Great analysis! I've been working on a startup merchandising site (StartupThreads) so I think about this a good deal. A few tips from working with companies:
1) Make it interesting - Tie your design into the ethos of what your customers care about. This can compelling beyond just your logo/brand (An Example by CloudMine: http://on.fb.me/MD9E0W and AngelHack http://on.fb.me/Mmn8CM)
2) Fit & Quality Matters - Spend the extra few dollars to get a decent American Apparel shirt, as it will fit better and stand out amongst the bargain bin shirts that most startups hand out (order womens sizes, guys). Increases the likelihood that they will wear it more than average shirt.
3) Make It Exclusive - There are varying strategies on this, but it is my observation that giving it out to those who really care about your company is the best way to go. You don't want the receiver to feel like you are bribing them to try their service/product. It works better if you are thanking them for their loyalty, as you help to solidify that relationship.
4) Reduce Your Colors - Most t-shirts that you are going to put out are screen printed and the cost varies based on the amount of colors used in the design. You can get great results with a one color shirt (see stripe: http://on.fb.me/MHG7Hp or foursquare: http://foursquarestore.com/shop/product/SH_MayorCrown). Even if you have a multicolor logo, there are likely ways to represent it with less colors and still communicate the same things.
5) Ask If The Printers Save Screens - If you are screen printing, have your printer save the screens they use, as it is a fixed cost that can save you money on subsequent runs
6) Vector Artwork (!!!) - Make sure your designer creates vector versions of your logo for print (bonus points if they do different versions for varying amounts of colors). You do not want your logo to look jagged when printed, so the best bet is to play it safe with a vector version of all artwork. (I spend most of my time dealing with this, actually)
Working on a full post on this, will get it out soon!
(1) 12 times is way too low for number of wears. I'm probably extreme, but I've got tech shirts that could apply for a drivers license; I've easily worn some of them hundreds of times -- let's say every other week for the last 500+ weeks. (The older shirts are too worn thin/ragged to wear in public any more though.)
(2) I'm not sure CTR is going to be as high as an explicit ad. Clicking a link is a fairly low barrier, and still only gets a 0.14% conversion. CPC is probably the wrong model to apply; you might want to look at it as traditional advertising and apply a CPM. For what it's worth, I'd bet that your conversion rate would be higher than average since the visitor has already qualified themselves somewhat more than just a casual clickthrough.
(3) For all of the times I've worn those tech shirts, I can only recall a couple of times when someone has actually talked to me about one of the shirts. Once (a nice polo shirt with embroidered logo) I was asked if I worked for the company (which was somewhat well known) -- the answer was no. The other time was when a fan of the by-then defunct company mentioned on the shirt remarked that it was too bad the company had imploded.
(4) I've gotten free t-shirts that required very little engagement on my part with the company. E.g. rackspace sent me a shirt several years ago I think just for answering a survey; I've never been a customer. Definitely consider giving shirts to people who can actually talk about you if someone does ask about your company. IOW, the CPM on a "talking billboard" is worth more than on a static billboard.
> (1) 12 times is way too low for number of wears. I'm probably extreme, but I've got tech shirts that could apply for a drivers license; I've easily worn some of them hundreds of times -- let's say every other week for the last 500+ weeks. (The older shirts are too worn thin/ragged to wear in public any more though.)
Depends greatly on the quality of the shirt. I have shirts that have held up like you describe, and I have shirts that became unwearable after a half-dozen times. If you want to provide shirts, don't use the lowest bidder.
Know what you mean. One of my favorites is the Netscape Lighthouse T. Only nit I have with what you say is that you should get over the 'too worn thin/ragged' thing. Far as I can tell, no one cares anymore. How often do you see jeans without holes anymore? Admittedly since I come from Idaho, there are a certain class of cowboy who wouldn't be caught dead in un-starched holey jeans, but every study needs outliers yes?
It's a really interesting post - I think one market segment you gloss over is college students.
Some products will obviously be better suited to others than this, but a great example I saw recently was Dropbox. Dropbox were sponsoring a CS project fair at my alma matter which I stopped by at. They handed out t-shirts (quite nice ones) to everyone there.
For dropbox, college students represent both their potential customers and their potential talent pool. It can be a bit of a bun-fight to recruit the most talented graduates, and the more you can do to get your brand across the better. And of course, students are going to wear those t-shirts across campus, getting hundreds of eyeballs every day.
Not to stereotype college students, but they if they're anything like I was they like free stuff, hate doing laundry (so an extra t-shirt is appreciated), and spend lectures frequently googling random crap (like the name on the person's t-shirt next to you).
(PS: If anyone from Dropbox is reading this, I am very sad that my less than a year-old Dropbox t-shirt developed a hole in it. Subtle hint!)
Aside from it not being ROI, I thought it was as good as any guess. What I was hoping for was a clever way of measuring some of the things you assumed.
I think what you calculated is called customer acquisition cost. ROI is how much you made from your investment in a given time, usually as a multiple of the investment -- sometimes it's quoted in time to breakeven.
Once I read this comment, I think it clicked. The result of what we calculated is customer acquisition cost. But, we were trying to figure out if T-shirts were worth the upfront cost. I am not a marketer, so I am 100% sure I am confused by the terminology.
To get ROI, you need to also know the value of a customer. If each customer pays 2 x Customer Acquisition cost in the first month, then the ROI is 2 in the first month. If it's a subscription, and 100% stay for a year, then it's 24 (or 2400%) in the first year.
Pro-ROI calculations take into account the time-value of money and taxes.
You have calculated Cost per Acquisition. ROI would be (Income / Investment - 1).
To get to ROI from CPA, just divide it into your Customer Lifetime Value. I.e. if you think a conversion is worth, on average, $40 over the life of their relationship with your company, then your return would be 40 / 19.05 -1 = 110% or roughly 2-to-1.
I think this is a cool calculation, and the numbers seem pretty reasonable. One benefit from tshirts that isn't taken into account here is simple exposure -- even if someone who sees the shirt doesn't follow up on the company themselves, let alone become a user, they are at least more likely to remember and recall the name. (Coming from MIT, it seems like Palantir and Dropbox had a lot of success getting good word-of-mouth campaigns going because of their stylish shirts).
Also, quality does matter. An extra couple of dollars per shirt can make a disproportionately large increase in the number of times the shirt is worn.
Quality is important. We thought about the "branding" aspect of the shirt and didn't have any idea on how to calculate it. Do you know of any resources to determine how to calculate brand value?
Given we both just posted within two minutes of each other about how great the Dropbox t-shirts are I think we can conclude it was quite the investment for them!
Why use Click-Through Rate to calculate return? Cost Per Impression would make a lot more sense to me, since it is closer to how traditional advertising is measured.
Excellent point. We aren't traditional marketers and some of the terms are still newish to us. I will revisit the terminology of the article and update it.
The formula you are using looks correct, but the value estimations are mostly wild-ass guesses, particularly the CTR. My own guess is that CTR for t-shirts would be much lower than traditional online ads, but who knows?
Suggestion: create a unique URL on the t-shirts so that you can test effectiveness-- a landing page for the t-shirt. That way you could even test alternative t-shirt designs, if you wanted to.
To me the CTR seems reasonable, if anything I would guess that 0.14% is a low estimate. The conversion rate on the other hand is sky high. Even for a free software product expecting a 7% conversion rate on fairly untargeted traffic is unreasonable.
1) Make it interesting - Tie your design into the ethos of what your customers care about. This can compelling beyond just your logo/brand (An Example by CloudMine: http://on.fb.me/MD9E0W and AngelHack http://on.fb.me/Mmn8CM)
2) Fit & Quality Matters - Spend the extra few dollars to get a decent American Apparel shirt, as it will fit better and stand out amongst the bargain bin shirts that most startups hand out (order womens sizes, guys). Increases the likelihood that they will wear it more than average shirt.
3) Make It Exclusive - There are varying strategies on this, but it is my observation that giving it out to those who really care about your company is the best way to go. You don't want the receiver to feel like you are bribing them to try their service/product. It works better if you are thanking them for their loyalty, as you help to solidify that relationship.
4) Reduce Your Colors - Most t-shirts that you are going to put out are screen printed and the cost varies based on the amount of colors used in the design. You can get great results with a one color shirt (see stripe: http://on.fb.me/MHG7Hp or foursquare: http://foursquarestore.com/shop/product/SH_MayorCrown). Even if you have a multicolor logo, there are likely ways to represent it with less colors and still communicate the same things.
5) Ask If The Printers Save Screens - If you are screen printing, have your printer save the screens they use, as it is a fixed cost that can save you money on subsequent runs
6) Vector Artwork (!!!) - Make sure your designer creates vector versions of your logo for print (bonus points if they do different versions for varying amounts of colors). You do not want your logo to look jagged when printed, so the best bet is to play it safe with a vector version of all artwork. (I spend most of my time dealing with this, actually)
Working on a full post on this, will get it out soon!