The revolution in digital photography is making an unprecedented number of spectacular, attention getting, photos available to commercial Website designers, who frequently can't use them because of the Internet's current bandwidth limitations. In an ideal world, every consumer will have their own dedicated 1.5 megabyte per second connection, and a screen full of 24 bit graphics will load in a split second (ideally, a timeframe far shorter than a customer could decide to go to another Website in.) Unfortunately for Internet advertisers, much of the home and smaller business market isn't going to be connecting to the net at speeds above 28.8K for two or three years, at best. At the same time, the most affluent members of the target markets will be using high speed Internet connections, capable of delivering high quality graphics, audio, even video.
It's annoying enough to have to consider how Web pages will look on two fundamentally different kinds of browsers, at a variety of screen resolutions, but it appears the additional issue of connection speed is about to become a required consideration. Websites will have to be designed for users connecting at both high, and low, connection speeds. Short of having a separate Website for each speed, advertisers will need to think about how to reach an effective compromise of the two; a Website that meets the minimum level of appearance for high speed connections, and is still fast enough to work on a low speed connection (28.8Kps.)
The real problem, of course, will be to optimize load performance for low speed connections. Much of the current focus on Website optimization concerns issues like load balancing and load performance, which are valid considerations, but only part of the whole picture, and taken by themselves, possibly misleading about the site's effectiveness.
The most important performance issue for any commercial Website is how long it takes for its home page to load. If it takes too long, the customer will gone, and it doesn't matter how any other part of the Website performs. Given how much money some companies are spending on developing Websites, it's rather surprising exactly how little time, thought, or money, if any, companies are willing to spend to optimize this aspect of their Website's performance.
Some of the techniques to optimize home page load time are simple, and actually don't even cost any money. For instance, a big improvement in the perceived load time of a Web page can be made by carefully determining the order that the graphics and text will load. This is actually fairly easy to do, and a lot of major Websites don't seem to do it at all. Browsers usually load text and graphics in alphabetical order, so by varying the name of each graphic you can make sure that graphics at the top of a page load before graphics at the out-of-site bottom of the page. Done right, this can allow a page with a lot of images to "seem" to load considerably faster than it actually does. If there's much text on the page to read, the user usually won't scroll down the page fast enough to notice the lower page graphics may still be loading.
Other ways to improve Web page load time are available as well. For esoteric reasons, a large graphic image loads faster if it's divided up into parts and sent as a series of smaller images, (or tiles, as they're called) than if its sent as one big image, even though the total file size is the same. For instance, a 50K image file sent as ten 5K "tile" images would load somewhat faster than if it were sent as one 50K image. Generally, it's a good idea to avoid using graphics big enough to need this process, but if you have to, this technique can help. There are commercial and shareware applications that will create tile images from most common file format images.
Advanced file compression is another technique that can improve page load time, and it should be seriously considered, even though the software needed to do it costs one or two hundred dollars. In general, this software works by highly compressing the file size of graphic image, without the usual resulting trade-off in image quality. Some of these programs are smart enough to allow parts of the image to be compressed less than others, so for instance the part of an image with the advertiser's product could lose less image quality than the surrounding background.
Ultimately, just having the tools & techniques available isn't enough, it's also going to take a plan to use them. One of the big challenges facing Internet advertisers now is coming up with a formal and effective process for optimizing their Website, including determining what areas will be covered, and who will be responsible for them. Given that most Websites are a team development (or worse, a committee project), this is not an easy task. For instance, is it the responsibility of the artist to maximize image compression, or is it the responsibility of the Webmaster? Who pays for the extra software needed to compress the images, the art department, or the Web development department? And who decides what trade-offs will be made between load time and content. Is this a decision for the Web designer, or the marketing department? These are essential questions, yet to be addressed by many major Internet advertisers. Until they are, a number of commercial photographers had better hope their photographs have a long shelf life. On the Internet, size does matter, and until that problem is resolved the market for images on the Web will remain limited to just the photos that will fit.