Earlier this year, Uwe Hennig, CEO of Detego GmbH, spoke to us about his analytical technology Suite, the trends sweeping our industry, and the need for open platforms.
Name: Uwe Hennig
Occupation: CEO, Detego GmbH
Likes: Hiking, race-bicycle riding, cooking, fast cars
Dislikes: Disrespect and dishonest people
Words to live by: Live and let live
Lydia Hanson: Uwe, can you start by giving us a brief overview of Detego, in your own words? Your software solutions provide real-time analytics and merchandise visibility for retailers?
Uwe Hennig: We are a software company, providing a platform with a lot of out of the box services. It’s a SaaS-based system running in the cloud, which comes with a lot of standardized functionality. We are focusing, in retail, on the fashion section, which to us means clothing, footwear, accessories, and even cosmetics. Because we recruit from the fashion segment we have a lot of experience there. This doesn’t mean that, in the future, we won’t expand to any other retail segment, but this is our focus at the moment.
With our solution, the retailer is getting what we call a real-time single stock view and, in real-time, in-store consumer behaviour. Let’s be honest – when you go into a fashion retail store, you like the websites of these companies, but often coming into a physical store you feel like you’ve stepped back into the ‘70s: there’s no information available if you ask an employee for a skirt one size smaller, for example, or they disappear into a back room and come back just to say “I don’t know”. So everything that spoils consumers – and we’re all shoppers – on the Internet, is not really present in a store.
What we do as a platform is first to analyse the availability of merchandise across all locations and channels, and analyse how the consumer is connecting to the merchandise, all in real-time. These retailers have IT systems in place, absolutely, and even tablets at POS in the store, but all of these systems are made for reporting purposes – weekly closing, monthly reports from ERP etc. – and don’t put the consumer in the middle. They are not customer-centric. Because this data is missing, retailers are often making decisions based on gut feelings in-store; we think they also need to digitalize the stores and make decisions based on database information, rather than the gut feeling of a store manager. And that’s what we’re providing: real-time insights into consumer behavior, availability of merchandise, and how the consumer is connecting to the merchandise / the brand.
There are different levels of applications, or different levels of analytics, for different users; the store associate needs different information than the store manager, than the merchandise planner or marketing department in the headquarters. The basis really is having the data – using IoT technologies like cameras and RFID – and analyzing it and providing insights in real-time, to make a decision now and not read a report six weeks later when it’s too late.
Lydia Hanson: Analyzing customer behavior in-store makes perfect sense with the direction the fashion industry is going, with everything moving towards digitalization and the customer experience. How many aspects of in-store behavior do you – or can you – monitor? Is it only what customers are buying, or is it what customers are taking into the fitting room and then subsequently not buying, as well as what merchandise is in store and is in stock?
Uwe Hennig: Save actually counting them …I think what’s important for fashion is, even with a lot of eCommerce surfing, in fashion retail 80% of the revenue is still delivered through a retail store. And that’s really important. The real purchase is happening in the store because you want to touch the garment, try it on, and see the colour in natural light. So 80% of sales are made ‘offline’. I actually think it’s quite amazing that retailers dub their store business as their ‘offline business’ in an online world, where the consumer is always online.
The biggest problem for the fashion retailer is that they don’t know where there stuff is, so they can’t commit to a client. If you were to go to a store and request a certain shirt in a size XS, the shirt exists in XS but the associate doesn’t know where it is. So they’re disappointing you.
In our InStore analytics platform the first thing we measure is the on shelf availability of colour and sizes. So, if you run a marketing campaign for, say, Valentine’s Day with only pink shirts (or whatever) – logistics is doing a great job shipping to hundreds of stores – we can measure if the items on the shelf are enough in quantity and in size. Every time a size moves to a fitting room and is purchased we ensure that, because there are enough in the back room, replenishment from the back to the front is happening in a minute, ready for the next customer. We are doing this with our customers to a level of 98%, ensuring items are really in front of the customer.
Then if we move onto something you said, which is very important, about the fitting room. When clothing goes to the fitting room, the deal is almost done; the customer is investing time, taking an item to the fitting room to try it on, and you do not want to lose this customer. You want to give him or her the best service. So when he or she is asking for help (perhaps wanting a different size), by knowing what the consumer is doing in the fitting room, we can provide many more behavioral aspects. For example, what is he bringing to the fitting room and not purchasing (as you said)? Let’s say everyone likes a particular red shirt, and is trying it on in all sizes, but the size M is not arriving at the point of sale; all other sizes that went into the fitting room are purchased but size M is not. Perhaps the style is wrong? Or something else.
What we are analyzing in a large German sports brand where we have implemented are the top combinations people bring into the fitting room. These trousers, for example, always going in together with this shirt. This is what we are feeding back to a merchandise planner, and then maybe the brand’s marketing department can make a bundle out of these products to sell on the website together.
In smart mirrors, we have a recommendation engine. So, when you bring in a red shirt, we are recommending a certain trouser, based on analytics from all other fitting rooms of that brand. Then we can analyse if the customer purchased these recommendations. The system is learning, so if we have been recommending white trousers to go with the red shirt but nobody is purchasing these, then we stop this recommendation, and alert the marketing planning team that it isn’t working.
We analyse which items are requested once someone is already in the fitting room (but then not purchased), as well as those that are never requested, and are just sitting in the back room for weeks, getting older. This is a big risk, as they’ll later need to be marked down, meaning the business is losing margins. We are analyzing ageing structures of items – in the back room, on shelves, in the front room – and analyzing, when a new collection launches, the most used items (in the first week) in which stores, in which colours and sizes.
We’re really analyzing all of the insights fashion brands get from their eCommerce sites …but there it is easy to track. They track you when you move your mouse with cookies. We bring the same into a physical store.
Lydia Hanson: [Laughing] I suppose for my question of ‘how many aspects do you monitor?’ the answer is ‘a lot’.
Uwe Hennig: [Also laughing] Exactly. Each business is running on KPIs. The in store KPIs for a retailer are always the same – revenue per day, per store, per week – but we are bringing new KPIs, of how a consumer is engaging with a product: top ten items in fitting rooms, fitting room conversion rates. If you then put a benchmark you can see how many stores are falling below that, and visit those stores to find out, for example, that the store associates are using the fitting rooms as a storage location. These things are happening. At the end of the day, we want to provide a base of data that prepares a digital decision, rather than a decision based on, “I think the weather is good, so we might sell more t shirts”.
Lydia Hanson: Precisely. I just wanted to pick up on the recommendations you mentioned. Linked to this, you work with individual retailers and offer insights to them based on their own lines and their own collections. Do you also work on a more general basis? What I mean by that is that some analytical, insight-driven software platforms offer ‘bulk’ data sets…
Uwe Hennig: Yes, I see where you’re going. The idea would be that if we had 50 sporting brands connected to our platform, you could theoretically build a kind of index for this segment of fashion. I think there would be big value there.
The value sits, at the moment, in the real-time in store analytics for the brand. But if more brands are connected then, sure, you could build such an index. We have actually spoken about this internally with our data scientists and for them it would be easy – they find patterns in a lot of data. But there is nothing in place in the product today. Perhaps that comes later in the industry; it would be a big help for the industry.
Lydia Hanson: The fashion industry is famously private, so I guess you would likely come across issues with regards to brands sharing data between them.
Uwe Hennig: Yes, it may take a little more time. But if you go back some time, people were very shy to provide any private information to consumers, and then look at what is happening today with information on Facebook and Instagram –the discussion around privacy is almost completely off the table.
Although, we are prepared for that, and we always try to explain that we are not ‘tracking’ consumers. I think that’s a very difficult word. We try to understand what a consumer is doing in store, to provide the consumer with a better service, or offering. Using these insights, a retailer is able to better react. At the end of the day, it’s benefitting the consumer. It’s not to track people; nobody tracks people to track people.
Lydia Hanson: And at the end of the day, as you’ve said, everybody is working for the consumer. There’s that well-known phrase of ‘the customer is always right’.
Uwe Hennig: Indeed.
Lydia Hanson: You’ve mentioned the InStore part of Detego a few times, but I understand that the full Detego Suite contains a few different elements – InStore, InChannels, InWarehouse, and InReports. You’ve shared a lot already, but could you just tell me a little bit more about the actual technology behind this suite? Just try not to give away your secrets.
Uwe Hennig: [Laughing] Sure. The Detego Suite has these four products. InStore and InWarehouse stand for themselves. The idea was that if a fashion retailer came to us wanting to get more insight into their business, but not wanting to install a huge ERP system, for example, they could – so we split it into different products. So they could use InWarehouse to analyse their performance supplying to a wholesale business or a franchise business. Then later on they could say, “I want to go my first store” and use the Detego InStore product.
InChannels, in fact, is the product where you are combining the consumer data with the merchandise data. So InStore and InWarehouse are more B2B, whereas InChannels is more B2C.
The InReport sits on top of this; it’s the whole analytics. We can analyse a lot of what is happening in the warehouse, for example. You have people working in the warehouse on the inbound and the outbound sides. Sometimes it happens where you have a lot of trucks coming in, and you can’t offload everything as you don’t have enough people, but then the next day you have lots of people sitting around the warehouse, with nothing to offload and therefore nothing to do. So, we are analyzing how many people are there, and we know the picking performance per person. So, if we know there is a big order coming in we can recommend that tomorrow afternoon you move some people around. When you analyse in real-time you have a better chance to steer it.
The technology behind is really sensor IoT technology. The RFID tracks garments, and all these fashion items are tagged. There is a huge tagging initiative in the industry, with many fashion brands doing source tagging in the factories. It’s like a unique identifier – like a license plate on a car – that can be used just once.
We also use cameras in store. Then, if you have big department stores, you might see that the people flow in the store is not aligned to your campaign products. You can see heat maps, or do people counting. In the fitting rooms you wouldn’t want to put cameras in clearly [laughing], but you can have sensors to know when someone is in there, and then RFID sensors in the fitting room to sense what he or she has brought in. There are a lot of sensor technologies.
To us it’s an open platform so we have people asking, for example, if we can connect weather data. This could let them know whether, when a product purchase was increasing, it was raining or sunny outside. It’s just another data feed we could use to make the analytics even more detailed.
Lydia Hanson: You’ve jumped onto my next question without me! I was going to ask about whether you were an open platform, able to ‘plug in’ to other systems?
Uwe Hennig: Absolutely. We have to be open; our view is that the time for huge systems, with two-year rollouts, is over. The software needs to be as flexible as the consumer is. You need to be able to switch on a service in the Internet, use it in five stores for a week, say, and then for whatever reason you decide you don’t want it anymore, and want to combine the Detego product with another third party application because it brings huge value for your business and consumer.
So we built the system completely open. Everything inside the products is using micro services, so we can even isolate some very small services, and provide these to another software in the IT landscape of the retailer.
In the future, we have to all be open platforms. This is the demand from the customer, and if you aren’t following it you will lose out as a software company.
Lydia Hanson: I couldn’t agree more. We do come across closed platforms in this industry, some of which are opening themselves up to difficulty down the line.
With the Detego Suite, do most of your clients purchase the entire platform? Or are they able to purchase individual products within the Suite?
Uwe Hennig: Exactly that. For our InStore service, we even have a lean version. So if someone wants to start very small, with 5 stores, to check whether their inventory is OK, we can start there. So this person is purchasing a service on a monthly base, in 5 stores, and we would invoice this way. Then, it could be working great so he could add 10 more stores, or add InWarehouse, and a year later add another country with 50 stores. It’s a very flexible model.
Being in the software industry for so long, it’s obvious this big bang approach of developing a custom specific solution for a long time before you deploy it in 1,000 stores is over. Very flexible and agile methods are in use.
Lydia Hanson: And these different products within the Suite are of course connected but performing different tasks. And this is what the typical retailer or brand looks like today; they will have tens or hundreds of different systems within their business, each different, but all working to the same end. It’s important, certainly for larger businesses, to have ‘a bit of everything’ in a way.
Speaking of clients, where is your client base? Are you global, or predominantly European?
Uwe Hennig: It’s become global. We are from Europe, with headquarters in London. We started in Europe, and did our first projects here, but we are serving global brands so we work in Russia, in North America, and now even in places like Japan.
The beauty of software from the Internet means we don’t have to have physical offices all over the world – because our software’s in the cloud. We have a global network of partners, of system integrators, who then go to a physical store – in Tokyo or Brazil, or wherever you like.
Lydia Hanson: So since being founded around the turn of the millennium, you’ve had a great reception from the fashion and apparel space?
Uwe Hennig: Yes. I mean, RFID in fashion is exploding at the moment, but when we started there was just a little bit here and a little bit here. There were some concerns on costing, on whether it was really working. RFID is nothing new – in fact it’s a very old technology – but the performance now is in such a great place. Technology providers have made huge strides in recent years and, because the adoption rate increased so heavily, the price for this tech went down dramatically. This, in turn, gave even more of a push to adopt it. There is a huge RFID initiative in fashion retailers across the globe – if you pick the top 100 in Europe or in the US, everyone has an RFID initiative, or is already deploying one, or already live.
Lydia Hanson: It’s the same with the majority of IoT technologies. Again, the actual term has been around for a long time, but it’s only just starting to pick up traction. Retailers and brands are now beginning to really use it and to be less afraid of it (in a way) than they perhaps once were – linked to things like the privacy issues we were discussing earlier.
Uwe Hennig: I think it’s down to a combination of things: the technology is now working; the privacy issue is disappearing from the consumer point of view (although not 100% down); the consumer is used to using his smartphone all the time, and from the mobile carriers/networks the prices for doing so are dramatically down; cloud solutions are getting completely stable and very fast. All of these factors allow us to get data from anywhere, anytime, and analyse it to provide even better service.
I mean we are all using our smartphones all of the time, right? And we’re using a lot of IoT technology and real-time data coming from the Internet all the time.
Lydia Hanson: Exactly. Most people are constantly using a lot of what we would consider ‘IoT technology’ without even realizing.
Uwe Hennig: Precisely. I was chatting to a friend recently and explaining that, when I am going about my normal day – let’s take last Wednesday, for example – I am sitting at home doing some online banking via my laptop; I’m using Spotify to hear my favourite music in my car; I’m going into a retail store where they know – from my loyalty card and my eCommerce activity – who I am, and they, using RFID, can allocate my stuff and send a signal to DHL to pick it up and bring it to my office the next day. All of this stuff is happening every day. I’m ordering food online, and my wife is picking it up from DHL. This is all cloud, SaaS, Internet, IoT.
Lydia Hanson: We certainly live in a smart world.
Uwe Hennig: [Laughing] I like that.
Lydia Hanson: We’ve spoken about the vast amounts of data that can be available to companies from businesses like yours. We’ve been researching what companies can actually do with all of this accrued data, and have often found a bit of a disconnect there. Is there a particular use case or client that you can tell us about, in relation to their use of Detego, and the benefits they’ve seen? Or an example you can give, without ‘naming names’.
Uwe Hennig: Unfortunately, as you’ve guessed, I can’t disclose the name, but we are working with a large, global sporting brand. They are analyzing in all (hundreds) of their stores, with our product, the availability of the merchandise on the shelf. They’ve put KPIs against it – I think 98% availability – to ensure that when the consumers come into the store they can make that purchase.
What they are also doing is the combination of the eCommerce with the offline business – so, when you are surfing on their website and like a shirt, for example, you want to know if it’s available in Munich tomorrow. So the eCommerce system is querying our system because we are sitting on the real-time stock – we can confirm it is in Munich, now, at the same second, and the consumer is then pressing a button to reserve it for 24 hours, and our system is alerting (in the same second) a store associate to take the shirt to the reservation zone for this customer coming to try it on tomorrow.
The real-time element is so important. These processes exist in retailers, but they are offline. How good is the information you check in the morning, in the afternoon? The item could have been sold by then. Or someone who reserved it could have cancelled that reservation.
Real-time is the main driver, and this sporting brand is doing this across all of their stores.
We are tracking around 80 million fashion items per year, and connecting the eCommerce with the offline business in close to 1,000 stores.
Lydia Hanson: It’s great to hear concrete, real world benefits.
Are you able to give us a glimpse into any upcoming developments within the Detego Suite? Or any news in the pipeline for the coming months you can share?
Uwe Hennig: What we have done already (though it’s in the early stages) is to develop a chatbot. What we figured out is that when people are in the fitting room communicating with the smart mirrors – “I like this, but I want it one size smaller” – they do not want to share their private information; they don’t want to say, “I’m Lydia” into this mirror. But if they would use their smartphone then they can do.
So what we’ve developed is: you have brought some items into the fitting room, and we want to ask you whether you like the style, does the size fit etc. People do this on their phones and you move to the chatbot, so you’re scanning with your phone a QR code from the mirror, logging in (using your Facebook or Instagram login), and answering these questions. The machine can ask you, “Oh you didn’t like this skirt. Is this down to style, colour, size, price?” This is where we try to identify patterns. If you tell us it’s down to the price, £50 for example, then we can ask whether you would consider purchasing it at £40. This kind of thing gives a lot of insights to the retailer. If everybody says they would take it for 10% off, then the retailer could mark it down just 10% instead of losing the customer.
Also, if you didn’t purchase something, and you are sitting at home a week later, the retailer can prompt you an email knowing who you are from your chatbot use to say “we saw you didn’t buy this skirt; we are running a campaign at the moment where if you want to pick it up tomorrow we will give you a voucher” etc.
This is what we have in beta stage at the moment [Editor’s note: this conversation took place mid-September 2017], but we are working heavily on, together with a big fashion brand from the US. They are helping by providing us with the right questions, so it’s a combined endeavor. This is something we’re hoping to bring to market quite soon. Again, it’s service for the consumer, while providing insights for the retailer.
The other things we are seeing coming are self-checkouts for fashion retailers. Queuing at the checkout on a Saturday afternoon is a nightmare, and these points of sales (POS) are taking up a lot of in-store space. If retailers could provide a self-checkout in the fitting room, for example, together with mobile payments, they could change the complete store setup. Then the store staff can be working with the customers instead of staying behind the desk.
Lydia Hanson: Please do keep us updated on that.
I just want to finish with one final question, away from your specific business and sector …Aside from self-checkouts what would you like to see coming from other applications and software in the not to distant future, for fashion as a whole? What do you think is missing today?
Uwe Hennig: Something on the B2C side, with the consumer. We’re speaking with some people in our network. One example – you’re putting on a size S, which is usually your size, but in this brand a size S is always too small and you need to take a size M. That is very important information for the brand. There are companies now – one is called True Fit – that are analyzing just this feedback and giving it back to the brand.
There is another company – Wide Eye, from Spain – and I really like their technology. You’ll be surfing Instagram and see a celebrity wearing a yellow skirt and you really like it, so you copy the link of this picture and send it to their database, where they’re able to tell you that it’s from Ted Baker, for example. They have a huge database, but then we are coming into play, and our real-time data can tell that person that this yellow skirt is really available in their local Ted Baker store.
So, I see these other B2C fashion startups using our data to provide a better service from their system, and vice versa. And this circles back to my earlier point about having open platforms.
Lydia Hanson: It seems as though we’ve come full circle. So, with that, I’d like to thank you for giving up so much time to chat today.
*Keep up-to-date with Detego here.