therein lies some of the reasons why you see shelf out of stocks.
CF- Yeah well, this takes us very neatly and thank you for all of that context too. And now we live in the world of listicles and people like things divided neatly into lists and thank you for that. So we're going to get into what can actually be done about all of this. And you've very kindly divided that into eight steps, which we're going to go through if I might, and it starts with why now? Why now? How does an organisation knows this reached a kind of tipping point if you like, a point of no return and what are the metrics that are driving that sense of urgency?
CP- Yes, I mean, when I've looked at retailers and sort of seen, oh they've got a big push going on in on-shelf availability, is often that they reached a crisis often triggered by news media. I think there was one famous example with Tesco. And there was a time when not that long ago when they were going through some struggles, the change of CEO and others, but one of the things that new CEO, Dave Lewis, identified was the on-shelf availability had become a big problem. And so he had recognised that this was the point of no return. And if you remember at the time, there were lots of activities that Tesco put in place in terms of extra staff in the fresh produce aisle, new metrics in the organisation, the idea that these are the 2000 items that will always be in stock of, those are the fastest sellers. So there are a number of things that CEOs put it best, but there has to be that recognition, at least at the top. And then as I say broadly across the organisation, that there is a problem here and that there needs to be a turnaround. And if we don't do a turnaround, then we will lose trust with the shopper. And sometimes it is triggered by the media, just the consumers, just say, "I can't get what I want." And the consumer does have a much bigger voice these days on social media and other, so some of those metrics are just shopper metrics and some of those metrics will be "Why am I losing branch here?" Well, we asked the shoppers what they found most frustrating when they visited your store and why they don't come back it's because they couldn't buy what they wanted. And so these are really important changes that need to be thought about, but organisations, if they want to get started on the on-shelf availability journey, I think probably need to reach a tipping point and a point where people say enough is enough. And so that will be the first of the eight points on this listicle.
CF- Yes you've got a source cutter I believe who has a framework on organisation change?
CF- Can you take us through that? It sounds very technical, but actually it's not. So bear with us everyone because he has the perfect answer.
CP- Yeah I'll do the one minute version of it. I mean, it's basically just sort of start to bring visibility to the problem, put together a leading coalition of actors, a small group of people, who are sort of going out to champion change, identify where those barriers are to change that are going to get in a way that's sort of the terrorists if you like, who are going to sort of stop any progress, continue to document the small wins, sort of small wins and start to build it into your reward system, use a DNA system. It's a marvellous framework. And certainly something that's worth looking up for any of the viewers too, or just drop me a line and yeah, it's great. So, and the background to it was that he studied, I think about 75 organisations and how change projects did or did not work. And so these were, if you like, the things that you need to put in place to make successful change happen, and it's about sort of-
CF- And you're pulling people in, yeah, sorry. I was going to say but it does require from a cultural point of view pulling people in from quite a broad set of merchandising, supply chain, human resources, even finance, et cetera, et cetera. There's quite a broad set of people that need to be brought together to do this.
CP- Yes and of course there can be competing objectives of course. I mean the supply chain team are working on running a very cost efficient, fast operation. And when you say, well, actually if you could slow down a bit and make it more accurate and make the deliveries more frequent, that's counter to the objectives and the goals of the supply chain. Equally, when you say to stores why don't you just hire and put some more labour in this section? Well, again, there's a targets being set for how much labour you can have as a percentage of sales. So there's also, I think we should probably mention this as well you know, so one said to me, "Hey Colin, you will never get to a hundred percent availability. We can't afford it." And whilst that was referring to the whole store in general, specifically in the area of fresh, if you are a hundred percent available, if every single bakery item that has a one day shelf life at 6:00 PM, that's an awful lot of bread you're going to have to throw away. So we see retailers now starting to be a bit more selective in terms of, well, if we got 95 bakery items, maybe I'll just have five of those available at 5:00 PM rather than the full 95. So as to sort of get that balance between food waste and availability, so these are some of the sort of balances and trade-off. So it's not as sort of simple as we can, let's go after a hundred percent availability. And it does require organisations to think holistically across all functions. So that you're all gearing for the sort of one target of on-shelf availability.
CP- Go on question three.
CF- Well, like I say, I want to talk about how the organisation, the grocer can get, as you've said, inarguable metrics to measure and track online shelfability. And as we mentioned earlier, online availability.
CP- Yes. I suppose it's going to be, it's always a tricky one, right? Because everyone sort of says, well audits, there's a limitation of audit. That was just when you went in there, you went in there a Monday morning, not Friday when it's busiest. Then there's going to be cameras. Well, cameras where you can't check for the gaps being properly filled, then there's the sort of using data. Well, there's limitations on the power of the data because there's only sales one per day or something. And so to wait a long time before you could declare an out of stock. And I think the answer here is rather than search for one inarguable metric. I think the answer here is to seek, a number of metrics and start to look at a number of perhaps not perfect measures, but measures that can give you a sense of the trends and the movement. So you definitely pick, you look at your EPOS measures, you look at online availability measures in terms if you're picking from store, just what percentage of orders were picked the first time without substitutions. You might also look at customer service measures that you have on the checkouts where you ask people, did you get everything you want? So you think you need to have a look at a number of different metrics, but probably the one that is the most universal that covers all the categories is data. And that seems to be the one that's, at least all retailers have as they use their EPOS data to determine is that item there or not? It's a good proxy for availability based off the historical sales data.
CF- But the point I pick up from that is it's about ruling everything in, it's not really about saying, "Well, that old measure doesn't work. We're now using technology to move to something else." Actually, a retailer's data gathering and processing abilities are such that really they can process quite a lot of inputs, quite a lot of measures in order to get a better view.
CP- Yes yeah, absolutely. I think we have to live in that sort of framework in as much as, you know, as a CPG working for Proctor and gamble for a number of years, we wouldn't rely just on one health metric for the brand. You'd look at a number of different metrics, sales, brand share, consumer expertise, numbers from consumer research. So the health of the brand would be measured on a number of different ways. And I think the health of the shelf also has to be measured on a number of different ways. So just to rely on one, yeah, I think would be a mistake. I don't see retailers doing that thing as well. I do see them looking at a number of different metrics and I see cameras and robots as additional and more information that can give us more colour to the problem of shelf out of stocks, so rather than replacement.
CF- Great. The next question, number four, it looks like something to do with the almost the will to change back to the sort of cultural thing, which is where does improving the online and on-shelf availability problems sit in the context of a company's current and future corporate priorities?
CP- Yes, yeah, no, I think it's just good enough. Can sometimes be we've got a burning fire. We need to build up our online capability. Let's just get some sales first. We'll worry about the missed sales later, or let's just build some more stores. We'll worry about the lost sales later. Let's spend our IT time thinking about computer vision and how that can help us understand what's going on in the store. So yeah, I mean, the corporates have many, many priorities, and this perhaps goes back to the first point about, unless this one really gets to the top and you've got to a tipping point, sometimes on-shelf availability is just, well, that's just what we do as everyday business. We don't need it to be a top priority. It's just, that's what we do. We live and breathe and we're here to put stuff on shelves and sell it.
CF- Yes, yes. I can see, particularly in a competitive environment, which let's face, set in the UK it's a fairly killer environment but I'm just focusing on, which is the nature of grocery after all. It's all about the tasks in hand, trying to find time to kind of stand back and think about this is isn't easy.
CP- Well, and if your sales comp is growing 10% a year, so we're up 10% versus last year going after-
CP- We're doing something right.
CP- Yeah we're doing something right, and well, if there was an empty shelf, well maybe they bought something else. Do we really know how much we're losing? Can be one conclusion that retails have reached, and retails are all in different places. Some of them sort of say, "No, no, no, this is really important. We need to measure it. It's a vital health metric for our business and it needs to be right on top of the corporate scale." And then there's others who will have a sort of slightly different view. And that will change over time as different management teams come in and reset priorities. But in the context of lots of other things going on, it can be, well it's just business as usual. It's not a project. It's business as usual, get some clear metrics around it and make sure we can.
CF- Yeah. Next question kind of goes back to the different ways that are available to measure some of this stuff, which is, is the board, so again, from the very top, aware of the required trade-offs and barriers and improving these measures, so that they know the extent to which they're open to investments and choices required to deliver some of those improvements.
CF- Yeah it's a real, it's a big investment right? All of those, I was going to ask you that earlier, actually, is that cameras or robots, et cetera, et cetera. That will start to add up to quite a hefty investment, which then comes down to well what's, it's the opportunity that. We will come onto the opportunity by the way. I think you were speaking to us, sorry, I was going to say you were speaking to a senior retailer recently about, this goes back to your hundred percent on-shelf availability. Well, is that even affordable anyway? And am I just going to be chucking stuff away? And I wanted to keep going with that because that is becoming an issue because the consumers are saying, why are you throwing all this stuff away?
CP- Absolutely. It's this trade-off between, as we said earlier on, having it on the shelf in bountiful quantities, but then having to throw it all away because we wanted to achieve our hundred percent availability target. I just don't think that washes with the way we work in grocery these days. So there has to be some trade-offs, but there's also other trade-offs. We know that complex, you make the range and assortment, the more stores will struggle to execute. So there's a reason why Mercadona and Costco are such good retailers. They're very disciplined in their range and they'll carry very strict assortments. Whether where you see other retailers just, there are no bounds to how many items they'll carry, you know, how many tomato ketchups, or how many butters, or how many shampoos. But all of that puts pressure on the stores. And it leads to sometimes retailers having a productivity index of 150, which means there's 50% more work than they budgeted for the store to do, so increasingly these needs to be the sort of trade-off between the buyers and the assortment that they believe is right for their shopper that they need to have, you know, and stores and the ability for them to be able to sort of execute without any flaws. And so these are some of the trade-offs that I think the retailers need to sort of take into account. Again, I'll give you the supply chain example. Distribution centre managers can send a lot of products very quickly to stores at low cost, but it might come with an inaccuracy measure. Same with self-checkouts. Self-checkouts are great. They speed up the customer service. They can help with productivity, but we know that with self-checkouts that there is a level of inaccuracy that creates shrink. And our research indicates that for every 1% of sales that go through fixed self-checkouts, there'll be an extra basis point of shrink, but more importantly, that extra basis points of shrink translates into, "Hang on, we've got a negative variance in our inventory, so we've got less than we think we've got." And so that's inevitably going to lead to shelf out of stocks because the ordering will be late. Replenishment will be late, and then the shelf will be empty by the time the customer gets to see the shelf or online picker comes and picks it up. So these are all the trade-offs that need to be considered if you're going to run to a sort of maybe a sort of near perfect, on-shelf availability programme or shelf quality target.
CF- And you mentioned the brands. Of course, they have to be part of that conversation, don't they because they're in a competitive market where they're trying to introduce new products and get what have in many cases become very complicated ranges. And we've been reading about the likes of Kraft and others who have recognised the need to simplify, for other reasons to do with brand equity and clarity with the consumer. But presumably that is an issue to a retailer that wants to introduce some simplicity in terms of the ranges to make them more manageable.
CP- Yeah, I mean, I can, if you're a sort of a grocery retailer and you want to sell cosmetics, you're going to compete against Boots and Boots have got a tremendous range. And so you're going to feel compelled to carry a big range of cosmetics or shampoos because that's what the competition do. And that's what you need to do in order to be able to compete says the buyer. Of course the store operator might say something different. We don't need as many of the, some of these never sell one a month, one a week. So they're all slow sellers and they'll jam up and they gather dust. So it is problematic. I mean, in Europe, there's less reliance on perhaps the direct Salesforce moving in and talking to store managers in America, I think still 20 to 25% of the volume that goes through an American supermarket is supported by a direct store delivery representative. So the vendors over there are certainly not taking control of the shelf, but they're certainly really helping to make sure that shelf is full. But yeah, that's a trade-off.
CF- Right now, more politics in a sense, but I'm going to ask you who, and perhaps which function as well is going to be, or is currently the single owner and champion of these on-the-shelf and online availability programmes.
CP- Yes three different ones, really difficult one. If you call it lost sales, maybe it should fit within the area of the store operations to go and chase after those lost sales. After all one could imagine, that's where most of the problems lie is because the stores just aren't adjusting the inventory and they're not putting stuff that's at the back of the store in the shelf. So there's an argument for store operations. Then there's no argument for supply chain. Well, they're the ones that are supposed to be supplying stuff on time, they understand the metrics around service levels. They should be in charge. Then you can sort of say, "Well, the marketing people is core to our business." We need to, and so I've seen all those functions take ownership and that's not unusual. I mean, even if I think about the problem with shrink, sometimes it fits into the general council. Sometimes it fits into HR and sometimes it fits into store operations, sometimes it fits into finance. So there are these sort of problems that transcend an organisation, where it's difficult to find a home for them. And probably there's no right answer to this because ultimately this is something that everyone needs to take part in. So there needs to be some sort of certain leadership from the top to say called out, this is important, says the CEO. "We need to do something about it. Let's put some decent metrics in," and everyone needs to then play their parts in helping to deliver it, from HR, hiring the right people. Maybe you need to train people in different ways, technology, you need to provide ways in which technology can help from simple sort of perpetual inventory management systems to the ability to be able to see the inventory on your smartphone and the sort of mobile device all the way through to sort of loss prevention and supply chain, et cetera.
CF- It does sound like that, yeah, no, I can see that. Well, I was going to say, it sounds like, yes, good process is one thing. But it sounds like you still need a good old fashioned strong, capable, charismatic leadership, if you like to kind of push these things through, because it sounds as if it absolutely must be business as usual, but feels like it's kind of slightly to one side, if you like, in favour of just sales and operations and getting stuff out there.
CP- Yeah, so there are there, I mean, retails do put in again, if you're into a turnaround situation, so you've declared a crisis, you might have a task force and the CEO or someone senior might share that task force. You don't shuffle with the task force, but there are all sort of organisations that have a sort of centre of expertise. Here's a sort of think tank within central operations with a charismatic leader, who's passionate about on-shelf availability. The on-shelf availabilities are, and they can make an enormous impact in the organisation by helping the organisation see the problem and joining everyone up and trying to encourage best practises and sharing of ideas and exploring those data relationships, but also owning the metric. So a sort of good company policy, perhaps is also required here around what is our company policy on on-shelf availability. So that might be something that we've certainly seen work very well in shrink, where an organisation declares, who's accountable for shrink, what the metrics are, what the methods for improvement are, what technologies should the central organisation support.
CF- Okay, we've kind of covered seven, I suppose, which is this, to what extent will this become embedded in the organisation and therefore prioritised in decision-making? So I think we sort of covered that in terms of the need for leadership and good process and an understanding that it does work across departments.
CP- Yes I mean, I've seen some, I mean, on the problem of shrink, the way this has sort of manifested itself very tangibly is that every single function on shrink that this particular retailer in America worked for, they had 25% of their bonus tied to an improvement in the shrink number. So one way to bring the number seven alive is every single function gets 25% of their bonus tied to an improvement in the on-shelf availability number. Another way would be to give every store manager an on-shelf availability badge with a number on it, I'm at 95% availability, I'm at 92% and the lower your number, the closer you get to sit to the CEO at the annual dinner.
CF- Yes, I guess the rewards, the rewards can be significant because of the size of the opportunity. I don't know. Is it possible to define the size of the opportunity? Because I would have thought that if 50% of a particular item appears to be available, but actually isn't, then spotting that at least puts you in a position of saying, "Well, look, I need to fulfil the 50% that simply isn't there that therefore never gets sold." So is it possible to describe the size of...
CP- Yeah I mean, I think that, the numbers I've seen typically if you get on-shelf availability right, there's a 2-4% sales uplift per year for your business. You can sell maybe an extra two weeks of business, 52 weeks becomes 54 weeks. That sort of order of magnitude is not inconsistent with Carlston and Gruen but I don't think anyone in retail or anyone in CPG, wakes up and thinks I'm going to keep that shelf empty today. Everyone is absolutely motivated to serve the shopper. And there's just a lot of things that can get in a way of the product being on the shelf at the right time at the right place. Most importantly, I think it's around this inventory record accuracy and keeping those inventory records absolutely bang up to date. And how you do that? when we know that when you ask stores to adjust inventory, they don't always get it right. Is that a bit of a conundrum and RFID I think in grocery might not be an immediate solution in my career.
CF- Last question is where I suppose it's kind of a where to start, because clearly can't boil the ocean as you say in your notes, but is it simply about picking the most high margin, fast moving categories like fresh or what would you advise?
CP- I think you have to start, I think you have to start with waking up in the organisation to the realisation that there is a price to go after. So I think putting in place some form of clearly understood metric maybe around the top sellers is a great way to start, probably for scale and speed using your own EPOS data to come up with "Hey, we didn't set any of these this week. We must be out of stock." Simple, not simple to do, I don't think, but those sort of EPOS based metric OSA measures I think is probably the simplest way and start to declare, you know, this is the size of the opportunity, and this is the variation. It's the variation I think that everyone needs to start getting into why is it we got a problem with ice cubes, why we got a problem in this part of the country associated with this distribution centre. Why in this format stores, we've got a problem on this category. So I think it's about establishing a sort of relatively simple universal metric I would say, like using EPOS data, looking then for variation and getting the organisation behind, maybe some of those major sources of variation and doing some deep dive, maybe on one or two categories, is it we've got too much assortment? Is it because whatever? So that's why.
CF- I'm going to, I'm going to, I'm going to finish on technology, not because I want to make technology the hero. We've always understood that it is an enabler of the changes that are put in, and we've talked about management and culture and the kind of the will to change. But I am interested just to sign off and thank you for all of this Colin, fantastic insight, is some technology. So let's assume you've got everything in place. What are some of the technologies, I think you've mentioned cameras, that kind of thing, that can then allow you to go a bit further?
CP- Well, look, I've been around a while. So I remember RFID was one of the technologies that was really talked about in 1999 onwards. Yes, in five year's time, everyone will be using RFID. That was in 1999. Somebody even said that RFID will be bigger than the internet, but in grocery supermarkets, I'm not so sure within my career that RFID will be omnipresent is that some of those visions sort of suggested, but I think it may well have a role to play in some of the clothing areas. And we're already seeing retailers like Walmart and Tesco using RFID in their apparel sections. But so I think that that would be, and that's really played an important role RFID improving accuracy from 65% of inventory records being wrong to 97% being right. So that's been really good. What other technologies? Well, I guess there's going to be people talking about the use of cameras either mounted on shelves or robots, that's certainly something that's of interest. There are some limitations with those. There's also people looking at computer visions. Overall one camera looking perhaps over on the produce area. And that's another area. Another technology I think could actually have a very profound effect on fresh, will be the use of 2D data barcodes, where the expiry date is actually built into the 2D data barcode. And again, that could have a profound effect because it tells you the age of your inventory and then fresh, that's pretty critical for your forecasting to know that, you've got 30 chickens there, but they're all going to run out tomorrow. It means you better quickly get some more chickens there. So, I think that would be quite an interesting one. And there may well be some sort of smaller Bluetooth enabled RFID that may come along and be ultra cheap. So that's, I think would be something else. I mean, I was talking to a retailer in Australia this morning before our call here. One of the reasons why a shelf might be empty might be because someone's stolen everything. And as a retailer, that's using a software, very simple at Edge Algorithm Software, that's able to detect unusual movements from the shelf, and they've been able to reduce their shrink they say from 1.3% to 0.3% by deterring the prolific offenders. So if you haven't got that disrupting your shelf and making the shelf empty, that, again, it's a computer vision I suppose, is we're coming back to being something that could really help us understand what's actually happening on the shelf from being from a theft point of view. And therefore we can quickly correct the book stocks through to the cameras being able to tell us where the gaps are on the shelf with image recognition supporting that. So it's this specific ones in the wrong place as well. So I'm quite hopeful actually that computer vision, cameras, video will play a more increasing role in helping us and supplement the existing measures we've got such as online pick accuracy. So was the picker able to pick first time? And then the EPOS based measures that we've been certainly reliant on the students very well. You sell 10 bananas every hour, you sell none of that, that particular hour, probably it was out of stock. And then, with the cameras then we'll sort of say, well, they're out of stock because they're all off, they're all damaged. And so they'd be able to put everything together so that eventually you can start to influence upstream what the replenishment order is based off all these different senses to sort of say, well, accuracy problem here, and actually, should we bother with inventory records and perpetual produce? Can't we just rely on what we're seeing in the store? So I know there's some school of thought that says that's possibly a better way of generating a forecast rather than perpetual inventory.
CF- Right, we don't really have time for that, but that is interesting, of course, because then you're talking about technology that is able to, if you like, bring those different departments or disciplines together slightly more.
CP- Yeah. So I know an experiment that was conducted by a CPG. They put cameras into cabinets and they were looking to understand with the cabinets that had a camera that was linked to an ordering system. Would that be able to generate better, more accurate orders than the store manager who came around with their pattern then created particular orders? And they showed a 30 to 40% improvement in sales when the camera did the ordering, rather than the store manager, because the store manager comes with constraints, not least the fact they can probably only do the ordering for the store once a week, which means that if you go out of stock on the first day, you're out of stock for the five days, but also they also have constraints like I've only got so much money to buy stuff. And so not order any more of that particular product, because I've got some other ones that are similar. So yeah, so they saw a 30-40% increase in sales, not relying on the inventory position, but just relying on what the cameras could see. Boom Boom.
CP- Cameras are the future.
CF- Indeed. Colin, thank you. Thank you for your time. Very, very useful. There's a lot of information there. I hope that you've enjoyed that and that you'll get in touch with us if you wants a little more information and please do tune in again for more insights when we come back to you next time. Thank you very much. And Colin, thank you very much.
CP- My pleasure. Thank you.