Tom Chi  

*NO* Money Mo’ Problems

November 7th, 2003 by Tom Chi :: see related comic

Hey everybody. In our first full calendar month we used up well over 6 GB of bandwidth (lesson: color comics are big files), so KC and I spent part of last week looking at strategies to break even. Now that KC is a grad student again he’s back to eating Ramen, and for reasons unknown I’m still eating it too.

Anyways, the investigations were certainly interesting, and they got me thinking about usability in the context of advertising. Perhaps this is already a funny statement, since usability and advertising seem to be having a protracted war on the web. For the last couple years, each new type of ad somehow manages to break more usability rules than we even knew existed. But I think that usability has played a big role in why ads are still the dominant model.

On the web today we have a lot of options. We can go with an all-out offensive: paving the entire surface of our page with jarring Flash ads. We can test a user’s patience with 15 second interstitials which try to feed us a full-video commercials over our dial-up connection. We can pop over. We can pop under. Or we can simply deluge the page with blinking banners. From the littlest pages to the biggest, all of these approaches are exceedingly popular on the web today.

As HCI folks, we gotta wonder why people put up with this. Well let’s look at some alternatives…1. Micropayments. Scott McCloud is so awesome, and I have to admit he puts forth a great case for how the economics of micropayments might mean lower prices for consumers and better compensation for creative types. I also have to admit that I did not pay the 25 cents required to see his new comic, even though the teaser looked awesome. What happened? Well I had no BitPass account, and I just couldn’t muster the energy to go through all the registration steps and initial account setup. Right at the point of sale, the system wanted me to spend 5 minutes filling out forms which is a pretty good definition for ‘unusable’.

Perhaps if the onus was moved somewhere else… for example, if my ISP charged me an extra $5 a month so that I would have $5 of micropayment moola which automatically got deducted when I checked out micropayment-based content. There might be a little ticker running on a client daemon telling me how much I had left for the month, but other than that the mechanics of the system would stay out of my way.

2. Paid Subscriptions. This one is more popular than micropayments and also a bit more usable. You read the public content of a site for a while and eventually you decide to sign up to get to the exclusive content (and/or avoid ads). Salon, CNN, and IGN among others take this approach. It is more usable because once you sign up, you never need to think about paying for that content again (until your subscription runs out). With micropayments you pay bit by bit, so the cost of each bit of content being viewed must be considered. This might sound good because you get to pick what you pay for, but I’d argue its too much mental load to be thinking about prices all the time.

For the content creator, the downside of paid subscriptions is that you now need to spend extra time either creating exclusive content, or spend extra time delineating what is exclusive and what is public. Both are arduous, and both mean that you are limiting the size of your audience for exclusive content (which you probably worked really hard on, and which you’d really like to share). Lastly, subscription based sites are vunerable to competition from other sites with similar charters who opt to show all their content (behind a veil of ads, of course). Users will ask: why be a member at IGN if I can get all the same “exclusive” things for ‘free’ at 3 other sites?

3. Regular Old Web Ads. This is both the most common, and the most usable (!?) of the solutions which are currently popular. When you go to an ad-funded site, there is no hidden exclusive content. Everything is there. You don’t need to think about pulling out your wallet, or evaluating the cost of individual pieces of content. You can just browse and read whatever you want. If the ads are too annoying you can just leave and not feel as if you are wasting subscription money.

The dark side of this model is that as a content owner your site may not make any money if the ads do not attract attention. This is where all the bad things start. This is why we have mutating, blinking pop-ups from heck multiplying on our screens. Yet given the choice between providing my patience and providing my credit card number, it’s clear which one I will pick.

* * * * *
So back from the abstract to the practical. We’ve tried hard to pick a system that will be unobtrusive and deploy it in a way that will (hopefully) minimize the usability problems discussed above — afterall, we’re an HCI site.

Tell us what you think. Oh also, before I finish, I should point out some other models:

4. Donation / Wishlist model
5. iTunes model
6. afilliate programs (amazon book referrals, etc)

I’d write about them, but it’s better to defer that to the community discussion.

18 Responses to “*NO* Money Mo’ Problems”
Bob Salmon wrote:

Clay Shirky has an interesting and fairly meaty article about micro payments (summary: they don’t work, because even though the financial cost is minimal, the mental cost (decision making) is still high): http://shirky.com/writings/fame_vs_fortune.html
TC alludes to this in saying that his ISP could give him a $5 bucket of pre-paid micro-payments.

Sorry TC: whenever I type TC I’m reminded of Top Cat, the excellent cartoon from my youth (http://www.dfcom.freeserve.co.uk/hbw/topcat/ )

It’s also worth checking out his article on on-line communities and how they can self destruct: http://shirky.com/writings/group_enemy.html

Paul Reinheimer wrote:

The concept of having an ISP driven micropayment system sounds convient, but unfortunatly I think it is a pipe dream.

Lets say, NiceSmallCorp starts up this micropayment system and somehow (likely by giving free credit away) gets a few ISPs to join in and give their clients accounts. Eventually NiceSmallCorp will become BigEvilCorp (They simply cant get to a critical mass of ISPs and websites without becoming big, becoming big will require more money than they can skim off a startup micropayment scheme, so it will happen, either by a buy out, or AOL/MSN will build thier own system and use market dominance to destroy NiceSmallCorp).

At which point BigEvilCorp has market dominance, and a monopoly, rates will increase from a penny a transaction to 3%, then 3% and a $0.05 service charge for each transaction, etc etc.

Take a look at PayPal & Amazon gift funded sites these places are losing disgusting amounts of money in service charges.

Affero has been talking about a micropayment scheme, but they are not there yet. It does seem like a nice community support type system.

Paul Reinheimer wrote:

One other small comment.

While I love the whole donation/wishlist type system, I think it is underused.

I am active in several IRC channels, both as a help-er and a help-ee. Often people providing assistance have amazon wishlists, or affero accounts ( http://www.affero.com ). However, even when people spend hours helping someone, the thanks is breif and non-commital.

Someone helped me, so I hit up their amazon wishlist and droped $15, I then asked, how often does this happen. I was the first, in three years.

I value information & assistance. I perceive it tp have a high intrinsic and monetary value, the Internet mentality however seems to be everything you want now and free of charge. More people will have to convert to my mentality before PPV sites and micropayments go main stream.

Mike Hillyer wrote:

Well, what I can say is that I tried a PayPal donation button, it netted me nothing. I tried being an Amazon Associate, it netted me 85 cents that I will never see. I tried Google Adsense, it has netter me enough money to support keeping my site online and buying my wife birthday presents. So only ads have worked. I like Adsense because it is non-animated, does not take over the screen, and presents relevant advertising every time (and the ads are even in Spanish and German on my translated pages.

Tom Chi wrote:

Yes, we have hopes for the ad-based approach, though we are open to other models if our readers dislike this one.

After reading over the ad text, I began to think that they may even contribute to the discussion. Afterall, if we write an article about focus groups and then links to focus group moderators appear on the right, readers will have a quick way to compare our opinions to the actual industry practice. Maybe.

Joshua Kaufman wrote:

Continuing on my CSS parade from KC’s thread, you can also save bandwidth (and therefore money) by using CSS because you don’t need to include tables for presentation. The CSS file is (typically) downloaded once and cached. After that point, only pure HTML is downloaded.

Ivan V. wrote:

What about spreading the load between multiple voluteers? The simplest way would be to use a server side script which randomly selects a server and uses the image URL from that server. I can currently offer about 5 GB/month of bandwidth, but that might decrease in the future. However I can commit to at least 1 GB/month. Let me know if you’re interested.

PS: I don’t want anything in return (no strings attached, really), although a link back would be welcome.

Michael Buckbee wrote:

Something else to try might be to further compress the raw image size of the comics themselves. OKCancel20031107.gif is about 120Kb. I ran it through Photoshop and got a reasonable (to my admittedly non-graphically art trained eye) version out at around 64Kb.

You could probably cut that 6GB down to 4GB with just that one change.

Tom Chi wrote:

So I did a quick analysis of some of the possibilities. You can check that out here http://www.tomchi.com/sizeanalysis.gif .

Turns out that Michael is right. With a good JPG encoding we could cut down total size by around 20% (though it would require infrastructure tweaking). A good GIF encoding would bring it down 10% (no tweaking). By comparison, the total amount of table cruft on our page was only 1.2%. To Joshua’s credit, on a page that wasn’t so dominated by graphics, table optimization might play a significant role. Then again, a page that was primarily text probably wouldn’t run high bandwidth in the first place.

Ivan’s suggestion is an interesting one. We might need to take you up on that bandwidth sharing offer, since we have some truly large projects in the works. Whether we’ll go with an alternate host approach, or with bittorrent, etc, remains to be seen. Thanks to everyone for the suggestions so far.

Tom Chi wrote:

What a coincidence. A Slashdot article just got posted about micropayments and BitPass:

http://slashdot.org/articles/03/11/09/2051240.shtml?tid=126&tid=187&tid=95

Paul Reinheimer wrote:

I just read the slashdot article too. Bitpass looks okay.

What about a regular version with a compressed or smaller comic, and the google tower. And a bitpass version with a bigger nicer image and no ad tower?

Bob Salmon wrote:

You could always take the approach of BBSpot: http://www.bbspot.com/index.html - take the personal ads and then make fun of them!

Tom Chi wrote:

Yes, I love BBSpot, and we did have an ad like that in the works. So much for surprises. As for BitPass, from the slashdot thread it looks like its not quite working for content providers yet.

One comment mentions that BitPass charges 15% transaction fee and does not mail any checks until the value of the account is sufficient. That same poster notes that this is worse (and has performed more poorly) than PayPal.

As for splitting the comic into “low-quality comic with ads”, or “bitpass without ads”… this might be hard to pull off. I know that I’d keep looking at the comic for free at lower quality, unless a) the ads were so obnoxious that I had to escape, or b) the quality was so bad that I had to escape. In addition to a) and b), the quality of the work/writing would need to be so good that I would choose to pay over just not coming back to the site - (c).

As an HCI person, I find option a) terrible, and as a comics artist I find option b) terrible. On top of all that I’m not so sure (c) is true. We are reasonably funny, and our essays and reasonbly interesting but as an outsider I don’t know if they would be compelling enough to pay for (course I’m a little stingy, but whatever).

I guess my underlying point is that the web has always been powerful as a broadcast media. At the core of it, it was the idea that a dude in his garage could tell the entire world about his cats. No, but seriously, most of these models for monetization run counter to the spirit of what made the web so great in the first place.

They also create a significant barrier to entry for creative types who are just trying to establish themselves. No one wants your content until they’ve heard of you, but no one’s heard of you until your content gets out. That’s why a site like http://www.switchsessions.org (as well as most band sites) are free. Content creators need exposure as much as they need money, and a lot of the methods of monetizing the web have just thrown up walls which isolate them from exposure.

Tim wrote:

Another option for reducing bandwidth usage is to use PNG instead of GIF. See the “PNG tips for Cartoonists” article:
http://hackles.org/etc/png.html
Although JPEG will reduce the file size, it tends to make text and line art blurry as a result of the compression technique.

Tom Chi wrote:

Tim, you’re definitely right about JPG potentially causing ugly text and lines. I checked out the link and it recommends PNG for cartoonist because they assume simple cell shading techniques. Unfortunately, (or fortunately, depending on how you look at it) OK/Cancel features subtle shadows and soothing gradations of color. Aaahh. In conclusion, KC and I are probably going to go with the slightly smaller GIFs.

Anyway, while these file size comments are great, we should get back to the main topic:

What are good revenue models on the web? Which work for the content provider? Which work for the people? Which ones are just unbelievably lame?

Paul Reinheimer wrote:

I’ve been reading penny-arcade lately, and it would seem that they have a club, you dontate X dollars to join, then they mail out free stuff every once in a while to members.

Same sort of thing could apply here, stickers, Laminated HCI reference sheets, plastic spoons, etc.

Membership is driven by the want to reward the site, the free stuff is just a nice thanks.

KC wrote:

Both thegoodtomchi and I read PA every so often (regulars there may note the general ratio of funny comics to completely uninspired has taken a decline). The interesting thing about Club PA is that they actually just STOPPED doing it because they’re making enough money off their ads, merchandising and commission artwork for game magazines.

Tom Chi wrote:

I’m sure we could set up a place for people to just drop off money they don’t need. It’s not a business model per se, but about as good as most dotcom models a couple years back. Just wait until our superbowl ad!


Leave a Reply


OK/Cancel is a comic strip collaboration co-written and co-illustrated by Kevin Cheng and Tom Chi. Our subject matter focuses on interfaces, good and bad and the people behind the industry of building interfaces - usability specialists, interaction designers, human-computer interaction (HCI) experts, industrial designers, etc. (Who Links Here) ?