By Will Gu @ Mindless Blob
It sort of started with my girlfriend messaging me about a promo that StickerMule was having for $19 custom printed t-shirts. She asked me to think about a design I might want, and with Christmas just around the corner, I thought of secretly getting a t-shirt as a gift too.
I'm not a designer —but I had an idea that it might be funny to have a t-shirt with a design saying, "Karen's My Bestie," and something to do a white and brown bear as we often referred to ourselves as "The Bears" (my childhood nickname was GuBear and Karen becomes KareBear pretty easily).
I work as a freelance software developer, so I'd already used AI image generation quite a bit, and sure enough I was able to get a decent design after a few tries.
It was exciting. Just several years ago I probably would have needed to spend more time finding and tweaking a design to fit this concept or even hire a designer to get it done, but with the help of AI and a little bit of prompting I got a result I liked within a few hours for a dollar or two.
The final prompt to generate the image above was simple enough:
"a logo that says "Karen's my bestie" with two bears in the background, one is white and small, and one is brown and a bit larger"
Christmas came along and the t-shirt was a hit. I still wear it in public on occasion, and it's a great way of expressing that... "Karen's my Bestie."
The thing is, after Christmas, her birthday was just a few months around the corner, so I started to think of more gifting ideas. I've always been a believer in making gifts personal, and didn't want to just buy something from a store and call it a day. I wanted to give it a bit of thought and make something custom that she couldn't/wouldn't just go out and buy herself. Of course, I didn't really have any skills myself where I could crochet or knit or make the gift myself, but I wanted to make it special and at least come up with the concept or idea for the gift.
At the end of the day, I decided to go with photo puzzle of some of the moments / memories we had in 2024 and began researching where to get it done. Surprisingly, it was a bit harder (and more expensive) than I thought it would be.
First, I needed to combine all the photos in an artistic format. I didn't want to just line up the photos in a rectangle and call it a day, but there weren't a lot of options for making an artistic rendition of a collage available. Most puzzle making sites came with templates, but didn't actually offer any service for creating a design.
Ultimately I stumbled on https://collager.ai/ and spent $10 creating the image for the collage —which was more than I expected going into the project.
With my photo puzzle image ready, I needed to find a place that would actually print the puzzle. I thought this step was going to be easy, but actually since I was shipping to Canada I was bombarded with high shipping costs and less options that existed within Canada that could handle it. I ended up with https://piczzle.com/, and my final cost was $61.57 USD. For the 500 piece puzzle with a box, it cost $39.95 and shipping $21.62 (more than 33% of the cost was shipping). Add the cost of the image generation in the first step and my total cost was $71.57 plus the time it took me to research and ultimately place the order.
It was more than I expected, and I really felt that my total cost could have been closer to $50, not to mention the time it took me to look through all the different vendors and generate the photo image. All in all, it wasn't as straightforward a process as creating the t-shirt in December, and it dawned upon me that for future gifts I'd have to repeat the process for each type of gift I wanted to give. AI was helpful in some of my design requests, but it couldn't do everything, and it definitely couldn't make the product I wanted to gift, and it didn't seem like there was anything out there connecting these two pieces.
I started researching into AI design tools that gave me more control than just prompting in ChatGPT and sites that did more than just print/make 1 type of product and entered a deep rabbit-hole —what I found out was:
1. AI companies make money by focusing on the subscription aspect of their technologies and servicing millions of users. Producing goods would complicate that and make little sense for their bottom-lines
2. POD (Print of Demand) companies make money by focusing on being a production company that sells products to small businesses trying to turn a profit from up-selling from their set rates. Helping with designs would complicate that since small business owners will come up with their own designs to try to sell anyways.
3. Small businesses that are willing to take on custom / low-quantity orders need to mark-up their prices because they rely on marketplaces like Etsy / Amazon that charge them fees, hand-make their products (time-consuming), and/or use POD companies for production.
But in my mind, I couldn't let go of the idea that some of the prices were unjustified and the process could be simpler and more accessible.
I created a mock order in Printify, a popular POD company, and these were my costs for 1 custom printed magnet, t-shirt, and greeting card.
Other POD companies presented a similar issue where each different product added a $9 shipping charge. So while the t-shirt may only cost $10-15, the shipping cost was at a minimum $9 on top of it. For 1 t-shirt the price is still reasonable, but if you're making a gift box for someone, you'll likely want to throw in different items which is where the shipping fees start getting crazy.
For adding a mug and another t-shirt, suddenly my shipping cost went to $42.25.
Ecommerce really is amazing since it offers us options we may not have had in a physical store, but while shipping costs can be expensive for companies, it didn't seem right that my shipping costs were literally more expensive than my production costs.
I was also baffled by Etsy stores selling custom wall art for almost $100 for an A4 sized piece of paper.
Or Etsy stores selling AI generated art for $20.
Now I understand the need for businesses to make money, but it has to make sense. When I looked at the shipping costs for Print of Demand companies, the overall cost of items from independent / small business sellers, and the lack of innovation in getting AI designs into consumer use-cases, I felt like there was definitely a void to be filled.
At the end of the day, I wanted to create a custom, personalized gift with the help of AI, at a price-point that made sense, and I realized that the reason why prices were like this is because everyone in ecommerce wanted to scale globally. Reaching lots of people meant a few things:
1. Incurring high shipping costs due to global shipping
2. Spending money on ads to reach people
3. Creating a website and/or paying platform fees to have an internet presence
Each of these added to costs and didn't include things like buying product materials, making the product, shipping boxes, and any other labor and material costs associated with getting a real product to someone.
And while the costs feel necessary for any business, they aren't ultimately valuable to the end product that's being delivered to the customer. Maybe it's naive, but I want to believe that cutting these costs and focusing on the core product and service will be the main differentiator in a saturated market.
That means offering shipping only to Canada for now to keep shipping costs low, relying on organic / social advertisement over paid, and cutting out as many middlemen as possible so we can pour most of the value into our service and product: making custom gifts that are truly special.
We're combining the latest technologies in AI and being conscious about our choices in manufacturing to make this a reality.
I hope we can help make your next occasion memorable and special, and bring your heartfelt intentions into reality.