Old school vs shiny new technology

I think people are too happy to dismiss “old school” in favor of new technologies. Old vs New should not be the singular criteria but rather, appropriateness given the situation. This being one of my pet theories, I was delighted to find (via Treehugger) that there is an active traditional knowledge inventory of older technologies being redeployed toward sustainable ends. At the forefront is Low Tech Magazine; consider this counter intuitive argument that bicycles are faster than cars (circa 1978):

The model American male devotes more than 1,600 hours a year to his car. He sits in it while it goes and while it stands idling. He parks it and searches for it. He earns the money to put down on it and to meet the monthly installments. He works to pay for gasoline, tolls, insurance, taxes, and tickets. He spends four of his sixteen waking hours on the road or gathering his resources for it. The model American puts in 1,600 hours to get 7,500 miles: less than five miles per hour… With his much lower salary, the Chinese acquires his durable bicycle in a fraction of the working hours an American devotes to the purchase of his obsolescent car. The cost of public utilities needed to facilitate bicycle traffic versus the price of an infrastructure tailored to high speeds is proportionately even less than the price differential of the vehicles used in the two systems.

fitting_robotCaveats abound, no need to argue the point but it is interesting to think about. Shortly after I read this, I found notice of a new exciting technology courtesy of BBC News. An Estonian firm has engineered a fitting robot with the purported intention of spurring online clothing sales.

“Only 7% of all clothing sales in the world happen online,” says Heikki Haldre, chief executive of Fits.me. “And the reason is that there’s no way to really try clothes on before you buy”.

Although it is irrefutable you can’t try on clothes virtually, it is not a given that a robot could provide this virtual function. I argue a better approach involves analyzing assumptions of the nature of sales and how to maximize them. Which is not to say the technology isn’t shiny (there’s a video, very very cool) but appropriateness is key. My approach would be to analyze who is buying clothes online, why they do it, which brands they’re buying and which they are not. For the record, it’s women aged 35-55 buying clothes online. It would be more useful (old school) to discover which vendors are better at capturing this market to reproduce its effects.

A fitting robot is not going to solve enough problems for enough people to be cost effective. First, the solution relies on a robot having volume and pattern measures of myriad styles across manufacturers embedded in its database. Most manufacturers don’t have that information internally so how could it be supplied to a third party (assuming they even would)? Why would manufacturers go through the considerable expense and time of gathering and supplying data when styles change every season?

You could argue a fitting robot would be great for a perennial style such as the Levi’s 501 but it is not so simple. Because each of Levi’s contractors is charged with making their own pattern, Levi’s would need data sets from multiple facilities they have in each country (Argentina plant 1, 2, and 3, lather, rinse and repeat for Brazil, Mexico, Turkey etc). The customer is only served if the jeans displayed, exactly matched the country of origin and supplier code from the care label to cause the given data set to be loaded into the robot’s working memory -and for said exact pair to be shipped to the customer. It is such a heady challenge as to be insurmountable.

An old school person would look for ways to increase Levi’s 501 sales by (arguably) more cost effective means such as standardization. If Levi’s used one pattern for all of their 501 production (destined for US consumers), their fit would be far more predictable. If their fit were more predictable across contractors, a customer wouldn’t need to try the jeans on. They could buy it online and be done with it. But no, if you want a well fitting jean, you are compelled to go to a store and try it on.

Technology is great to solve problems but it will never be a cost effective or appropriate application if its primary use is to compensate for inefficiencies or mistakes in our processes. If your refrigerator stinks, you can use some kind of high tech deodorant to mask the odor but the better solution is old school -clean the fridge. New technologies should be used to solve complex problems, not relatively simpler ones of our own making or within our control to mitigate. If your sizing is consistent and predictable, more people are going to buy it online even though buying online is a trade off. A consumer is trading the cost of their time in exchange for less uncertainty and a trouble free transaction.

Summary: the difference between old school vs shiny new technologies is that old school seeks to reclaim that which we once knew to solve problems that were never solved when newer technologies were lain over them. Both are useful and have their place provided the problem has been analyzed so you know you are trying to solve the appropriate one. It is less a matter of having the right answers than it is having the right questions.

Leave a Reply

20 Comments on "Old school vs shiny new technology"


[…] 60% of returns are made due to the incorrect size, 15% due to the fabric and 7% due to the colour no… […]

4 years 10 months ago

It’s been such an interesting discussion, it’s a shame recent big developments didn’t let me answer sooner.

We’ve launched the new version of the fitting room – which I invite everyone to critique. It’s accessible here: http://www.hawesandcurtis.com/mens-fashion-shirts-blue-plain-ref_YOPGE001-B40
(Red “New! Visual Size Guide” link next to size selection buttons).

Kathleen, your calculations are absolutely correct – 60% of garments sold online are not returned; on average 25% of garments are returned instead. Of those returned, 60% are due to poor fit. Indeed, Fits.me can address only those returns. Is it important? Unquestionably – returns are clothing retailers highest cost factor and reducing returns is the easiest way to impact P&L.

However, as we all know, sizing scales for clothes aren’t consistent, and that makes shopping on the Internet risky. Remove the risk, and sales increase. Fits.me simply answers the customers’ final question “which size should I place into my shopping cart?”, without this answer the sales suffer. Just imagine a world of brick-and-mortar without fitting rooms.

Jeff Bezos, CEO of Amazon.com, recently declared “we make it really easy for people to try clothes on,” explaining that they make returning products, after trying them at home, guilt free. Guess, their acquisition of Zappos.com, always welcoming returns, was for the better. Making returns easy increases sales, but is a very costly option for the retailer. It’s better to sell the right size in the first place (oh, the customers will like it too).

Today, apparel online sales remain only 9% of total apparel retail sales, but it’s share is estimated to increase to 35% by 2018. Or put another way – one in every four brick-and-mortar clothing shops will close, a trend observed today in the sales of books, and with travel agencies.

Several issues remain: size consistency of the brand, errors in self-measurement, and timely access to garments.

Many brands require their suppliers adhere to stringent size specifications. Building the target market takes years, educating the customer base to the sizing is invaluable, and yet the suppliers often fail to deliver size consistency. Saying this, some brands are better at quality control, some not so much. Fits.me is most effective with brands with high sizing quality control – and still delivers benefits, albeit lower, to those with lower consistency. At some point, the benefit becomes so low, there is no justification for the investment though.

Errors in self-measurement are a problem. As people make errors, they keep receiving clothes that fit wrong. While we are double checking the measurements entered comparing them to some 10’000 3D body scans, they will not be perfect. However, the idea is not to reduce the returns by 100%, but just deliver enough to benefit the retailer the most. 20% reduction in returns, will still increase the online retailer’s profits by 10%, on average.

Access to samples is the major issue, thus Fits.me focuses on only 4-7 generic fit sample garments. Timeless styles which will remain the same across seasons. The garments are handled once, and turnaround time can be longer, without hurting the business. True, the customers won’t be able to try on everything, but will still get a great idea how this brand fits their body size and type.
This approach requires that the retailer has even more focus on size consistency, but again, the end result will need not be perfect – we simply need to increase the sales a bit, reduce the returns (and costs) a bit, and the profits will increase a lot.

4 years 11 months ago

So if UC Berkeley can make a towel-folding robot
http://news.cnet.com/8301-17938_105-10471898-1.html
how come my UC Berkeley grad husband can’t do it? ;-)

Has anyone seen how well those optical measuring systems work in real life?
While they can measure distance, can they measure firmness? That is, some bodies are firmer than others and the ease requirements will vary.

kathleen
4 years 11 months ago

On average, in apparel online sales, rate of returns is 25%. Polos and t-shirts have the lowest returns at 8-11%, the highest rate at 40% would be high ticket fitted fashion items….I would stick to saying that 60% of returns are due to poor fit and am afraid, I do not agree with the 1999 research that cited the 17% figure.

The 17% statistic dated tho it may be works in your favor in that it is higher than the 15% of the whole (60% of 25% = 15% returns due to poor fit). I don’t wish to be argumentative but I want to make this clear because I don’t want my readers who can’t afford your technology to throw in the towel before they even launch thinking they are doomed to have 60% returned of what they ship out when actual fit related returns is closer to 15%-25%. Zappos, which practically begs customers to return stuff, only gets 25% back. But that’s beside the point. For the sake of argument, let’s leave everyone with the impression that 60% of what they sell will come back if they don’t have your fit technology, it’s still a deal breaker to use it. In addition to the points I’ve already made, here’s why:

I can’t see how the average ecommerce retailer can work it into their schedule as a matter of logistics. I have a friend who sells strictly ecommerce, I don’t know how many skus she has but it’s a lot (thousands). One of her problems in picking up a product line is product photography. When she selects a style at market, she has to beg/borrow the item from the vendor’s booth and run it over to the sample photographer so she can set up her web pages to feature it when it finally comes in -because from the time that product ships, she’s paying for it. She needs to turn it very quickly. Now, according to information on this site (and with which you concurred in your comment there):

In order to use the technology, Fits.me requires samples of each clothing item (in each size). They then take up to 2,000 photos of each product. They digitalize the whole thing, set up a virtual fitting room, and charge retailers on a CPC basis. The whole thing takes 2-3 weeks of work.

The 2-3 weeks doesn’t include ship time. A retailer would have to get their product in the door, process it and then pick and pack representative samples to you. Assuming Fit.me had the facilities and infrastructure to process the samples the second they got there, we’re looking at a 3-4 week turn around minimum before the retailer can even load the data on their site. Are you familiar with retail’s mark down schedule? Many retailers start marking things down after they’ve had the item for 4 weeks. Many retailers charge back the vendor for discounts on merchandise that isn’t moving. In short, I can only imagine that manufacturers are going to be very upset about it and it won’t take long for push to come to shove and something will have to give.

Your own stats make it a difficult sell for retailers (tons of spam comments there, might want to look into it), with estimated 3% of revenue generated by ecommerce. Fwiw, this is close to what my same friend told me, she says that even flagship brands with immaculate sites, only generate 2% of sales online. Based on your case study report of a 3 fold increase in online sales, the best retailers can hope to exact from the process is 6% to 9%. Considering that they’re incurring payables on their inventory for the 3-4 week turnaround for Fit.me to process it and even if their online sales did treble, they would still lose money.

As an aside:

but using the robots we can easily, cheapily, and en masse, reverse engineer these fit types.

It doesn’t matter whether I agree that Fit.me can “easily, cheapily, and en masse, reverse engineer these fit types” it only matters that manufacturers may fear or suspect you can, meaning they aren’t going to send you anything.

The deal killer on the whole thing is that somebody somewhere somehow is going to have to eat 4 weeks of stagnant inventory. That is going to take a lot of capital and with times as tight as they are, I don’t know who has it. To throw an extra month into the cycle would mean pushing the whole production schedule into disarray. With ZARA eating everyone’s lunch by leading the industry with a 2 week turnaround from sketch to product on the floor, everyone is trying to find ways to shorten the cycle, not lengthen it.

4 years 11 months ago

Hi Kathleen,

I find this conversation so interesting. Indeed – how about going over to e-mail or doing a conference call where I’d call my CTO to join in.

However here I’d like to write some which is of more public interest – about the return rates.
On average, in apparel online sales, rate of returns is 25%. Polos and t-shirts have the lowest returns at 8-11%, the highest rate at 40% would be high ticket fitted fashion items. Trousers have also had high return rates 30-40%. In some countries, like Germany, the return rates are much higher due to open invoicing – it encourages more customers buy multiple sizes at once to return the wrong ones.

While a return is a nuisance, the real cost for a retailer is the cost of time – usually more than 4 weeks from the initial sale. The garment sold in August, and returned to the sales cycle in October bears significant discounting costs.

I would stick to saying that 60% of returns are due to poor fit and am afraid, I do not agree with the 1999 research that cited the 17% figure. There certainly are retailers with return rates where poor fit is much lower concern, but of hundreds retailers we’ve been spoken to by now – and of those most collect the data about reasons for returns – an average between 50-80% has emerged. In the end – it’s up to the retailer to decide if it’s the fit, or are there other issues that need to be solved first (for the record, I do believe that for most online retailers there are many other issues that should be prioritized over the fit).

One of the concerns many commenters have voiced is the sizing differences throughtout different manufacturing sources, and as you point – the trade secrets on the fit of the brand. I guess, it might be due to some poor communication that Fits.me is projecting, but using the robots we can easily, cheapily, and en masse, reverse engineer these fit types. As it goes beyond the publicly interesting topic, I’ll stop here.

I really love all you voicing your opinions! Please keep them coming! And if you’d like to get in touch personally, please do so at heikki/at/fits.me (remove the slashes).