Show HN: I mixed selling t-shirts with a little bit of game theory

I had this idea rolling around in my head for the last couple of years, and I finally decided to build the damned thing just to get idea relief.<p>The backend is a typical LAMP stack with the Symfony2 framework backing the P. Varnish sits in front of all the Chef managed Linode boxes. The site will work without Javascript (up until the checkout), and the style degrades really nicely depending on your viewport (try it out by resizing your browser).<p>I'm open to all comments and critiques, so let me hear them. And if you are into graphic t-shirts, here is a code for a Hacker News 20% discount: 4803-7940-0816-5604!

I like the idea, but I have the feeling your t-shirts are overpriced.<p>I had what you could call a t-shirt problem, I purchased over 150 in the space of a year, anyway one common thing I noticed is that a <i>brand</i> (eg: Threadless, Bustedtees) that sold based on merit and not gimmick could sell at a much higher price than somewhere that was relatively unknown or sold based on some sort of gimmick. For example teefury.com is a site that does a sort of similar thing to you with time based sales, they sell at around $10 each but would never do well at $25 each.<p>My (anecdotal -- myself and friends) understanding of why people aren't comfortable spending $25 on a time-sale t-shirt is that you're selling selling based on 2 things, the t-shirt itself and the idea that it's time-exclusive, this adds pressure. If I purchase a t-shirt from somewhere like Threadless I know it's likely if I wait 2 weeks it's going to be there still, so I don't feel like "I must buy now!" and then I don't feel like I'm paying for something I might not want in a week, whereas if I see a t-shirt on your site with 1 day to go: "That t-shirt is cool! Hmm, but I have to buy it right now... and $25 is a premium t-shirt price... I might not like it in a few days... I'll skip". This behaviour is especially so with the added idea that I'm possibly buying a t-shirt nobody else likes, it makes me think maybe it's just a crappy t-shirt.<p>Your site is a site that customers would constantly be coming back to, so while you might not make such a good margin on a $12.50 shirt vs. a $25 shirt, it would be made up for in the average customer purchasing 20 shirts a year vs. 5 if they had to be considerate about their purchases.<p>I'm not sure if I explained that well, if not I can try and re-word it.

I really like this. I found myself with a gradually widening smile as I realized the ramifications of the whole concept. The brilliance from the commerce standpoint is that the freebie is by definition minimized by the number of shirts sold, and a typical loser product will see more activity than usual. From a customer standpoint, you have effectively 20% chance of getting a freebie, making the perceived purchase price 80% of $25, or $20, a very reasonable price for a graphic t-shirt. However, since it's the least purchased number getting freebies, I'd expect your actual revenue to be more like $21 per shirt, thus creating profit out of thin air. Customer pays a decent marginal price (for a hopefully quality product), you get a built-in profit from the mechanics, and the game mechanics nearly guarantee a larger volume of purchases than normal. Overall, a great strategy!

I saw this site last week or maybe the week before, and to be honest, I was put-off by the obnoxiousness of trying to get people to buy the "least popular" t-shirt. To me, it came across as you trying to unload your shittiest product through some clever trick. It had a Zynga-esque feel to it, and I didn't like it at all. The fact you give a clue like "this was the worst selling t-shirt in the last hour" makes me feel like you're tricking me, and I'd rather do business with someone else than risk getting played or ripped off. Sorry, but that's the feeling I get when I see this site.<p>If you want some additional feedback:<p>Instead of the gimmicky "you lose I win", I think you need to make it some sort of win-win type gimmick. Like the price goes down the more people buy it, and whoever buys it earlier will get some money back on their next purchase.<p>You need to make the "game" shorter, I don't have patience to wait a week until I know if I "won".<p>$25 is something I would pay if the t-shirts were good quality, like American Apparel, but you didn't list the manufacturer.<p>You need a lot more designs. I was looking for the button for additional designs, but I finally figured out you only have those 5. If I hate all 5 of them, then I won't likely remember your site, and won't come back.

I love the idea. It's really great to see something innovative here. I was almost going to say a little bit of genius, but i have a few more reservations now i have thought the idea over-- mainly that i don't think i'd see value in the 'game' (that is, saving potential) unless i really liked all the tees on offer. So that sets the proposition to me[1] as very close to any other online t-shirt retailer (would anyone suspect their favourite was going to be least liked?).<p>BUT in the very worst case scenario, you have a fantastic marketing gimmick to drive people to the site, and keep many of them coming back perhaps. I really wish there was a 'control' to compare this to, to see what effect this has on sales- it seems it will be hard to know how much the game is driving sales vs people just liking the tees. My current feeling is the idea might function better slightly tweaked.. eg. offer the shirts at $18 normally, or $20 with the freebie game.. only count the $20 sales toward the totals.. might offer some balance-- for example, the 'better' (popular) one won't currently have a lot of appeal WRT. the game, because most will rightly assume it wouldn't be least popular. Add the tweak and there is an element of double-guessing that might make more people believe they have a chance of 'winning'? Hope that made sense.. it's not a fully worked out idea, but worth thinking about perhaps.<p>Also, consider following OKcupid and having an onsite blog.<p>[1] perhaps not everyone is like me (or us here), though

Adding some type of effect to show that the shirts are actively being bought (I imagine every few seconds half a shirt 'flash') would give the sale more of an active-feeling. Right now I don't feel strongly urged to make a decision right now, but rather that I could come back later.<p>Second comment: Rather than giving away the least popular shirt, you could apply some type of variable pricing depending on when a user purchased a shirt given its popularity?

After you do this for a while, please please release your data. I'd love to see what kind of influence the star has on things, and the variation between least popular and most popular.

I like the game mechanics.<p>A story that comes to my mind is: there was a grocery store in London in the 1950s who didn't have the best of locations. They ran a contest: They would randomly select a half an hour time period (eg: 2.20 to 2.50pm). People who had shopped during that time on the previous day could come back with their time stamped receipts and get cash back.<p>The store tripled their number of customers per day.<p>This however worked because people didn't go out of their way to buy new things. They just went out of their way to buy the things they needed at a different store to get a chance to get cash back.<p>So my suggestion is this: source out the "winning" designs on threadless etc in the past 3 months. And list those t-shirts on your website only. Don't ask your audience to submit their designs.<p>And announce this fact. Re-assure people that they are buying the same "hot" stuff they would buy on other t-shirt sites. But they could just earn a free t-shirt too.<p>Take away the neg that comes with the "least popular tshirt" tag.

Cool idea, agree with the comment about regrettable purchases so I think you need to figure something out for that.<p>I think there's a lot of space to apply game theory / gamification in e-commerce. Still need to get a clear understanding on the regulatory framework for this.<p>Also, which fulfillment service are you using?<p>Good luck!

Shouldn't that be "I mixed selling t-shirts with a little bit of <i>gamification</i>"? I don't really get the game theory angle (as in <a href="http://en.wikipedia.org/wiki/Game_theory" rel="nofollow">http://en.wikipedia.org/wiki/Game_theory</a>). Very clever idea in any case!

I'm friends with seekely, and we have gone back and forth about the idea of framing the "winner" as the "least popular", which as a negative connotation, vs. using terminology like "unique", which has a more positive connotation. It could be that by using the more negative language, people would believe that the free t-shirt is a consolation prize for having bad taste, as opposed to being excited about being special somehow. However, it's a completely subjective assessment in either direction. What did you all get as a first impression (or did it even register with you in that way)?

On first glance, I feel like people would simply choose the shirt that they think looks the least fashionable, which just leads to regrettable purchases.

One slightly annoying thing I noticed. If I select a t-shirt ie: <a href="http://getnifty.com/market/1001/product/1002" rel="nofollow">http://getnifty.com/market/1001/product/1002</a> the sizes appear<p><pre><code> 0S 0M 0L 0XL </code></pre> which for a second till I ran the mouse over them realised it was a form selection. Can you pre-populate or add spacing?

Interesting, but I'm not sure I buy the "I'm unique" logic. It reminds me of "big-choosers" from the Toilet Paper Problem: <a href="http://gi.cebitec.uni-bielefeld.de/teaching/2007summer/jclub/papers/toiletPaper.pdf" rel="nofollow">http://gi.cebitec.uni-bielefeld.de/teaching/2007summer/jclub...</a><p>If everyone was a big-chooser, you get into a dangerous situation because rolls are being emptied at roughly the same rate.<p>Similarly, I am led to wonder what percentage of your users will be making their choice based solely on the coupon (i.e. least popular shirt at the moment). If it's a high percentage, then each shirt should approach 20% of sales. Hardly unique at all.<p>By the way, "Virgin" is an awesome design. Reminds me of Dali. I can't imagine opting for any of the other shirts in order to try to win a coupon.

I so wanted to buy the bicycle one for my girlfriend.<p>When checking out my cart, it didn't allow me to choose another country than US, and I'm in Europe. Too bad for me (and for her) (and also for you, who lost a sale ;) )<p>Great post, great idea, though ! Keep it up !

There are quite a few sources in this thread for quality t-shirts, but I don't wear cotton anymore.<p>Curious if there are any t-shirt sites that sell on Nike dri-fit, Under Armour style shirts. You know the kind, moisture wicking etc. They tend to be polyester or even bamboo.<p>My favorite at the moment is this one I got from Lids. It is AWESOME. <a href="http://www.lids.com/NCAA/Hawaii-Warriors/20304981" rel="nofollow">http://www.lids.com/NCAA/Hawaii-Warriors/20304981</a><p>I just wish I could find more! Anyone know of any sites like that? And I of course I know I'd pay a premium for that, so no worries there.

Interesting idea! Graphic tees are a great market to do an idea like this, as well.<p>People who have the disposition to wear a lot of graphic tees seem like the same people who would spend $25 on some generic tees just for the game aspect. Personally, id rather bargain hunt on Gilt or something for some high quality, well designed t-shirts, but im not your target audience.

Interesting, but my reaction is, I'll just pick my favorite (if any) and not worry about getting the special. It's not worth buying one I don't really like just because it <i>might</i> turn out to be free.<p>On another note -- would really like to be able to pull up some much larger images of the shirts.

This reminds me of the game mechanic in Puerto Rico ( <a href="http://boardgamegeek.com/boardgame/3076/puerto-rico" rel="nofollow">http://boardgamegeek.com/boardgame/3076/puerto-rico</a> ) where money accumulates on roles not chosen to incentivize people to pick that role.

My hands are itching to write a scraper that checks the site every 30 minutes or so and tells me just before the countdown ends which shirt was the least popular.. What's stopping me from doing so other than not being a jerk?

What a fantastic idea, I am a long time customer of Threadless, TeeFury etc. You see a lot of clones springing up but you have done something really interesting and unique, will defiantly give it a go!

You sick bastard!!!! Now I had to buy 5 shirts just to make sure I get my free one. How dare you! KHAAAAAAN!!!!<p>edit: set self destruct code: kirk0001 &#60;-- lol at complexity.

The game theory stuff is neat but to me inconsequential.<p>I just went to buy a really cool shirt but can't - USA shipping only :(<p>Pity! I doubt I'll ever come back to this site.

Does this count as a Keynesian Ugly Contest?

Uber cool concept. Time limits + freebies = good bag of tricks and sorcery. +1 for your logo.

It might be just me, but I can't see what the shirts look like in Chrome.

I would really love international shipping for the tshirts.