ECR Retail Loss
Enabling the Retail Sector to Sell More and Lose Less
The latest discussion in our self-checkout working group, attended by over 120 participants, revealed ground-breaking advancements in retail technology and improvements in the Product Lookup Menu (PLU) process. As retailers move towards sustainability and an increased offering of loose products, the importance of scales and a PLU menu becomes more important than ever. This shift not only addresses environmental concerns but also brings unique challenges in inventory management and customer experience. One of the key insights discussed at the meeting was the significant role AI is starting to play in enhancing the accuracy of product identification. Retailers are adopting AI-driven systems at self-checkout stations and weigh stations in the aisle which are proving increasingly effective. Some of these systems have demonstrated remarkable accuracy rates, as high as 99.96%, in recognising products like fruits and vegetables. Such technological prowess not only reduces the likelihood of mislabelling, said to be 4% by one retailer, but also streamlines the customer's shopping experience. However, discussion at the meeting also highlighted the ongoing issue of product misrepresentation. One retailer noted selling more carrots than they actually bought into the business! Some customers are clearly mis-identifying products at weigh-stations and at SCOs, which underscores the necessity for more sophisticated recognition systems. The discussion also touched on the challenge of distinguishing between organic and non-organic produce, suggesting the use of stickers for easier identification. As the industry moves towards more sustainable practices including using brown paper bags, the technology must adapt to maintain its effectiveness. The video below recaps the key findings, with the transcript below. Colin: Well, this week we had our self-checkout group. I think we had over 120 participants registered for a discussion around the product lookup menu. Increasingly, retailers, in order to move away from plastic, are looking to use loose products, perhaps, fruits and vegetables, bakery in other areas. And that requires increased use of the product lookup menu, which is integrated into SCO, but also, we know it's also in the aisle as well. And as we've documented, this is an area of vulnerability for retailers in terms of mistakes and errors. But we also know it's a huge consumption of time and effort. And we have also learned with our online grocery team, that when online pickers look to pick loose items from customers online for their online orders, again, they have to use the product lookup menu, so it's also a significant point of friction for online grocery pickers in the stores too. So, we've got all that going on. It really was a topic of very high interest, hence the number of people registered to attend and who participated. And we had some excellent speakers sharing what they're doing to try to at least detect some of the problems that might be arising and helping to make the whole process more easy and faster. What did you take as your key notes? Adrian: Yes, it was a very popular session. And I think the reason for that, Colin, is it is a real and present issue, I think, for all the grocers who are using self-checkout. The vast majority of whom expect their shoppers to select loose fruit and veg. It's one of their attractive propositions of groceries - you can choose your own. Isn't it? So, it's been a long part of the use of self-checkout. How do you get the customer to weigh the products that they want to purchase, but critically also to identify what they are, so they can get the right price. It's now almost become legion, hasn't it, in terms of arguably one of the most common ways in which people abuse SCO is to misrepresent one piece of fruit for another, or one piece of vegetable for another. Colin: [joking] Yes, everything looks like a brown onion. Adrian: Exactly, and I think a number of retailers have shared over time now just the extent to which this is potentially being abused. There was one retailer who came out with a startling statistic, that they sell more carrots than they actually bring into their business. Which was highly indicative that perhaps people are choosing carrots when they may not have been carrots. We know it's an issue, but it's not only an issue in terms of abuse. As you rightly say, it can be complicated for customers to use these. When you have potentially, some readers could have upwards of 50, 60 different types of fruit and vegetables in their stores, try to organize a product lookup screen that enables you to get to that papaya or whatever it is, that sweet potato that you purchased. It can take time. There can be multiple menus, can't there, for people to work through, which isn't easy because nobody's ever trained us to use these technologies. There's that issue around just how difficult this can be for customers to do that, and there's the potential abuse as well of, as I say, the older carrots for grapes trick. Then the third element that you mentioned was this productivity piece, which actually came up in the session, and it wasn't something we'd really thought about very much in terms of that audience benefiting from this. Of course, what a lot of the interventions and solutions focused upon this are to do with product recognition, aren’t they? They've got to be able to say, that's a banana, and I can choose that for you on your behalf from the product lookup. That is the critical element. How well can these video analytics systems identify fruit and vegetables accurately to be able to deal with this issue. And so we heard from three retailers who shared their experiences. The first one was slightly different in that it was mainly focused upon the weighing of fruit and vegetables away from self-checkout. So it was related, but it was different. It was in their in-aisle weigh stations, wasn't it? And they were using a very simple system, which was a camera above the weigh station. And it had AI linked to it. And it had a learning algorithm in there. And they were using it simply to look at what was in the weigh tray and say to the consumer, they're bananas. Is that what you want to buy? And they claimed they were getting 99.96% accuracy of the system, recognising the things that were put into that weigh scale. Colin: So that was a remarkable statistic. What was also fantastic about what they did, I thought, was they were able to put it into silent mode and actually work out how many people were mislabelling, which was incredible. You know, so I think the statistic was around 4% were being mislabeled, which is a lot of carrots or a lot of, not carrots. Or whatever it is. Adrian: Yeah, that's right. I like the surveillance mode ideas you can have with these things, because what surveillance mode is really good at this silent mode is it really helps you to get a measure of the size of the prize, doesn't it? Because you can see - and retailers often don't like it, or at least don't like having it on for too long - because they can see the loss. Visibly across the screen. But it's a great insight into just how much is happening. They were recognising something like 4% of the products that were put into that weigh scale were incorrectly chosen from the product lookup. So that's a really big win just in terms of reducing that. But they also talked about the reduction in customer time. They reckoned it would reduce customer time by as much as 80%. Just because if you put the bananas in it, it just says bananas. And all you have to do is say yes. And that's so much quicker than going… bananas, need to find bananas, click on bananas, and then print. So that was really good. And then the other two were looking more specifically at how they were using it at self-checkout. So it was integrated more into their fixed SCUs. And that was a little bit more checkered, wasn't it? I think one had gone through a number of iterations of this technology, trying to get it to be sufficiently accurate to offer value. And then the other one was a pretty archaic system, really, which it wasn't very good at recognizing things other than if it was perfectly square or had square edges. So it's trying to identify where people were putting things on the weigh scale that were patently not fruit and veg. There were very few fruit and vegetables that were perfectly straight and have got square edges. And so that system was trying to say, you've put a bottle of wine on there. And you're claiming it's bananas. We don't think that's the case. We saw different ways of doing this. But there were lots of other retailers on the call who were clearly very interested in this and trialing it out as well. Because I think everybody recognises there's much work to be done here. And a lot of benefit can be derived from these much improved product recognition systems that are coming to market. Colin: Yeah, as you say, I think it really was an area where there will be much more. And that's why I think we said we would anniversary this meeting. And the cloud did seem to make a big difference. I think we heard a couple of times if you can actually have all this intelligence. Every machine learns at the same rate. Yeah. And perhaps that prize of, you know, recognition, and speed, might trump some of the other reasons we might want to do this around the sort of malicious types, you know, putting Lego on the scales. I don't know how often that happens, you know. Adrian: No, but a couple of other issues came up. One is: what do you do if you want customers to put things into bags? Obviously, there's a drive to move away from C2 plastic for sustainability issues. So if you're saying to customers, we want you to put them into brown paper bags, the best product recognition system in the world is going to struggle to look through a brown paper bag and see what's in there. So that's an issue that needs to be addressed. It's not the end of it, but it needs to be thought about. And then of course, there's a thorny old issue that we always talk about, which is organic versus non-organic vegetables and fruit. And how can these systems try and possibly identify the difference between an apple that is organic and an apple that's non-organic? And I think there was basic recognition that you need to help the system in that place by having some sort of sticker on there. So it can actually hone in on the sticker and say, I can see that that's organic because I recognise that particular sticker. Otherwise, I doubt whether there's going to be any systems because the idea that there are subtly different shades of green between them is asking a lot. I think it's much easier if you just put a sticker on these things. Colin: Yeah, hopefully you can just grow the sticker on there as well. Yeah, it grows with the sticker on it. But this seems to be the solution for one retailer, although I guess putting a sticker on it is an extra cost. That’s one way round it, but the brown paper bags, again, are a different thing. Well, thanks for that, Adrian. We're next together on December the 13th. There should be a QR code somewhere. And on December the 13th, we're looking at self-checkouts in non-groceries. So we've got a couple of retailers who are saying, yes, it's working. Here's some differences. And a couple saying, well, it didn't really work for us you know and for these reasons. So it'll be very interesting to hear the different travails of retailers as they have experimented with self-checkouts outside of the grocery. So looking forward to that. Thank you again and have a great weekend. Thanks Colin, bye bye. Adrian: Bye bye.
ECR is on the search for new innovative ways and technologies that can help retailers and CPG's improve OSA. The search starts with a clear set of problem statements. These problem statements emerged from interviews and discussions with the OSA working group members, made up of the OSA experts from the retailers and CPG's represented in the group and the academics who support the group, who have have in depth knowledge of the OSA problem, including Professor Daniel Corsten and Thomas Gruen. Broken down into four areas, upstream design, supply chain, back room and shop floor, there are in total twenty two problem statements, where new innovation, technologies and approaches would be welcomed. By mid-December, the retailer and CPG judges will have received a curated long list of innovations that meet the briefs. By early February, the Top 30 OSA innovations, as rated by the judges will be published. On May 22nd, the top ten will pitch at the online showcase finale. If you would like to help with the short-listing process, please email email@example.com
Following our recent OSA working group meeting, Professor Daniel Corsten shared his takeaways from this well attended (41 retailers / CPG's) online meeting and the key insights on the challenges and strategies crucial for retail success during the festive season with Colin Peacock. Here are some of the key themes: Importance of Execution and Discipline Execution and meticulous planning remain key themes when it comes to how retailers plan for the holiday season. Daniel pointed out that, despite technological advancements, the bedrock of successful festive retailing remains rooted in these fundamental principles. A focus on delivering the essentials well, rather than an overreliance on complex tech solutions, emerged as a key theme. Empowering Store Managers: A common theme shared by all the retailers was the high degree of autonomy given to store managers over this festive period, on both the assortment and on the planning of the retail shop floor. This strategy allows for more localised, customer-centric merchandising however, there was also a recognition that it can bring extra complexity in planning and inventory management. The balance between autonomy and operational efficiency was a focal point of the discussion. Daily Coordination and Communication: As Christmas nears, the need for intensified daily coordination in retail operations becomes critical. Retailers reported shifting their planning and communication from a weekly to a daily schedule, a strategy likened to 'air traffic control' for its precision and necessity. This heightened level of coordination is vital for managing the festive rush effectively. Post-Holiday Strategy: The transition back to standard operations after Christmas involves strategic markdowns and a reassessment of inventory. This phase is crucial for ensuring a smooth shift out of the holiday mode, balancing stock levels, and preparing for the upcoming business cycle. Vendor Challenges and Solutions Finally, discussing the unique challenges faced by vendors, Daniel also picked up feedback from retailers on the need for creating durable, 'Christmas-proof' displays - which arrive ready to use with minimal store labour hours required to place on the floor. This insight reflects the practical challenges of the festive season, hectic, busy and not enough hours to get everything done on time. The OSA meeting provided a number of suggestions for retailers aiming to maximise their festive season performance. You can read the full transcript of Daniel and Colin’s meeting below or watch their discussion here;. If you would like a recording of the whole meeting, please email firstname.lastname@example.org Colin: Well it must be about two or three weeks ago when we had our On Shelf Availability Working Group meeting. We were actually talking about the festive season. There's the holiday season where sales in grocery stores, especially goes through the roof. And it's sort of make or break time. It's one of those real adrenaline-fueled periods for retailers where you know you try and get Christmas absolutely right and we had someone from with huge operations experience in retail to kick off the meeting. And then we had inputs from lots of other retailers in the meeting as well as a few CPGs. It was a very very good meeting. I learned a lot and took lots of notes, but what about you as the academic? You know, I think you've got an academic notebook somewhere. What were your key takeaways Daniel? Daniel: Well, my academic notebook is starting to be so full that it overflows. No, jokes aside, it was interesting as you correctly pointed out that we had a wonderful expert here, a veteran retail expert and what was interesting to him, you know with all 38 years of working on retail at Christmas season. Pretty much everything boils down to discipline and execution. So he said first of all stop all complexity. No more shippers. No more nothing, right? And I thought that leads me to the first point. Apart from the fact that everything is about discipline and execution - so that wouldn't have been something special to Christmas. What was interesting is that we actually talked very little about, you know, the tech stuff and while there is lots of tech that is being given away as presents around Christmas. It turns out that technology doesn't seem to be helping particularly in the tech Christmas, you know tech doesn't really help with the Christmas season. As a matter of fact a lot of what we discussed was around (and here I have to give all my all my regards to this veteran retail expert), it's all about you know discipline and planning. So I think you and I agree. What was interesting, was this thing that in the start of the year around Boxing Day you know they do a first you know debrief to figure out what went well, and what what has to be improved. And then later the year they start, after having done the first real orders of the Christmas presents whatever the stuff that comes from far away. What was interesting is that towards Christmas something changes and what changes is the sort of the planning and the communication schedule. All of a sudden it starts changing from weekly to daily. So one person mentioned they have something like an air traffic control call every day. A daily call across the whole company. Everybody should join. It starts December 6th where 6-8 store managers dial in and every one of them gives the three top issues on that day. So that's the daily call and they also discuss what their rivals are doing. And that wasn't just one person. There was another UK retailer they also said they forecast the sales per day to manage labour so it's a daily view. And there was another UK retailer who says they actually have availability targets per day and I thought that was kind of interesting that all of a sudden you see the intensity of the coordination completely accelerate and I thought that was probably right. Colin: Yeah there was a lot of intensity as you know I think we heard you know it's a really exciting period if you're a retailer and trying to make Christmas really really successful. What were some of the lessons do you think first, for maybe vendors who are listening and we had a few of them on the call? But if you were a vendor what would be some of the takeaways you think you might have got from the meeting? Daniel: One of the memorable things that I want to mention here is that they clearly said suppliers, vendors, please make displays that survive the Christmas transportation chaos. Because they said there's so many times that they actually have the displays arrive and they haven't really survived the whole transportation, whatever it is. So they arrive and the retailers look at them say no, we can't put them up there. Make the displays Christmas-proof, in essence, was some of the things that they mentioned. And then the usual stuff, you know, plan tightly, coordinate, all of that, because you're not the only vendor that wants to place your display. You're not the only vendor that wants to make a fantastic and amazing unheard of promotion. And all of that has to happen in this very short time frame. Which brings me to an interesting point that I think we both find very interesting. There's some retailers give their store managers a lot of discretion and autonomy to change and localise some of the assortment. Totally fascinating. One guy said there's 25% extra stock that store managers could order. Not extra stock, but 25% extra that they could actually do with what they want to. They could order in place on their own. They could change the merchandising according to the local know-how, entrepreneurial, strategically place the chippers. And obviously, you know, there you get into an interesting situation. There's probably more experienced store managers who do it really well. And then there's the others who may not do it themselves. But I think what you and I realised is that this disrupts the planning. And not just in a positive way. Snce all of a sudden, you know, shelf capacity changes, plan diagrams are no longer here. You know, if I would be in headquarters, I would love to give my store managers the discretion, but on the other hand for 25 percent of the stuff, I don't have clarity anymore. Shouldn't that influence tech? Colin: I think it does influence tech in terms of the ability to use it. We didn't hear of any special Christmas AI tech, did we? Coming out there. So I don't think that was a big thing. I really caught that there was clearly a process that starts in January every year they look to improve and get a better Christmas or better holiday season sales operations-wise and execution-wise. It was really fascinating. Daniel: Yeah, there's one more thing that I was talking about now specifically to on-shelf availability. I remember now one person mentioned that if you realise in Christmas period that you don't have enough of something, it's too late anyway. Unless of course it's butter. But if you really miss certain key items, there's no way you can get any additional, right? Which is actually from an academic point of view an interesting difference because there's product that exists all through the year where you have a certain way of inventory management and then you have these seasonal products where you have very often the planning, it's like a one order and maybe you can reorder once, but you cannot reorder as you wish. And so therefore you have to plan much more ahead and you have much more of an overage underage problem. It means maybe you have too much or very often too little, but you can't change it. That's a very interesting way of changing the planning and inventory management routine towards Christmas. Colin: Yes, and so we did touch on, didn't we get business as usual, getting back to business as usual, after Boxing Day or sort of all like January the 6th or January the 12th and just the need to do markdowns and discount those products you've got too much of. Daniel: Wait a second, it's not just people here from the UK and the United States on this show here. When exactly is Boxing Day if you give it a calendar day. What is that? Is that January? Colin:No, Boxing Day is the 26th of December but as you say in Russia it's probably not. In other countries it might not be the same. Daniel: Okay, just so we understand like for instance in Germany Boxing Day doesn't exist, I just want to make sure that everybody outside of the UK and the world understands what we're talking about. Otherwise they think this is the day when the Brits start going to the pubs and afterwards they start Boxing. Colin: Well yeah, that happens in provincial towns all over the place every Saturday night. So what's our next one? We've got January the 17th, I think we're next up. What have we got on January the 17th? Daniel: Well exciting, we're going to try to enlist the suppliers to help with on-shelf availability by sharing what they believe are ways they could support retailers on-shelf availability targets and a couple of things that we want to discuss First of all Christmas proof displays, data sharing as an option. A very interesting one is how suppliers use field sales force in order to support it and to what degree there is a room for optimisation if they work together better. What else is there? You had one or two more? Colin: Clearly there's metrics that the suppliers are working on. Forecasting and replenishment would be another one that I think they've been working on, especially around ranging per store. So yeah, it'd be very interesting to see what each of them are actually up to. Whether they're using technology as well, their perspective on some of the technologies that retailers are looking to introduce as well. Daniel: Well, in between this call and Chris and the seventeenth, we still have Christmas. And now the question is, did you get all your Christmas presents? Colin: I hope so. I hope so. I hope you do as well. Daniel, as ever, a pleasure. Thank you for your ongoing support and we'll see you on the seventeenth. All the best. Daniel: Thank you. Bye bye.
Background and Remit of the Group ECR Retail Loss is part of the ECR Community, a voluntary and collaborative retailer-manufacturer platform with a mission to ‘fulfil consumer wishes better, faster and at less cost’. Over the last 23 years, the group has acted as an independent think tank focused on creating imaginative new ways to better manage the problems of loss and on-shelf availability across the retail industry. Championing the idea of Sell More and Lose Less, the group is open to any retailer and manufacturer to join. In the past 23 years, the group has invested more than €5 million in research projects focussed upon meeting its core objectives. Details of previous projects can be found at: www.ecrloss.com. The group is interested in a broad range of areas relating to retail ‘loss’ and its control, adopting an inclusive definition to encompass any aspect of retailing that negatively impacts upon retail profitability and the methods that can be deployed to minimise their effect. For instance, the issues of food waste, on-shelf availability, inventory accuracy and ecommerce losses are as relevant to the group as losses due to staff and customer theft. In addition, the group is interested in exploring a wide range of approaches to address these issues, including those focussed upon, guardianship, technologies, retail processes, and design. All types of retailing and modes of operation are of interest – from grocery to apparel and physical stores to online/omni channel. In addition, the group is keen to explore problems of loss across the entire retail supply chain – including shipping, distribution and reverse logistics. Of particular importance to the work of the group is generating new insights and approaches that offer practical value to the retail industry to enable them to sell more and lose less. As such, its research agenda is very much driven by a desire to generate high quality, independent, rigorous and methodologically robust studies that the retail community can have high levels of confidence in. Research Funding Opportunities The group is interested in receiving innovative and exciting research proposals from academics working in any higher education institution in the world (students are excluded). To guide your application process, the value of an ECR research grant can be up to €100,000. The group has a simple two-step application process: 1) Initial expression of interest 2) More detailed research proposal Initial Expression of Interest In order to keep the application process simple and to avoid unnecessary burden, the group is keen in the first instance to receive a short outline (no more than two pages) outlining the proposed research, covering the following issues: · Overall aims and objectives of the research · Relevance to the work of ECR · Proposed research methods and timing · Likely outputs and deliverables · Your experience and expertise · Approximate total cost of the research There is no deadline for submission – applications can be made at any time. Academics interested in applying can contact ECR Retail Loss to discuss their ideas before submitting an application. The group will review all initial expressions of interest within 2-3 months of submission. Detailed Research Proposal Successful applications will then be invited to complete a more detailed research proposal (5-6 pages) which covers the following issues: · Project summary · Case for importance of the project · Aims and objectives · Methods, approach and activities · Outcomes, outputs and dissemination · Staffing · Legal and Ethical Aspects · Timetable · Budget (e.g. Salary, University Overheads, Equipment, Conference Attendance, Travel and Subsistence etc). Each detailed research proposal will be assessed by a panel made up of the ECR Retail Loss Board, invited retailer representatives and the ECR Academic Advisor. Applicants will be notified within 3 months of submission. To find out more, contact Colin Peacock: email@example.com
The research priorities are determined by its members – they drive the agenda to ensure ECR delivers research that meets the need of the industry bringing new insights, tools and techniques that enables retailers to sell more and lose less.
ECR Community a.s.b.l