Hacker News Reader

Comment by melanie_io | Hacker News Reader

I was actually discussing this very point with a good friend of mine who also runs a fashion startup the other day.

To refine what I said: it's not that discoverability is not a problem, it's that all of the new fashion apps (from Lyst to Fashism to Inporia to Svpply to Google Boutiques) seem to have only increased the "noise" versus decrease it. In fashion, customers pay for the edit: a small, curated collection of products that an editor has determined best fits her customer's profile. To argue that we can somehow replace this very right-brained activity with crowdsourcing or algorithms is untenable.

View on HN

> To argue that we can somehow replace this very right-brained activity with crowd-sourcing or algorithms is untenable.

For now maybe. Ten years ago if you asked me if I'd spend 2 hours watching a movie because a computer said so, I'd laugh. Netflix recommendations, while not perfect, are still very good. Had I not told Netflix that I loved District 9, it would not have suggested Torchwood: Children of Earth to me. Had it not suggested Torchwood, I wouldn't have discovered that Doctor Who was back on air. Right now I get about 50% of my entertainment discovery done via algorithms. Sure, fashion is difficult and subjective but I don't see it as being any different from music or entertainment in the big picture sense.

Btw, that was a very well-written article. Thanks for sharing.

It is different. When buying clothes people ask questions like "Will my girlfriend think I'm sexier?" or "Will the kids at school think I'm cooler?"

Fashion is a kind of performance -- people see what you wear, after all -- so there's an inherent social dimension that isn't obviously present like it is with movies on Netflix, which you watch in the privacy of your own home.

People, not robots, will win the day when it comes to shopping online.

The core of these recommendation algorithms is often a flavor of nearest neighbor analysis. In this case, the computer is using attributes to find people who are similar to you, and then suggest to you things that they liked.

All the "robots" do is find people who it thinks you would like to emulate. That sounds like the same thing the fashion industry has been doing for decades.

Robots don't merchandise, that's the problem. For many people decisions about what to wear is very intimate, but algorithms are bloodless.

The who, where, and how matter much more in fashion. Algorithms miss that.

I'm not really opposed to what you're saying, but it does seem to me like you dismiss the big human factor in these algorithms. It's not like it's just a system saying "You bought a red dress last year, here is a list of other red dresses you may like" -- Because the quirky frock she bought for work last year doesn't mean she wants to see cocktail dresses that just so happen to be the same color, or from the same designer, or in the same price range.

Instead, it's correlating prototypical woman A to prototypical woman B based on available demographic data and past purchase history, and suggesting to "A" some of the things that "B" purchased. The emotion and intimacy there was the human act of "B" purchasing clothing that she connected with.

I think a strong algorithm may surprise you.

I remember reading about the Netflix contest and how one of the algorithms categorizes movies into it's own statistically relevant categories*.

Once a comparative algorithm is used and the dataset gets large enough, I have no doubt the recommendations will start to get real good real fast.

http://devlicio.us/blogs/billy_mccafferty/archive/2007/01/02...

What I'm trying to say is this: it's not about a "match" between the product and the potential consumer.

The who, what, and how of the recommendation matter almost as much as the product itself. Just the fact that I know this recommendation came from a human counts for something, especially if it's a human I trust or respect when it comes to fashion (not necessarily a friend).

e.g., a celebrity wearing a shirt and having it sell out the next day.

Movies are different because their consumption isn't inherently conspicuous. I do it alone and talk about it with friends if I choose, but everyone sees the clothes I wear no matter what.

For example, I can choose to hide the fact that I love Katy Perry, but I can't hide the fact that LVMH made my handbag.

How you dress is a performance, and so the decision to wear something is filtered more rigorously through a social dimension than watching a movie or listening to a song is.

In the context of my most previous reply above -- how would you feel about reccomendation algorithms that pull more visibly from your own social graph -- Friends, yes, but also people you follow on Twitter and Like on facebook.

That seems almost a perfect marriage of my comments on the innate human quality behind a nearest-neighbor algorithm your comments here.

Whether it's a friend or an impersonal algorithm making the recommendation, you're still watching the same trailer before deciding to purchase a film. On the other hand, no amount of precise algorithmic matching of tastes and sizes can substitute for feeling the softness of the fabric and being reassured by your reflection that your bum doesn't look big in it before you buy. The costs involved are quite different too...

I suspect the simple solution to that is to have low cost models in a range of sizes try on each item. Let's say it costs 100$ a per size * 10 sizes = 100$ per item in your catalog * 10,000 items in the catalog = 10 million. Which is within the range of a well funded start-up. So it adds 1k/item and some lead time it's also a huge barrier to entry and probably well worth it.

PS: Don't forget Zappos sells shoes online where fit is both harder to come by and more important.

If I'm in the target market for high fashion, I'd probably list "clothes shopping" as one of my hobbies. I don't; I buy cheap clothes which retailers make minimal profit on. If I was in the market for designer clothes then my tastes wouldn't be price elastic, so saving retailers' costs by substituting cheaply made model videos for the user experience of actually seeing and touching the clothes in a stylish location isn't much of a purchase incentive. By contrast, unless the local Blockbuster outfit offer exceptional advice, actually going and physically picking up DVDs is a mere inconvenience and I'd much rather pay less to skip it altogether.

I'm not arguing it's impossible to sell fashion items online; it evidently is. It's probably also possible to sell wallpaper over the telephone. That doesn't mean the economies of selling over the internet necessarily lend themselves well to disrupting a sector that enjoys massive profit margins through making far more effective emotional inducements to purchase than a photograph and a like button.

I would argue that there is a huge market for people living in Arizona that can't get to any of those high end stores without flying there that still buy 'high end' fashion. The market is fit 34 year old doctors making 300k that don't live near such stores it's really poorly catered to. When a catalog sell 5,000$ dresses based on pictures that look nothing like the customer that's just ripe for disruption over the web.

You could also do the same thing for the mass market, but I don't think the average American really wants to see someone that actually looks like them trying on the clothes. Granted, there are main stream markets other than fat that this could work just fine, baby clothes, teens, big and tall etc.

Yep - there is a huge market for this, one that Net-A-Porter and Mr. Porter dominate (~$200 million in revenue). They take care of the sizing issues with flexible returns and personal stylists. Since 80% of their revenue comes from only 2% of their client base, they can afford to hire personal stylists for higher-volume clientele. Check it out.

I totally agree about the increased noise making discoverability worse. In fact, I have spent very little time on any of the new sites we're talking about just due to sheer exhaustion from market oversaturation. And that's crazy, because I'm probably their ideal customer.They definitely aren't reaching me, and they should be.

I tend to agree with you about curation versus AI, but I am not 100% sure. Years ago I was skeptical that collaborative filtering would ever work at all. As the poster below notes, it does pretty well for Netflix and sometimes for Amazon too. I could at least imagine some sort of AI recommendation system working for the masses, but I doubt it would ever appeal to the highest end customers.

Is this a symptom of startups wanting the homerun rather than building a smaller business around the edit? Seems to me that a business built around an edit fit to a customer's profile takes a personal intuitive touch or time and a ton of data whereas the current crop of fashion apps is throwing everything against the wall in hopes of attracting large numbers (of less discerning) users in the belief that # of users = high valuation (as opposed to the quality of users).

I hate to read startup blogs/news and they talk about exit strategy. What if Jobs, Gates, Page/Brin, and Bezos worried more about the exit rather than building a sustainable business. (I know it's off topic, just irks me.)

What are your thoughts on Pinterest? I use it all the time - recent purchases as a result of Pinterest include a bunch of apartment furnishings, an iPhone case, some new t-shirts, etc. And now when I need something I have started using it as a search engine. I do think it has solved the problem of both breadth of options and targeted results.

Disclosure: I am female.

Pinterest is one the coolest/simplest site I've discovered this year. I have yet to buy anything straight from it, but I search for styles through it before going out and buying clothes.

If I must. Disclosure: I am a guy.

Reply on news.ycombinator.com