On Web Advertising

Avatar of Chris Coyier
Chris Coyier on (Updated on )

DigitalOcean provides cloud products for every stage of your journey. Get started with $200 in free credit!

On this week’s screencast I talk about online advertising. What it is, how it works, and mostly, my opinions about it. I thought I’d recap here and touch on some things I forgot about.

Content Websites vs. Product/Service Websites

There are some websites that spend their time building content. They build it to keep visitors coming back because it’s funny or interesting or important or otherwise compelling. Content websites, by and large, are free to visit and consume. Creators of content website spend considerable time and effort creating and maintaining the site and its content.

Some websites exist to support a product or service that is for sale. The creators of these sites spend the majority of their time and effort working on that product or service. Making it better, improving features, etc. They aren’t working on building content, that’s not their forte.

Content websites get the traffic (but need money). Product/service websites make the money (but need traffic). Advertising is a way to trade and share some love back and forth.

Not Evil

There is nothing “evil” about advertising. Creating content takes blood, sweat, and tears. Creating products takes blood, sweat, and tears. Both deserve to be paid for that. Advertising makes sure they do.

Advertising can certainly be overdone. though. Having to watch a 30-second video before reading an article… I feel that’s too much. A page-peel that you come within 100 pixels of and it comes down and covers half the page? Too much.

Except when it is

The “evil” stuff only starts entering the picture when websites start do things under the table. Selling emails/spam, that’s clearly evil. Writing an article about how great a book is, without ever reading it, so you can link to it with an Amazon affiliate code, is evil. Filling a page with bullshit keyword-strewn content in hopes to get some search traffic and covering the page in ads, is evil.

Honest websites creating honest content displaying ads for honest companies creating honest products, that’s always OK. When any one of those things loses the “honest” part, that’s problematic.


On this site, I use BuySellAds. They specialize in the design website community, which means that they have a base of advertisers with products and services that cater to this community. That’s awesome, as it means that ads on this site will be filled with things that web designers might be interested in. They take 25% of sales, a hefty fee at a glance, but completely worth it when I consider that I spend almost no time at all thinking about advertising. I can use that time building content instead. Otherwise, I’d be spending my time finding/communicating with advertisers, invoicing them, keeping track of live dates, flipping in and out graphics, etc. And because BSA comes in over JavaScript, it keeps paid links away from Google bots which frown upon that. Also great about the BSA model, advertisers pay for what they get. They look at ad spots, they see what kind of impressions the site has, they see a fixed monthly price, and they make the call. It’s as straight-forward as it could be.

A much more popular advertising service on a global scale is Google AdSense. AdSense reads the content on the page it resides upon and serves up advertising relevant to that content. In general, I think that’s great. Relevant ads benefit publishers (more people click them), advertisers (the people clicking are a targeted audience), and the users themselves (they find something they are actually looking for). On my site, I don’t need it, since I already have a service that is targetted. But it’s great for something like Ning.com, where people can build their own social network for free. Someone starts a knitting site, content on the site is about knitting, there are Google ads on the site about knitting, everybody wins. Forums are also good targets for AdSense, since topics are set by your visitors themselves, and the ads can mimic those topics. Personally, I don’t like Google ads. I think they are ugly. I think people know exactly what they look like and know how to avoid them. I think it attracts lots of crappy advertisers. Most of all, I don’t like how it pays via clicks.

There is one place on CSS-Tricks that uses them, and that is next to the search results. I love Google Custom Search Engines. It does a great job in searching this site. I tied my AdSense account to the CSE so that I earn from the clicks on the ads in search. It amounts to a few dollars a month. The reason I do it is because it costs $100 to remove them and I don’t really think it’s worth it. They sit harmlessly over to the right which would be blank white space anyway.

The cool kid club

There are a couple of ad serving services in the web design niche that are exclusively for the cool kids: The Deck and Fusion Ads. They are pretty great. Publishers using these services display only one ad per page. The ads are typically very well designed and classy. I like this idea very much. Unfortunately, you have to be a lot cooler than you or me to get in on it. I also tend to think that sites serving these ads aren’t primarily concerned with advertising income. While I’m sure they pay fairly well, I’m betting one high paying ad can’t beat 9 medium paying ads.


Affiliate programs are another form of online advertising. Product/service sites tend to love them, as they only have to pay out when successful sales are made. In traditional advertising, they have to pay no matter what and if the ad does poorly, it’s a loss for them. If an affiliate ad does poorly, it’s a loss for the publisher.

I like the idea of affiliate programs. If someone goes through the trouble of spreading the word and sending in sales, they deserve to be paid for that work.

I’m also wary of them. I think the advertiser gets a pretty significant advantage. Advertising isn’t just direct sales, it’s brand exposure and mindshare as well. The advertiser gets this part for free, as well as only paying for direct sales.

I think if you become an affiliate for anything, you should do it because you like what you are selling and you think you’ll sell enough of it to turn the balance in your favor.

I have affiliate programs both for Are My Sites Up and Digging Into WordPress. I like how they have worked out.

Common sizes

I have no idea what these sizes are based on or what the history is, but the Interactive Advertising Bureau offers up this list (based on what is “commonly sold” in the marketplace). They say that the goal is to provide some standardization to reduce stuff like having different publishers using sizes only trivially different like 300×95, 300×100, 300×105 etc., which is a damn fine goal.

Recommend Max File Size Recommended Animation Length
300 x 250 Medium Rectangle 40k :15
250 x 250 Square Pop-Up 40k :15
240 x 400 Vertical Rectangle 40k :15
336 x 280 Large Rectangle 40k :15
180 x 150 Rectangle 40k :15
300×100 3:1 Rectangle 40k :15
720×300 Pop-Under 40k :15
468 x 60 Full Banner 40k :15
234 x 60 Half Banner 30k :15
88 x 31 Micro Bar 10k :15
120 x 90 Button 1 20k :15
120 x 60 Button 2 20k :15
120 x 240 Vertical Banner 30k :15
125 x 125 Square Button 30k :15
728 x 90 Leaderboard 40k :15
160 x 600 Wide Skyscraper 40k :15
120 x 600 Skyscraper 40k :15
300 x 600 Half Page Ad 40k :15

Personally, I think the sizes are rather asinine – with one ad size having nearly no relationship to the next. Standards are great, and of course, it will be hard to fight against the stream now, but I think this whole “common sizes” business needs a re-think.


If you offer a number of similar sized ad blocks in one area, the fair thing to do for the advertisers is to randomize their position within the block. I’ll post a snippet of how to do this soon.

A/B Testing

You may not have the ability to always do this if you are using an ad service, but a good idea in web advertising is to use A/B testing. That is, have two versions of the same ad and measure which one does better, everything else being equal. Randomly display each one, and add different tracking information to the end of the URL that each links to. If you are using Google Analytics, you could use their link builder.

When to start

A classic problem of online advertising is deciding how to approach it on a brand new site. There are some pretty different schools of thought.

Some people say you should wait to put ads on a site until it has grown up enough that it’s “worth it”. I can buy into that. You are going to make pennies when launching a brand new site with no audience. Your ads aren’t worth anything, so why put them there? It also preoccupies you with thinking about advertising when clearly the most important thing for a fledgling site is growing that audience and traffic.

The flip side is putting ads on a site from day one. Even though they aren’t worth anything, at least you are ready, in the design/layout of the site for when they are worth something. It also sets up user expectations on the site. If you build an audience on a site with zero ads, then one day fill the site with ads, that’s rocking the boat quite a bit. Users might not take kindly to that. Having ads from the beginning sets expectations straight from the beginning.

For the sake of offering advice, I’d say if you think your site will have advertising on it eventually, plan for it design-wise and throw some placeholder ads up for now (perhaps display ads for your friends sites?). Then once you’ve grown enough (perhaps 50k pageviews a month) then start working on paid ads.

Feel free to chime in with your own thoughts about online advertising!