Chapter 5: How to Build a Personal Web Site -- the Broader Implications of Search Engines

by Richard Seltzer , seltzer@samizdat.com, www.samizdat.com

Copyright ©1997, 1998 Richard Seltzer



This is the fifth chapter of a book entitled The Social Web. Permission is granted to make and distribute complete verbatim electronic copies of this item for non-commercial purposes provided the copyright information and this permission notice are preserved on all copies. All other rights reserved. To correspond with the author, send email to seltzer@samizdat.com Comments welcome. You can buy this book on diskette from Amazon.com.

My Internet: a Personal View of Internet Business Opportunities by Richard Seltzer, on CD, includes four books, 162 articles, and 49 newsletter issues that will inspire you and provide the practical information you need to build your own personal Web site or Internet-based business, helping you to become a player in this new business environment.


So far we've talked about why personal Web pages are important, how to design them, and what kinds of content you might want to include. But how do you put the pieces together? When you get free Web space, you typically get between 2 and 10 Mbytes of space -- that's enough space for four to 20 copies of Huckleberry Finn. You aren't going to be creating just one page, but rather a whole set of pages, each of which does a necessary job, and all of which work together for your benefit.

If you surf around the Web, sampling Web sites to see how they are designed, you are likely to get a very distorted view. The name, commercial sites use high glitz, high expense techniques. You'll see frames (a stationary frame section, with graphics, ads, and links to other pages) and the main content inside the frame. You'll see pages created dynamically on the fly from databases. You'll see banner ads with revolving, flashing images. You'll see text scrolling across various parts of your screen like a ticker tape, or like an electric sign. When you do see text, there will be very little of it on any given page. If you went to a site with the intention of getting a particular piece of information or software, you will typically have to click a half dozen or more times before you get what you want. If you arrive at a site from a search engine, you are likely to arrive at a page that makes very little sense out of context. The graphics and the links of the pages to one another seem not designed to help you get what you want, not to serve you, but rather to dazzle and impress you, to keep you spinning for a while. Many of these sites get revenue from selling banner ads; so the longer you stay at the site, the more times you click to another page at their site, the more revenue they get.

On the one hand, you can't afford to compete with sites like that on their own terms. On the other hand, you shouldn't want to, because you can do much better, providing real benefit to your visitors -- at much less cost. 


Attack of the search engines

On Halloween back in 1995, I had the good fortune to see part of an excellent teleseminar "Competing in the Marketspace: Profits on the Internet and Beyond", produced by Kathleen Gilroy Associates (now known as The Otter Group).

The part I saw included a video case study of Virtual Vineyards (http://www.virtualvin.com/). Harvard Business School professors John Sviokla and Jeffrey Rayport (http://marketspace.hbs.harvard.edu) put the case into historical perspective and shed light on the extraordinary success of that Web site.

They spoke in terms of product, context, and infrastructure.

Consumers could buy these same premium wines at nearby stores. And the infrastructure (the Internet) was available to competitors as well. So the company's value to the customer comes primarily from the context -- the carefully constructed experience that they lead users through at their Web site, to help them understand, appreciate, select and purchase the wines they want. (Keep in mind that the presentation at the teleseminar was far richer and more compelling than my gleanings which I am summarizing here.)

While the presentation was compelling, I was left with an uneasy feeling that while what Virtual Vineyards is doing seems very right for now, and obviously works, it might not be a good business model for others to follow. At first, I couldn't articulate that uneasiness. Then, on the next day, I chanced on some interesting statistics about visits to my own little Web site and got some anecdotal feedback from a couple of readers on how they had navigated to my site.

Several months earlier, I had posted an old article I had written about Halloween (http://www.samizdat.com/hallow.html). And much to my surprise, over the couple weeks before Halloween that article got more visits than my home page. In fact, over the last few days before Halloween, it got three times as many visits as my home page. And many of the unexpected visitors were surfing using search engines and looking for references to Halloween.

What did that mean?

Search engines were making "home pages" less relevant. People were finding what they want deep in the bowels of Web sites, directly, without having to navigate the way the designer of the site intended. And what was true then is even more true today. Only about 9% of my visitors look at my home page; the rest go straight to the information that they want, probably connecting by search engines or by hyper-links that go straight to the relevant document.

In other words, search engines help people get in touch with what they want, the way they want it. Users create their own context, and blow right past the carefully constructed contexts that companies try to create for them.

AltaVista (http://www.altavista.digital.com), the grandaddy of the search engines, first went public in December 1995. It indexes every word on every Web page it finds -- and even keeps track of the order of the words, so you can search for phrases and sentences. With it, users can go from pages deep inside one site to pages deep inside another, without the help of hard-coded hyper-links. They can drill straight to what they want when they want it -- bypassing the carefully constructed contexts of sites like Virtual Vineyards.

A couple years ago it was reasonable to assume that people navigating to and through a Web would start with the homepage, just as physical visitors enter through your front door. So you would naturally want to do the cyber-equivalent of carefully placing your displays and the signs announcing special offers, arranging the merchandise, and providing shopping carts, and perhaps putting impulse items on racks near the checkout counters. The visual effects on your home page -- the front door area -- would be extremely important. The number of hits on deeper pages would simply be a subset of the hits on the home page (unless you went out of your way to publicize deeper layers as separate entities and encourage people to go there directly).

While we realized that users can come and go as they please, on a whim, we presumed that the structure/context of a Web site was important to hold the user there and to lead users through a controlled sequence of experiences and choices.

The pages at some sites were like the old Burma Shave ads on highways -- each page providing a small piece of the message that made little or no sense out of context and out of sequence. Other sites were more like branching adventure stories -- providing apiece of the story, then a couple of choices, then a little more information and a couple more choices. It was like walking through a maze of hallways, opening one door after another.

But now search engines have blown that model away. Everybody's coming in through the back door and the basement and the windows.

You cannot own/control the context. The encounter with the user will not be serial/sequential. In other words, search engines change the basic principles by which you'll want to design your Web site.

There is no point in putting the bulk of your effort into designing your home page when most of your visitors won't see it first and some of them will never see it at all. And there is no point in setting pages up so they need to be seen in sequence, if visitors will enter at any page, randomly out of your control.

Every page is a potential entry point for visitors. So every page should be able to stand on its own, clearly stating its context and purpose and providing helpful navigation buttons (hyperlinks that take the visitor to other key pages at your site.)

Keep in mind that there are two exceptions to the rule that you can't control the context.

First, if you use techniques that block search engines (e.g., databases, frames, dynamic pages, and pages behind sign-in/registration forms), the only page that visitors will be able to find by search engines will be your home page. But at the same time, that means you are sacrificing all the traffic that might have come your way -- at no cost to you -- if all your content had been indexed.

Second, the web crawlers (robot programs) that search engines like AltaVista send out to find the text of Web pages usually abide by the "Robot Exclusion Standard." This means you can create a file on your Web server and name it "robots.txt." In that file you can list particular files or directories that you do not want these web crawlers to see. If you want to do this, and your site is hosted on someone else's equipment, just tell the Webmaster what you would like to do. It is very simple. For instance, such a file might

read --

User-agent: *

Disallow: /test/doc2.html

This would tell all robot programs (the asterix * is a "wildcard," standing for all) to stay away from one file named doc2.html in my directory /test. Hence, for instance, you could use robot exclusion to set up a two-page sequence and regain some modicum of control over the context of the user's experience. Make sure the first page is indexed, and use an exclusion command to keep web crawlers away from the second. And have an attractive, tempting hyper-link from one to two. The visitors will be able to find the first page by way of search engines. But the only way they're likely to find the second page is by clicking from one to two. 


Degrees of Separation: a design goal

The movie "Six Degrees of Separation" is based on the premise that everyone in the world is only six people away from knowing everyone else in the world (by the chain of I know A who knows B who knows C...) There are some games on the Web which play on that same notion -- you start at any random Web page and try to arrive at a predetermined destination using as few clicks as possible.

Having often experienced the frustration of having to click again and again and again within a Web site to get to information that I need and that I know is there, it occurred to me that "degrees of separation" should be a design goal/principle for both Web sites and Web software. Far too often, the user has to go from one menu of links to another to another to another, clicking again and again, each time having to wait for useless decorative graphics to load.

Sometimes Webmasters set up their sites this way out of a mistaken notion that small pages make for ease of use. Hence they deliberately limit the size of their pages to two or three screens full of information at most.

Basically, the fewer the clicks the better. My Web site design goal is to provide the maximum content for the minimum number of clicks.

At my own little Web site, http://www.samizdat.com, which now has over 800 Web pages, some of which consist of entire books, I aim for two clicks as a limit. In other words, no matter where you enter, you should be no more than two clicks away from any other page at the site. (It's not perfect - there are some cases where it would take three; but two is the goal).

Clicking is not a pleasurable activity; it's a means to an end. And navigating through a Web site shouldn't be like finding your way through a maze. Keep it simple.

Design is a mixture of taste and utility. What makes Web page design particularly difficult is that the utility of different designs depends on your computing platform. And while many Webmasters use UNIX workstations, an increasingly large majority of the potential audience for Web pages uses PCs.

The notion of limiting Web pages to three screens makes good sense from a UNIX workstation point of view. But from my PC, navigation within a page is a breeze. I rarely use the scroll bar. Instead, I use CONTROL-end and CONTROL-home to go to the bottom and then to the top again; or Page-up and Page-down; or I use internal links within the page and the Find function in my browser. Hence, as a user, I delight in large, content-rich pages. And the more relevant links the better.

I'm a PC-person; and at my little site, the overwhelming majority of visitors are PC-people. And they keep cheering me on to make my already monstrous home page even larger and more content-rich. And AltaVista and other search engines help immensely by indexing it all for me, and making it easy for people to find what's at my site without the need for tree-like directories.

As you design your Web site, focus on utility first and process second. In other words, do whatever is necessary to make sure Web pages look and work the way they ought, rather than simply accepting the way today's automated tools make them look. 


Modular distributed Web sites -- a kludge for personal and small business Web sites that offers richness and variety at no cost

A year or two ago, if you wanted to experiment with your own Web pages on the public Internet, you had to have an account with an ISP. Yes, America Online and lots of local ISPs offered free Web space as a come-on for dial-in customers. But if you didn't already have an account, that was an enormous

psychological barrier. You could get/see whatever you needed from the Web using your account at work or at school. Why should you spend money to have an ISP account in addition?

Now there are Internet Presence Providers -- Web-hosting sites -- that don't offer dial-up service and that offer free Web space to anyone and everyone. The three largest -- Tripod http://www.tripod.com, Xoom http://www.xoom.com, and Geocities http://www.geocities.com -- each already have about 2 million members (and they're still growing fast). As of today, Tripod offers 11 Megabytes of Web space with a free membership; and for $3 a month, you can get 22 Megabytes and your own chat room. (The stats and the offers change often. Check the sites for the latest.) These companies, and others like them, have on-line tools to make it easy for you to build your pages. And since they cater to newcomers, their instructions and help files are clear and simple. You could have your first page up on the Web in less than half an hour.

Unintentionally, these services open a new and intriguing opportunity. There is no reason for you to limit your activity to a single such free site. Rather you should explore the possibility of putting together a "distributed" site -- using free space wherever it is available (preferably for specific, clearly identified pieces of your overall effort), and linking them together with hyperlinks and a common look-and-feel. You could run a regularly scheduled chat program at Xoom, moderate a forum at Delphi, and have your heavy graphics at Tripod, etc. In other words, at no cost or on a shoestring you could run a rich and varied modular Web site

When I started my personal Web site, with a local ISP (tiac.net), my 10 Mbytes of free space seemed enormous. I was mainly interested in posting text, rather than pictures, and for plain text, that was the equivalent of about 6000 book pages. But once I started getting feedback from visitors, my imagination got clicking, and I had far too many little projects that I wanted to do than that space would allow. And as soon as traffic started to build, I find myself banging up against obscure limitations that I hadn't even considered when I started -- how much material (in megabytes) visitors accessed at my site per month. If I went over the limit, I would quickly go from zero cost to very substantial charges. I couldn't control the activities of visitors and really didn't want to make it difficult for people to get to my pages. (The whole point of the site was to build an audience and learn from them.) Fortunately, another local ISP (acunet.net) offered me substantially more space, at no cost, because they considered what I was doing valuable. All they asked was that I provide an acknowledgement on my pages, with a link to their site. If that opportunity had not come along when it did, I would have systematically created my own "distributed" Web site, putting discrete chunks of content at different Web-hosting sites.

As it is, much of my Web activity already is "distributed." Each week I conduct a chat session about Business on the World Wide Web at web-net.org, the site of a non-profit organization of Internet entrepreneurs and designers. The live chat takes place using their custom software, running on their system. And I save and edit the transcripts, posting them at my site (http://www.samizdat.com/#chat) -- with links going both ways. In addition, whenever an article of mine appears in an on-line publication, I provide a link from my site to the publication's site, and seek a link from there to mine (or at least inclusion of my email address).

For a couple of my chat sessions, I had visitors connect to two different sites at the same time. One of these was for a demo of new and very promising software that combines the best of both chat and forum (AltaVista Forum). The other was for a demo of a MOOs (another kind of software that is particularly good for regular interactions with a fixed set of people, like a class; we'll talk more about that in Chapter 9, about distance education). In both cases, the experimental software we were checking out was on someone else's system, but while looking at it and using it in a second window, we could also chat about it with developers and experts in a normal chat room. And, of course, afterwards, I captured the entire dialogue in the edited transcripts which I posted at my site.

Chat and forum are two very important "community" activities that you might very well want to add to your site. Fortunately, Web-hosting companies today are expanding in the direction of not only providing space for static Web pages, but also providing chat rooms and forum bulletin-boards to all comers at no cost or very little cost. We'll discuss the importance of chat and forum, and provide advice on using them successfully in Chapter 8.

What else might you want to use additional free space for? First consider graphics, since those tend to eat up lots of space. You could put your most graphics-intensive pages on a separate free site. Or if you have lots and lots of pictures, organize them and put different kinds of pictures at different sites -- always maintaining the same overall look-and-feel for the design of your pages, and always providing hyperlinks to all the other parts of your site from each and every page.

As an alternative, you could put the text for all your pages at your home site, and link from there to your distributed sites for all the photos, graphics, etc. That approach not only expands your total disk space, but also distributes the traffic. For instance, say you included a public domain photo of a planet from a NASA site on your home page. You might store a copy of that picture at a distributed site of yours, so whenever someone accessed your home page the only traffic that would show up in the logs of your home site would be the text, with the photo showing up as traffic at your remote site.

Second, you if you can get free space in a different country or, preferably, on a different continent), you might consider mirroring your own pages -- creating a complete copy of your home site at a remote location to make it easier for distant visitors to get to your pages and also to distribute your traffic (for faster access during busy times and also to avoid traffic charges).

You also could use distributed Web space for experimentation -- trying out new kinds of content and new kinds of activities and learning without disrupting your home site. If the new stuff works well, you might later bring it in and tie it more closely with your main content.

As ISPs and Web-hosting companies continue to expand their free and low-cost offerings, this approach is likely to become very popular. In all probability, you should soon be able to have an on-line store hosted at one site, and audio and video files hosted at another, and link to visitors/customers through an Internet telephony application at still another.

When I speak about the Internet, people often ask me for advice in setting up Web sites for towns and public service organizations. They have heard of towns spending millions of dollars a year on their Web sites. They are tempted by the potential benefits, but simply don't have that kind of money. I suggest that they sign up with Tripod or Geocities or Xoom, and that they put most of their effort into text rather than space-consuming graphics. If they need more space than what any single Web-hosting site offers for free, then I suggest that they line up volunteers -- split the town up

into areas of interest like schools, churches, town government services, etc. The bigger the town, the more areas and the more volunteers it will need. Each volunteer signs up for a free Web space account. They all agree on a common look and feel. They all hyperlink to one another. And the town gets a substantial modular distributed Web site for free. (Keep in mind that the technology is the easy part here. The tough part is working with volunteers -- keeping everyone motivated and focused.)

Whether the distributed site is your own or one your are building for a town or non-profit organization, always remember to "Add Page" or "ADD URL" at AltaVista and other search sites for each and every page you create. And provide clear descriptions of the overall context and handy navigation links and use a common look-and-feel on all your pages, so your multi-location Web site holds together.


Chapter 6

The rest of The Social Web by Richard Seltzer

My Internet: a Personal View of Internet Business Opportunities by Richard Seltzer, on CD, includes four books, 162 articles, and 49 newsletter issues that will inspire you and provide the practical information you need to build your own personal Web site or Internet-based business, helping you to become a player in this new business environment.

Web Business Boot Camp: Hands-on Internet lessons for manager, entrepreneurs, and professionals by Richard Seltzer (Wiley, 2002). No-nonsense guide targets activities that anyone can perform to achieve online business success. Reviews.

Can we help you build an Internet business? Richard Seltzer is an independent Internet writer/speaker/consultant. Click here for details. or send email to seltzer@samizdat.com

This site is Published by B&R Samizdat Express, 33 Gould St., West Roxbury, MA 02132. (617) 469-2269. seltzer@samizdat.com


Return to B&R Samizdat Express

Check our sitemap page www.samizdat.com/sitemap.html from which you can get to any other page at this site in one click.


<
Internet Business Showcase: