February 25, 2018
Evo Chairman Robert Diamond recently took part in the ESCP Europe 2018 panel discussion on Industry 4.0. Among the questions were, “Will Amazon kill the fashion industry?”
What does Industry 4.0 mean for the fashion industry?
Let me talk to the space that Evo spends most of its time in. Which is about using tech to deliver continuous improvement in everyday decisions.
So, the basic principle is that people are not best at making a large volume of repeatable decisions with lots of different data inputs. So, it’s about identifying where in a value chain people add value which is I think generally in innovation, and in selling, and where machines and algorithms add value, which is largely in what I would call continuous improvement.
Small gains of everyday decisions, So, that for me is … the key part of 4.0 is we’ve now got a massive increase in the number of data points being generated and then the ability to use machine learning to turn those around and do things through different touchpoints on a near real-time basis.
What are the 4.0 technologies which have the biggest potential in the fashion industry?
This is a really critical question because my take on the history of the fashion sector is that, compared to other sectors it’s about a lot more uncertainty within the business model.
I was thinking about the fact that more artificial intelligence goes into making decisions about this bottle of water than there does a $2,000 leather jacket.
I sense there are three reasons for that. One is that the perception that the touchpoint between the product and the consumer in a traditional retail store is a moment of magic. It’s a moment of high art and not a moment of science. And I think from a journey perspective there’s an increasing realization that actually it could be a moment of science.
I think the second point is that because so much – especially in apparel – so much of the offer changes season on season. Again, there’s an assumption that it’s art not science.
And the third thing I think, just rather simplistically, is in the multi-channel world so much stuff gets returned. Nobody really knows what’s going on. And, actually understanding the feedback where [I don’t know what it is for Allsaints] typically 30 35% of the merchandise that is sold online is then returned either online or through a store again I think has been used as something of an excuse for kind of not understanding customer needs in a more scientific and precise way. It’s kind of my take on it.
In fashion, where are companies more ready to apply these analytics and all that on the chain? Is it more on the customer relationship on the retail part today or can you observe other areas?
My take is that, historically, data has really focused more on the supply chain I think than the demand chain. So people are very familiar with using data to support demand forecasting, so right the way through the production process: how much do we need, how many pieces in what combinations?
And to a degree, they’re taking the same logic into replenishment as well. And that is either in season replenishment or store to store transhipment.
Personally, I think where the next generation of conversation will be is in the demand chain, with decisions especially around pricing, seasonal markdown, and other forms of promotion where I think that data has been used in a very, very loose way by many retailers.
And essentially, if you buy the thesis that you sell 30% of your collection at full retail margin and that makes all your profit, the bulk of it sits around and you try to break even through the season and then you lose a chunk of profitability at the end of the season when you have to go into markdown. I think the logic around that is shifting to say, “How can we use insight into buying patterns of different types of merchandise in different markets, and even down to different stores, to sharpen price so that we clear as much product in season at full retail?”
It can be a reduced full retail price, but at full retail, and I think that’s the new battleground. Which is using it to support what, traditionally, is very Excel-based trading and merchandizing decisions where somebody sits there with a spreadsheet on a Monday morning and literally marks through, “Oh that’s coming down this week.”
That’s a pretty blunt instrument and I think that, increasingly, we’re going to find ways of supporting – I don’t think we’ll replace those decisions – but we’ll find ways of supporting those decisions. So that people are just making more informed choices and leaving less value on the cutting room floor.
What do you think is the level of maturity and acceptance of fashion of the different tools of Industry 4.0 and digital transformation? Where are they compared to other industries where you can work with and what do you think are the key potential and where should they invest in?
I definitely agree that compared to other sectors I’ve experienced I think fashion is in some ways a long way behind. It’s pretty simple.
Again, it’s all about the model. Which is, if the outcome of the decision tomorrow is pretty much the same as what happened yesterday, then there’s a chance for machine learning to help improve. Frankly, within fashion there are decisions – I don’t know about the people in the audience – I don’t think demand forecasting is a particularly exciting job, and replenishment even less so. The Monday morning markdown is probably at the bottom of the pile so I kind of welcome the day when a machine can take over.
That said, the thing that I think where fashion’s truly led the way through technology is in building the brand and building the brand experience. And I think, just as we are moving into a great commoditization of the retail offer, and retail fulfilment, I think one thing that lots of other sectors are learning from brands is how to create an online and physical retail experience which integrates technology and brings brands to life in a way that other retailers are really struggling with. I think if we’re seeing an influx of brains from outside the traditional fashion sector coming in, I think one thing we’re going to see going back is this ability to build brilliant brands and brilliant brand experiences at point of sale.
At the point of sale, and all the customer experience, where do you think is the biggest potential, what will change in terms of the kind of experience, what do you expect?
For me, it’s mainly around the interactivity, which is the ability to deliver dynamic, increasingly personalized messages, both online and instore, and then for a brand to learn from those interactions so, while I think traditional loyalty programmes and things are really in decline – they are on the way out, I think that technology will replace them.
It’s a far more replacement as a platform for learning about what people like and don’t like. And I think we’re going to see a huge injection of smart technology, obviously online but also in physical retail, help build that connection between people and brands.
How can fashion brands compete with Amazon? Will it kill the fashion business?
The fact is that Amazon is just a very smart reseller.
They aggregate demand, so my interest is largely in their ability to aggregate demand, aggregate a curated range of suppliers with, in theory, lowest possible market price, and bring that together with massive footfall and the ability to dynamically cross-sell.
Will they get more and more into the vertically integrated chain?
Well, if there’s money in it they’ll do it. They already do it. It’s analogous to grocery retailers offering own label product. If fashion originators can’t come up with any particularly new ideas then they deserve to be taken out. They deserve to be commoditized.
So, in basics, I can absolutely imagine Amazon going there, but it’s no different to how Target in the US or Tesco in the UK, or Sainsbury’s have brought in their own relatively undifferentiated, low price, everyday clothing consumables. It’s no different to what they are doing in the grocery part of the store and I think that Amazon will continue down that path.
Do I think that they’ll go out and buy luxury brands or try to innovate or even mainstream brands?
Not for a while.
Just to close off on the Amazon discussion. I think the huge difference between how an Amazonian thinks and how a fashion, especially retailer, might think is: in Amazon there is only one answer which is more technology, more data. There is no resistance. And I think the question for fashion operators is: is the answer for every question, more tech?
And people like me believe it’s the answer to almost every question. Certainly any question where the question tomorrow is the same as the question yesterday the answer is technology will almost certainly do it better.
The interesting question is that does that then free up humans to go and do new stuff? So what are they going to go and do? That’s what fashion does brilliantly – is freeing up people, now how many people – it may need fewer people to do the new stuff than are doing the everyday stuff today, but technology will win because it’s faster, it’s cheaper, and it makes fewer mistakes.
How close are we to using AI to personalize the shopping experience?
I think if people are worried about sharing their information and worried about what people are doing with it, I’d get over it. It’s done. It’s happening. It’s everywhere.
The difference is, we allow it with technology. We allow intense personalization of every decision we make online. But the moment it moves into the physical world we start to worry that it’s kind of stepping over the line.
Lots of the information you’re talking about is available already. It’s available through credit card companies – so Visa and Mastercard sell aggregated information about who buys what. So if you are a multibrand retailer you can go to them – for a price – they will tell you whether an Allsaints customer also shops in these other stores in the local area.
The same for the telcos – they can do that. And then the over-the-top brands of Google and Facebook, and some other operators can do that too. So that information is really, really useful.
I think the traditional stuff I touched on – loyalty cards – it’s hopeless. These have all gone. But I think that smart data will identify people very early on. It will focus them towards decisions they typically make – and that could be a size in a store, or it could be a colour, or other types of merchandise and it will just make their visit more relevant so they don’t have to walk through racks and racks of merchandise that just isn’t for them and will never be for them.
I think that absolutely is here now and it’s just about the cost efficiency of rolling that out to the mainstream.
You said previously that Amazon was all about ‘more tech’, so how do interpret the fact that they recently opened a bricks and mortar store and what do you believe it shows?
I’d be careful about opening a single store because to have a meaningful retail footprint in America you’ve between two and three and a half thousand stores – to have a national footprint – so it’s very different.
I come back to: if the transaction is commoditized, I’m just buying more stuff and it’s the same stuff I bought last week and it’ll be the same stuff I buy next week, which lends itself largely to grocery, then absolutely I think the way forward is to more automation.
And you can look at Target in the US which recently launched a concept where they took a store – and remember Target has big stores – this is 115,000 square feet, which is Target’s average – it had 2 doors, and the door on the left took you into a full service, mixed merchandise browsing environment and the door on the right took you into a limited range convenience store.
So basically, one of them is “I’m going to think” and the other one is “I’m just going to buy more of what I’ve already bought.” I absolutely see that as the way the world is going. That some decisions will be replenishment, and some will be choice.
I think that the biggest practical difference that I see in the fashion sector is, where there is choice, I just see there being less physical product in store. I don’t think a retailer needs to have every item in every size in every colour combination in every store. And it doesn’t need to say “OK all stores of this size have the same merchandise offer. I think what you’ll see is people say well I really like that and I’ll have it in blue and somebody says “I’ll get it to you within the hour”.
And I think that will just massively reduce the amount of inventory sitting in stores, which I think is great because that inventory is doing nothing. I don’t think it’s making people buy things because, once they’ve decided to buy, it’s getting the right item, size, colour combination in their hands. And that can be increasingly done outside of the physical store environment.
To finish off the discussion, what, in your opinion, is the biggest challenge for the implementation of Industry 4.0 in the fashion industry?
For me it’s all about the human-machine alliance, which is let machines do what machines do best and let humans do what humans do best, which is engaging with people and coming up with ideas where tomorrow’s solution is different to yesterday’s.
About the author
Robert Diamond is Evo Pricing’s Chairman, and a successful entrepreneur with a track record of creating value from customer data.
After 13 years in the retail data space, in 2001 Robert founded what became emnos – the leading provider of ‘customer centric’ services and decision-support tools. Clients included top retailers (Target, Walgreens, Morrisons, Waitrose, Boots, Co-operative Group) and their FMCG suppliers. The business was sold to Palamon Private Equity in 2007 and then sold on to American Express in 2011. Robert moved on to an executive role at private equity-owned RAC Motoring Services, as well as overseeing a portfolio of private investments.
Outside of this Robert manages a private portfolio covering early-stage retail technology start-ups; lifestyle & hospitality investments; commercial & residential property; and classic cars. Robert is an elected member of the Young Presidents Organisation and regular media contributor (and TED talker) on the use of customer data. He lives in the UK and US with his wife and four children.