Connecting to LinkedIn...

The big interview: Michael Ross


Michael Ross is Co-founder and Chief Scientist at OrderDynamics, a big data prescriptive analytics software company that believes that the order is the moment of truth: when profit can be made or lost. Retail Gazette caught up with him to talk about the retail landscape from his point of view.


Until recently, I associated the name Michael Ross with a leading character from Suits, an American legal drama that follows a highly intelligent and cool college dropout. The Michael Ross I meet is no college dropout, but he is certainly intelligent and he is definitely cool.


I meet Ross at the OrderDynamics head office, following him into a room that overlooks stores off Oxford Street. As I enter the room and glance down at Britain’s busiest shopping hub, I comment on the view, bringing up the ‘death of the high street’ and asking Ross if it’s a ‘thing’.


“It’s definitely a ‘thing’,” says Ross, “but the high street isn’t dying, it’s just changing.”


Ross probably has the credentials to write a bestselling autobiography. He has a maths degree from Cambridge and prior to setting up OrderDynamics, was CEO at lingerie e-tailer Figleaves. He is credited with launching Elle Macpherson’s lingerie line in the UK and once, organised a tennis match between Anna Kournikova and Amazon CEO Jeff Bezos.


I tell Ross that he’s fascinating and he agrees.


At 13, he played chess for England, a strategic game which can be an analogy for all sorts of things. Does he ever compare chess to retail?


“Oh yes,” he nods. “It’s a big theme for us, what we call ‘the automation of retail’. So this isn’t the idea of robots as shop assistants, but how a lot of decisions in retail that are currently made by humans, will end up being made by computers” he explains


Today’s data is complicated. In the physical world, humans are quite good at decision making, but in the digital world, masses of data make up decisions that computers are nimble with. During the Second World War, Alan Turing, a British pioneer in artificial intelligence who helped break German ciphers, envisaged a computer that would be able to play chess. It was his belief that if this could be made possible, it would demonstrate computers are incredibly smart. What followed was an enormous amount of effort, put into working out whether or not computers could be made to play the game. For around 30 years it was said that any child could beat a computer at chess, until PCs started to become better and better and began beating grandmasters.


In May 1997, an IBM supercomputer otherwise known as Deep Blue, beat then undefeated, chess world champion Garry Kasparov. Today, it’s fair to say that computers are better at chess than any human will ever be “and yet, there are retailers who are doubtful computers will ever have the judgement of a buyer” Ross says with surprise, “but these naysayers usually reach a point where they recognise computers’ ability to crunch data and the realisation kicks in that it far outstrips what humans can do.”


Ross also likens chess to retail when he talks about the limited number of moves a player can make in the game. “The order of moves is really important, doing one thing before the other makes all the difference in the world,” he tells me.


“I suppose one must always be a step ahead,” says this journalist, who doesn’t play chess. Ross nods.


All retailers buy inventory and depending on what they’re selling, they are not trying to have empty shelves by the end of next week. What they’re looking to achieve, is a rhythm of sales throughout the year. Any success comes off the back of tactical moves. “I see retailers spend millions and millions of pounds on marketing, sending traffic to products that are sold out, traffic to products that aren’t converting, or traffic to products that would happily sell out on their own without any additional push,” Ross comments. “Currently there’s a frenzy of re-targeting, so a consumer could visit a website only to have that product chase them around the web in an incredibly unintelligent and unsophisticated way. Advertising companies will make their money from it but it doesn’t reflect well on the retailers.”


The battleground of making decisions in a world of ultimate complexity and almost unlimited data is his life’s work, but how did Ross, once a manager at McKinsey, get into the business of underwear?


At the end of the ‘90s when the dotcom bubble was in full swing, Ross was advising on the internet, which he describes as similar to “selling shovels during the gold rush”. Google had just launched and there was a real buzz around e-tail. Ross was introduced to Easyshop, as it was called back then, a tiny business trading lingerie and health and beauty products. The latter categories turned out to be grey imports so once that was slashed all that was left was lingerie which Ross tells me was interesting but in the end, a very bad category to sell online.


“Really?” I ask with puzzlement, “but wouldn’t the success of the company say otherwise?”


There’s a pregnant pause before Ross responds, “the story of Figleaves has never really been told,” he says quietly, “but it wasn’t successful.”


He continues: “Figleaves was bigger than Asos and Net-A-Porter in the early 2000s but Asos is 40 times the size of Figleaves now. Its success is in the sense that it still exists and people shop from it, but in terms of what it could and should have been…”


There were many things that went wrong. In the first instance, there was no real understanding of the business’s growths dynamics. According to Ross, the key to success in any online business, or in the online division of any multichannel business, is how much you sell to your loyal customers.


“The difficulty with underwear is that our most loyal customers didn’t buy that frequently. The way these businesses grow is through customer acquisition, customer retention and by reducing churn rates and if the customers you retain don’t buy very frequently, essentially your growth naturally plateaus.”


Businesses that do well over a long period - the Amazons and the Net-A-Porters – they share the characteristics that their loyal customers are also very frequent customers. Another challenge for Figleaves revolved around the economics of selling third party brands versus own brands. The business model at Asos makes money selling Asos’ own brand but when stocking third party brands, it’s incredibly easy not to make money. “By the time markdowns, promotions, picking, packing and postage are taken into consideration, it’s very easy to give away all your margin,” the Cambridge graduate adds.


Elle Macpherson Intimates, the hugely successful partnership with lingerie giant Bendon, and Australian model Elle Macpherson, was launched in Europe by Figleaves. “She was very thoughtful in how to leverage her brand,” cites Ross, who describes Macpherson as “very tall” and “attractive.”


He comments: It was one of the good things and bad things about Figleaves, and we wanted to be the category killer, offering the best underwear from around the world.”


​When Ross was connected to Macpherson’s branding, he and his then team, pursued it. “It was exactly what we wanted to be doing,” he says. Elle Macpherson Intimates could be discovered in Selfridges and a couple of other retailers but for those who lived outside London, there was limited access to it and that was the Figleaves premise. “For many years it was our number one brand, it was exactly why Figleaves existed and why we set up the business,” continues Ross. “Where we went wrong was that we ended up with far too many brands. Elle Macpherson Intimates was absolutely the type of brand we should have been distributing but before long we ended up with 100 brands that we really shouldn’t have been selling.


“Without the constraints of a store’s four walls,” he adds, “we were lulled into a false sense of security where we assumed we could sell everything: every brand, in every size and colour. Photography costs are pretty low so why not?”


So essentially, Ross and his partner were on a journey of ‘let’s sell everything’ but the costs amount to thousands and thousands of SKUs, which is overwhelming for consumers, and several brands that don’t generate incremental sales and are cannibalised by the popular ranges.


“It seemed to make sense at the time, but ultimately it was a flawed strategy,” reflects Ross. “I talk to a lot of retailers now about what the optimal range size is. For Amazon it’s very simple: you sell everything in the world, but if you’re any other retailer, you’ve got limit the number of dresses, shoes, televisions and bags that you sell.”


Elle Macpherson Intimates proved fantastic for Figleaves and in hindsight, while it was precisely what Ross wanted for the online retailer, he just didn’t know when to stop.


I am told that once retailers find a formula that works, growth and profit growth, although an executional challenge, is conceptually very simple. “What I see is a lot of retail death spirals where overly aggressive plans are set,” Ross warns. “Commitment to inventories are made but retailers slip behind on their plans and in turn, pull one of two levers: either a promotional lever or a marketing lever. Online marketing is almost like a slot machine, it’s easy to pull the lever more than once and drive more traffic and spend more money. This way, retailers can deliver on revenue but then margin and profitability is killed and customers become trained to expect discounts.”


“There’s a lot of retailers who I say ‘You’ve kidded yourself’ to,” says Ross. “They may have generated online revenue over the last X years, but actually, the right strategy is to reduce online revenue and focus on profit. There’s a retail quote that ‘revenue is vanity, profit is sanity and cash is reality’ and in my mind, there’s still too much vanity, there’s not enough sanity and reality,” the data scientist comments.


When I ask what he would have done differently at Figleaves, Ross tells me he would have needed to do more than one thing differently. “The key is understanding growth trajectory and planning. If you plan well and truly understand your revenue potential, you make sensible decisions, but if you plan badly and are unrealistic about your opportunity, everything falls apart,” he says wisely.


“I talk to several CEOs and often ask what percentage of orders they ship on time,” Ross continues. More often than not I’ll hear answers such as ‘99.76%’ but once I ask what constitutes ‘shipping on time’, I realise that these retailers are marking their own homework,” Ross explains. “One told me if shipment is made within four days of its fraud check, that’s considered on time. This could mean an undesirable customer experience but the measurements read well, almost a corporate approach of kidding themselves.”


According to Ross, measuring the wrong thing can drive completely the wrong behaviours, to the convincement that things are okay. “At Figleaves, we would look at our product availability at SKU level – I thought we were 94% in stock, as our in stock SKUs were over our total number of SKUs. Then we hired Robin Terrell from Amazon who looked at our data and said, it didn’t matter that we were 94% in stock, were we in stock of the products customers are looking at? We needed to look at availability weighted by page view, i.e what was the customer’s experience of availability. When we did that, we were actually only 70% in stock, so we were systematically bouncing in and out of stock of our bestsellers,” he says. There’s only a very small percentage of retailers he ever meets who measure page weighted availability but it’s “fundamental to having any idea about what’s going on in your business” he tells me.


Seven years in, Ross decided it was time for a change. “Having now spent a lot of time with people who I would consider to be good retailers, I can’t say I’m a natural one” he says modestly, as he tells me he’s a huge fan of Nick Robertson (CEO at Asos), Robin Terrell (now Group Multichannel Director at Tesco), who Ross cites as “part of the answer,” and Chris North (Amazon UK Boss).


He recognised that what he enjoyed, and what he was good at, was making sense of the data. Ross set up OrderDynamics so that he could bring this sort of data driven approach to retailers, helping them avoid the mistakes he made at Figleaves.

Published on Thursday 26 February by Veebs Sabharwal

Articles similar to Michael Ross

Articles similar to OrderDynamics

comments powered by Disqus