Fashion chatbots are back – but not without conflicting results

Fashion chatbots: when did you last use one? The AI-powered chatbots for fashion made a brief appearance in the business a couple of years ago. Both Burberry and Louis Vuitton launched Facebook Messenger services primarily for shopping, while the chatbot Epytom did something similar but with styling tips instead, built on the contents of your own wardrobe. Somehow this phenomenon never really caught on though. The services were percieved as too clunky and and not very user friendly.
But now, a few years later, things might be changing.
Improvements in AI technology and the increasing tendency for shopping on mobile, is now cited as reasons to why AI-powered services might make their way back again, Vogue business reports. The development is described as a way for brands to speak with Millennials and Gen Z in their language. And the ones driving this right now are said to be personal shopping services such as Threads and Jetblack, who has seen an increase in interest in customers.
What I find a little discouraging though, is that these bots mainly seem to aim at increasing sales in a time when climate issues are at the top of the agenda, and increasing fashion consumption is a big problem. But hopefully there is a way for brands and services to use the bots for other purposes as well.

Physical stores are adapting to the digital consumer

How to serve a customer that want the convenience of online shopping while still wanting the human touch of the offline experience? That is the number one question for brick and mortar stores these days.
According to The Current Daily, London’s Westfield shopping mall just made an attempt to steer in that direction with its AI powered pop-up store. AI provides data on what’s currently trending online – and every morning staff will stock the physical pop-up store with those items. The idea is to give customers the opportunity to experience the items live and really find out if they are right for them. The Trending Store, as it is called, will open on July 3.
Meanwhile overseas, American department stores Kohl’s and Nordstrom are adapting to  new consumer behavior by offering services and not just products. Kohl’s have partnered up with Amazon to allow customers to return products purchased online, in their physical store. Nordstrom on their hand have opened several smaller Local stores where they provide online-order pickup, gift wrapping and other services while not carrying any inventory, Digiday reports. 

Stil & Teknik taking the stage at Breakit Beauty & Fashion Tech event

On April 4, it’s time for me to take the stage again! This time I have the honour of being one of two moderators at Swedish startup magazine Breakit’s event Beauty and Fashion Tech.  Along with Breakit’s business development manager Camilla Björkman I’ll be introducing an exciting line of speakers and panel members, including Dr Amanda Parkes of Future Tech Lab, Erik Lindvall of Streamateria and Erica Blomberg at Swedish Fashion Council. See the full program right here.
Skärmavbild 2019-02-19 kl. 10.16.00

A new take on long lasting fashion

Imagine a shirt that gets passed on from mother to daughter to granddaughter over the course of 20 years. During this process, the garment changes print and colour and ultimately gets back to the brand for reusing in new ways the last decades of its life, first as a jacket lining, then as an accessory.
This thought experiment is part of the Circular Design Speed project conducted this year by Swedish clothing brand Filippa K, research body Mistra Future Fashion and University Of The Arts London (UAL).
“If you can make a garment last through the process of reinvention in reasonable, commercially available and viable ways, you replace the purchase of a new product. The lifecycle assessment of the service shirt against a standard polyester blouse showed significant climate change savings.”, project lead, prof. Becky Earley, Co-Director, CCD, says to FashNerd.
I think this kind of thinking is exactly the industry needs right now. What if we could reimagine the whole idea of what fashion is supposed to be? What if we could rethink the whole set of attitudes surrounding it, keeping the idea of fashion as something living and constantly changing, while still managing to apply this to considerably fewer items? Applying the concept of change to the item itself instead of constantly changing the item?
It’s an exciting idea.
Personally I’m very excited to see in what ways digital technologies could help enable this change. AI is already being used by some as a tool in the design process, maybe it could be of assistance in this context as well, reimagining the next phase in the item’s life cycle? Not to mention AR, with its ability to let us enhance reality.


Can this AI sizing tool reduce returns?

Almost half of the clothing items bought on online in the UK is returned, mostly due to incorrect sizes. This is a huge environmental problem. And many are trying to solve it, as I adressed in a recent post.
Now, online fashion retailer Asos is introducing a new online sizing tool that attempts to  take a step in this direction too. Their fit assistant combines machine learning with a visual questionnaire in order to take into account a wider range of body types and shapes. That is, not only height and weight, but bust size, belly shape, hip width, age, and also what kind of fit the customer is looking for.
Then they blend this customer data with recommendations from the customer’s previous purchases and returns, along with what other customers with similar body types have been happy with.
That’s lot of data though, and quite personal too, which means you have to be ok with giving away all that to a retailer.
The question is also if getting the fit right the first time, necessarily helps lower returns (and thereby the number of harmful transports). Maybe we’ll just shop even more? Worth thinking about.
Skärmavbild 2018-11-26 kl. 10.22.03
Screen shot from Asos Fit Assistance.

Yoox starts new label generated by AI

AI as a creative tool? The idea is getting more and more attention among labels and designers, as I have adressed in previous blog posts. The other day the italian online retailer Yoox (who is a part of the Net-a-porter group) announced a new, yet unnamed label. The new collection will be informed by data, but still be designed by the human creative team, The Current Daily reports.
“By using the data, we think the creative team can interpret better our customer needs going forward, Federico Marchetti, CEO of the Net-a-porter group, said when speaking at a Wired Smarter conference on Monday.
The YNAP aren’t newcomers when it comes to AI. Their logistics center is already fully automated and they have also given their personal shoppers an AI tool to help them give better advice to customers.
Using it in the creative process though, opens up for concerns about what it does to creativity.
Do human creativity and machines really mix? Marchetti wants to explore that. Maybe in the future we will see labels like ”Made in Italy” replaced by ”Made by humans”, he suggested.
“Man is about emotions. It’s about beauty. It’s about feelings. The machine is about speed, information power and the future. Can these two worlds co-exist? We must make choices to strike the right balance.”.

What algorithms might mean to style

The fashion world is currently being flooded with all kinds of tech. Some of them aim to make online shopping easier, others to technically enhance the clothing itself, while yet others are all about the holy grail of it all: style.
That mysterious quality close to the concept of taste – hard to grasp, hard to just acquire. More like something you either have or don’t have, a personal trait that just seems magic in a way.
But what happens when technology starts to claim to know something about style? Is it even possible? Well, there are several attempts actually. Like Amazon’s style assistant Echo Look that analyses your outfit through a combination of algorithms and human fashion specialists. Meanwhile StitchFix sends users clothing suggestions by cross-referencing a client’s preferences with what others of similar age and demographic have bought. Meanwhile works with personalized avatars that can try on digital clothing samples to help you see what it might look like on your body, while Net-a-Porter is trying a type of tech that will creep right into your data looking for info on what you plan to do next (events, trips and the likes) – and suggesting purchases accordingly, as reported in The Guardian.
But is that style though? And does all this help us require it? Well, I guess that depends on what we feed the algorithms. If we tell them to check for mass approval (likes on social media for instance or whatever has worked in the past) we will get just that and no progress at all. But on the other hand, are we comfortable programming algorithms to be audacious and bold the way our human style role models are?
Maybe that is the kind of trait we still prefer in humans.

Skärmavbild 2017-05-03 kl. 10.11.15Skärmavbild 2017-05-03 kl. 10.10.04