Be advised: this thread is not for serious in-depth discussion of weighty topics (we have a link for that), this thread is not for anything Culture War related. This thread is for Fun. You got jokes? Share 'em. You got silly questions? Ask 'em.
PaperclipPerfector
19hr ago
(text post)
724 thread views
Be advised: this thread is not for serious in-depth discussion of weighty topics (we have a link for that), this thread is not for anything Culture War related. This thread is for Fun. You got jokes? Share 'em. You got silly questions? Ask 'em.
Jump in the discussion.
No email address required.
Notes -
I’m writing this post on the behest of @100ProofTollBooth who asked me to explain my industry (e-commerce) and AI’s effect on it. I apologize in advance if this is not interesting or not what you had in mind when asking, but since I’m easily flattered I’m happy to talk about myself for a bit.
So to explain how I got where I am now I need to explain my background. I have a BA in fashion design. I started a fashion line around the time that I graduated college, which was very fun and somewhat successful. I sold some of my designs at boutiques in NYC and Asia which had always been a dream of mine and that was awesome. I maintained the line for about 7 years until 2020, when a variety of factors pushed me to end the line. It was partially covid making things weird and difficult, it was partially my waning interest in the aesthetic I was working in, but most importantly I was itching to travel and leave the US for a while which would be impossible while I was chained down to a studio with a proper line. So I pivoted around 2020 to focusing on my side gigs. I had done a bit of freelance design for a few brands but I really disliked working with other people (even if one of the designers I worked with was a dream designer who I still respect a lot.) My brand I had managed entirely online through a direct to customer model (outside of the wholesale boutiques I mentioned earlier.) So I was familiar with e-commerce through that (as well as having shopped online since 3rd grade as a customer- I still have the same ebay account I opened when I was 11, I’m proud of this.)
Anyway, while I worked on my brand I also dabbled in the Print on Demand industry. The earlier incarnations of this are Zazzle and Cafepress, later perfected by Redbubble and Teepublic and a handful of other platforms. Believe it or not I used to make a lot of money on Redbubble. (More on that later.) But anyway, as I wanted the freedom to travel and fulfill orders while making money away from a studio, I decided to pivot away from my physical brand’s business and move entirely into print on demand. This was a combination of through platforms like Redbubble and traditional marketplaces like ebay, Etsy and amazon.
Today I make over 90% of my yearly earned income from print on demand items that I design myself. Designing items myself gives me a bit of a moat between myself and the bulk of the drop shipper industry people who either have to buy designs from other designers or have to sell the same generic goods that everyone else is trying to sell so they must differentiate heavily on marketing, brand positioning, funneling, conversion tactics or whatever. All of these things are not very exciting to me so I am glad I can innovate on design and product offering as a designer rather than having to think about marketing (I hate advertising, I do not pay for ads for my products, I block every ad, I feel like dying when I see an ad irl etc etc)
Hopefully this all makes sense, I am being slightly vague in certain specifics just because it’s a highly competitive industry and I don’t want to be too helpful but I think you can get a broad sense of what I do from the above. Now to respond directly to @100ProofTollBooth’s questions, paraphrased slightly for format reasons:
A) What is my perspective on my industry?
I guess I can answer this from a few different industry perspectives. I will answer about fashion, about Print on Demand, and about e-commerce in general.
I think the fashion industry as a whole is not really terrible. I do think that the industry does drive innovation and prioritize creativity and artistry from people. In certain segments of the industry there is an attempt at conserving craft and tradition that I think is valuable. Being a female dominated industry it does have a tendency to foment woke witch hunts (John Galliano’s firing from Dior is still terrible, I suspect McQueen’s suicide probably had some degree of disillusionment from the politics of fashion for example) but people seem broadly to be over this currently and have some understanding of the cringeness of being that way.
The print on demand industry- as a designer, I value the industry a lot, as it offers me massive flexibility and a huge opportunity to make money without having to put in hardly any investment into inventory or development. I may only make a margin of 25% or less on each individual item I sell but the flexibility it offers is very good. Admittedly my switch from fully designed luxury goods under my own label to basically utilitarian POD items was a bit of a blow to my ego but the advantage of being able to get paid for very little work helps soften the blow.
Ecommerce in general. I think it’s good. It is basically a glorified Sears catalogue mail order service. It isn’t much different today from how that worked back in the 1800s. Amazon’s 2 day shipping is great when it’s available. Aliexpress and temu are really crazy, it reminds me of markets I shopped at in Thailand, where everything is incredibly cheap and abundant, of course giving those sellers access to the US market who are willing to pay American prices for their goods is a huge trade imbalance that benefits both third world middlemen and low and middle income Americans. Many Asia based drop shippers infringe designs that I’ve made and sell them on platforms like walmart.com. It is so ubiquitous that I have stopped looking and issuing takedown requests. I am not a fan and wish they would not do that.
B) AI’s effect on my industry
I suspect AI is changing a lot of things behind the scenes in fashion companies in ways that are not visible in their marketing, product offering or brand messaging. I have noticed a lot of shorts on YouTube are using obviously AI generated market copy which I think is glaring and tacky. I noticed this from brands like Sotheby’s and Balmain who should know better. I know there’s a somewhat trollish brand that is using AI generated imagery in their designs but I can’t remember the brand name (it’s similar to Praying but it isn’t that brand. Praying may use ai generated imagery too but I’m not sure.) Certain brands that are “edgy” can get away with using AI generation, a handful of other brands are getting screeched at by their social media followers for using gen AI, it just depends on the customer’s opinion when it comes to high end brands.
The first casualty of AI in print on demand was Redbubble. They had already been slowly tapering off the payouts and royalties given to artists but as soon as generative AI came out they clamped down hard, introducing a weird tier system. I had like 7 RB accounts at the time and they put 6 of my accounts in the crappy/low royalty tier, and one account in the high royalty tier. The one they put in the higher tier had like 12 rather bad, early gen AI designs, so if they were trying to put all the AI accounts on the lower tier, they failed. Naturally I only uploaded designs to the high tier moving forward, which they then deranked to the low royalty tier a few weeks later. They introduced a terrible system where the higher you price your items, the greater the take that RB takes from you, de-incentivizing artists to come up with designs that people are willing to pay more for. I have not uploaded anything to RB in quite some time as a protest to their system.
I will interject here and say that I do use gen AI for a small portion of my print on demand work. I would probably estimate that only about 10% of my yearly income comes from anything that AI has touched creatively in any way. This is partially because I have thousands of designs I generated before gen AI was even a thing that continue to earn me the bulk of my income, and partially because scaling gen AI is still quite slow and slower than scaling non-AI designs.
AI’s effect on e-commerce. Again I believe most of the innovation is behind the scenes here. I have used gen AI to generate product images on Etsy. These images look really good, in my opinion, but they have not increased my sales at all. In fact many of them are off-putting to people. The ones that do work are ones that look like casual iPhone photos. I can generate beautiful high end imagery of things and the crappy fake iPhone photo will outperform the beautiful one every time. It is just what the customer trusts, is used to, and attracted to. I don’t really relate but it’s not really up to me.
So, generative AI is still proving to be relatively ineffective for design generation, and for marketing purposes. But what I do use it for: brainstorming relevant terms and keywords, writing marketing copy (with good SEO. Users on themotte accused me of generating slop here- no, that’s bad SEO. Good SEO is concise and has a few highly relevant keywords. Bad SEO is a bunch of irrelevant slop. Regardless, these are short sentences that I guarantee hardly any human reads - it is mostly read by bots at this point.) I built an app in Gemini ai studio the other day that will generate good SEO titles and brands and descriptions that output to CSV which saves me time uploading designs. Before I did this by copy pasting formulas in google sheets, this will save me quite a bit of time moving forward.
C) I'd love to understand how you go about designing a new product, testing for demand etc.
It is a numbers game. I have made over 40,000 unique designs over the past 10 years or so. In the beginning I didn’t know what people wanted so I made 100 different designs at a time, then had to wait and see what people bought. Out of 100 designs maybe 10 of them would sell at all. I would take the 10 that sold and make 100 more variations of each of those, then just keep doing that. In the beginning I kept thinking it was frustrating because I didn’t have any sales data to draw from- so I could be putting up 50 things that no one wanted and didn’t know it. I look back at those times and realize I was correct, that if I had sales data back then I would have been able to grow much more quickly. But now I have the sales data so I can use it to generate more and more desirable products for people. My sales history/data is the most valuable thing I am generating.
D) What the elements of success in ecomm are and what "pros" do versus what "chumps" do.
Hrmmm. You have to be able to spend a year or a few years making nearly no money from e-commerce. The first year I switched to all POD I only made like $9,000. I lived off money I saved from my earlier business endeavors. But once you have the data and the momentum you can just scale up and be successful in whatever niche you find profitable. Be willing to chase the money, none of the niches I sell in are at all interesting to me. At first this irritated me but I value my customer enough that I don’t really mind anymore. I mean I am grateful that there are people willing to buy the things I’m willing to design and that’s very valuable to me.
I guess the “chumps” and the “influencers” you mentioned earlier just lack the dedication over years to make it work. It’s not a get rich quick scheme (I guess it’s possible if you get lucky but it’s not likely to last over several years.)
I have always been a fan of really tedious games like Harvest Moon, Stardew Valley, or even worse, those desktop based games like Farmville or Factorio knockoffs. At some point in college I thought if I found these so addictive I might as well use that time and energy and pour it into something that actually makes money, and nothing is as tedious as managing a ton of POD designs across 15 different platform accounts, so in a way I did just make those tedious games into a career. By the way, at this point, my career is nearly entirely passive, I actively work on stuff for like an hour a week which is mainly ordering POD items manually from various producers. I could automate this but it’s still so little work that I don’t mind doing it myself. The largest bit of work for me is that once a year I generate and upload new designs which takes a few weeks of working every day. I’m afraid of automating this process because it violates terms for a certain site so I just do it manually still.
——
I hope that answered your questions and gave you food for thought. I’m happy to answer follow up questions as well.
In return, you said you know about strippers, bars, hookers, and the nightlife industry. How do you know about those things? Do you work with them or just spend a lot of time with them? I’m gay, you have any insights or experiences or information that I’d be interested in as a fan of, um, gay strippers and gay bars? Also, I never heard the term Hick Hop before you used it. I was imagining it was like mostly a wigger thing but I see some black people involved in this. Where is hick hop geographically centered: The South, the West, Appalachia?
EDIT: I forgot I wanted to add another paragraph at the end that gave a more broad response to the AI question. I answered how AI HAS changed my industry from my perspective above, which basically, it hasn't changed it very much. But moving forward I can see a few huge changes on the horizon. Amazon has already started AI generating tshirt designs and selling them on their website. They aren't very good but neither are a lot of the traditionally designed designs that they sit next to. Amazon has been doing a similar thing for years - competing with their own 3rd party sellers by cutting them out and sourcing from the same suppliers the same items and usually undercutting them in bulk purchases. The fact that they are using AI to compete with designers is not a terribly huge change but is slightly different. You'd expect marketplaces like redbubble and etsy to be overflowing already with AI slop, but it's only encroaching on certain segments. Bad AI slop is unpopular and doesn't sell, good AI slop is good enough that you can't detect it. Glaringly bad AI goods just don't do well in the algorithm. I have seen crazy AI generated products marketed on platforms like temu but they were already pulling ridiculous marketing with photoshop so it's not a huge change there either. I imagine all these changes will continue to ramp up over the next few years and eventually either the algorithm will just hide anything bad and ridiculous or every website will break in a deluge of AI nonsense. I'd bet on the former for now.
I am not worried about AI in the near term, because we're going to be in the stage where early adopters/people who pay attention (like me) can benefit for a little while. I am small and light enough to be able to adapt to the changing landscape. Once large firms get very efficient at using AI in ecom, the situation might change, which is why I diversify my business as much as possible and, failing all that, have always saved a ton of my income and can switch back to a business with a larger moat at any time (I want to have a proper luxury brand some day using my own name, and I don't see AI making this impossible barring some postindustrial all knowing sci fi tier AI which I'm skeptical of coming in the next 5 years at least) or, barring all that, simply retire.
Thanks for the effortful writeup.
Ball is in my court and I'll get to working.
More options
Context Copy link
More options
Context Copy link