Richard Seltzer's home page  Publishing home
 

Business Web Site Design by Richard Seltzer

Redesigning A Company Web Site

Every day my modem seems to get slower. No, it's not broken. Rather, more and more sites -- even personal ones designed by high school students -- use the latest and greatest fancy graphics with frames and images that flash and move repeatedly, and Java applets that make new things happen non-stop. There seems to be some law of human nature on the Internet that everyone needs to push the limits of the technology, using all the graphics and multimedia effects that they possibly can, to prove to themselves and to the world that they can do it.

What are they communicating? What's the content? And in many cases, there is no content.

Even large, high-tech-savvy companies get caught in this same human-nature trap. While the business objective of their Web sites should be to provide useful content to potential customers, they put most of their effort and investment into fancy effects, that add nothing in terms of content and communication, and in fact make their sites far more difficult to access for ordinary people, with ordinary equipment and software.

The point isn't that these effects can't be useful and can't enhance communication, but rather that, in most cases, they aren't used that way today. It takes skill, experience, and judgment to create or even to choose a picture that is actually worth a thousand words. And it takes even more experience to effectively use the more advanced capabilities. But rather than proceeding step by step and learning how to get the most out of each level of technological capability before advancing to the next, people infected with the technology bug rush ahead and dabble with the new stuff before mastering the basics of plain-text Web-based communication. And others who have something to say but aren't familiar with the technology or can't afford the latest hardware and software are intimidated, and back off -- convinced that Web-based communication is beyond their capabilities.

Another rule of thumb -- make your pages whatever size makes sense for the convenience of your audience and the nature of your content. Many companies follow arbitrary design rules in making their Web pages, and keep their pages very small -- only one or two screens of material per page. In some cases, that approach means that you have to click again to get to the information you want -- one page is a list of contents, then the text you want to read is broken up into a series of tiny pages. To get something that's about the size of an article or a brochure, you have to click about ten times, and every time you have to wait for the graphics to load. And if you want to print it, you've got to print each separate page, rather than the whole thing as a single entity. That approach is also very annoying to people who use search engines to find your pages. They'll arrive somewhere in the middle of the article, at whatever Web page happened to include the word or words they were looking for, and then will have to back-track to get to the beginning.

I prefer for a Web page to in fact be a "document" -- a complete document, with a beginning, middle, and end. The content, rather than a design rule, should determine the length. Then right from your browser you can do a search of that page. And the table of contents can be a list of internal links for that page, pointing readers to anywhere on the page they might want to go -- and they arrive at their destination immediately, without having to wait for another page to load.

I design my own Web pages the way I wish others would, focusing on content rather than fancy effects. And what I do, anyone could do. Doing it that way, Web-page creation is no more complicated than using a word processor.


Step-by-step details on how to create your own Web pages

This is how I go about creating my low-tech pages on my PC with Microsoft Word:

If you use Microsoft Word, you can start a new document, save it as "Web Pge Filtered .html"  You can convert an existing Word document to HTML, by using the "Save As" command and choosing HTML. Or you can use it to create new documents in HTML, which is the way I prefer to operate and what I'll describe here.

After you have saved your new or old document as HTML, you will see web page icons in your toolbar. These icons open up all the functions you need to create and edit Web pages.

First click on File, then on Properties (or click on the icon that looks like the letter "i" on a sheet of paper). A box opens up for you to enter the title of your page -- not the title that appears at the top of your document, but rather the HTML title, the coded name that search engines like AltaVista give high priority to when answering queries and that is displayed with search results. Just type the name and click "OK". If your software works a bit differently, click on View, then HTML Source. Then in the top section of the page -- between <head> and </head> type <title>, followed by the words of your title, followed by </title>.

The words of your HTML title should clearly and simply describe your page. You shouldn't waste space including your name or your company's name. Only use words that are signficant for search purposes -- that people wanting to see the information on your page might enter as a query at a search engine. From the perspective of a search engine the HTML title and the first couple lines of text are the most important parts of your Web page. If someone does a search and the words they are looking for appear in your HTML title, it is very likely that your page will come out near the top in the results list. Be precise. Every word counts toward your ranking, and your ranking will make a big difference in the amount of traffic you get. (Few people look beyond the first page or two of results -- in other words, beyond the first 20 items listed.) Don't include any words that are unimportant. Do include all important words. And don't simply repeat the same words and phrases from one page to the next. The HTML title is what search engines like AltaVista use as the hyperlinked title of your page in their results lists. The words should differ from one page to the next, making them unique, rather than making them look like duplicates of one another.

Now enter your content. If you have something new you want to say, just type it like you would any other new document. If you'd prefer to start with something you wrote before, open that other document, then under Edit, choose Select All, and Copy; then exit that document, and Paste the contents into your new document.

Once again, use special care when writing the first couple lines, which count almost as much as the HTML title for ranking by search engines like AltaVista.

Now enhance the looks of your content. Click on Format, then on Style, and in the bottom left-hand corner, under "List:" click on the down arrow and select "All Styles". A wide range of choices will now appear in the large box above. Use the scroll bar to see the full range of choices. To change the "Style" of a piece of text in your document, you simply highlight that piece of text, then go to this set of Style choices, select the one you want, and click on "Apply." Experiment to see the range of choices. First select styles for your head and subheads. (I prefer to work with Headings 1, 2, 3, and 4).

Keep in mind that the usual methods for highlighting text in Word apply here as well. For instance you can change words to bold or italics or underline just as you would normally. If you have lists, you may want to bullet or number them. You can do this by highlighting the text and choosing a particular style, or by simply using the numbered list and bulleted list icons from the toolbar, as you would normally.

To divide sections of text, simply position the cursor where you want the divider, click, and then click on the dark horizontal line in the toolbar.

To add a hypertext link to another Web page, click on the chain in the toolbar. Click on the space under "Text to display" and type the words that you want readers to see highlighted. Then click on the space under "File or URL" and type the Web address. Then click on OK, and your "Text to display" will appear blue and underlined like a hypertext link. As an alternative, you could highlight a piece of text in your document, then click on the chain, and that highlighted text will appear in the "Text to display" box; and all you need to enter is the address. If you are going to have more than one Web page, make sure that you have hyperlinks connecting them so readers can easily move from one to another and can always, from any of your pages, return to your preferred starting point or index page.

To add a hypertext link to an email address, simply enter under "File or URL" mailto:username@address, e.g., mailto:seltzer@seltzerbooks.com

If you would like to add special symbols, like ©, ®, or Greek and mathematical symbols, click on Insert, then on Symbol, and make your choice.

NB -- If you think you might want to eventually move to another service provider or might want to move batches of pages to their own separate directories, you should use "relative" rather than absolute addressing in the hyperlinks among your own pages. In other words, if your basic URL is http://www.seltzerbooks.com/ and you want to link to a page called lowtech.html, you can enter the URL for that page as http://www.seltzerbooks.com/lowtech.html (absolute) or as lowtech.html (relative). With relative addressing, a browser looks for a document of that name in the same directory as the current page, regardless of what that directory may be named. That way if you do move, you won't have go through the tedious and error-prone process of changing all the links among your pages.

If you have a long document, there is no need to divide it up into lots of smaller documents with links among them (which many people do). Keep it simple -- both for your own convenience and the convenience of readers. The fewer the documents the better. And there's nothing wrong with having a Web page that is literally as long as an entire book. If it consists of just text, with no graphics, it should download relatively quickly. Just make the page easy for readers to navigate. Near the top, enter a list of the contents. Highlight the first entry in that list of contents and click on the chain in the toolbar. The highlighted words will appear in the box "text to display." This time instead of entering a Web address, and go down to "Bookmark location in file". Microsoft uses the word "bookmark" in a different sense than Netscape does. This "bookmark" means a hyperlink inside instead of outside a Web page. Enter a one-word label for the piece of the document you want to link to. Then click "OK". Your selected text will appear blue and highlighted like a hypertext link. Now scroll down to the text you referenced in your contents. Highlight the chapter title or subhead or first few words of text and click on the "open book" icon to the left of the chain. Enter the same word you entered under "Bookmark location in file" and click on "Add". Now go back up to your list of contents, and click twice on the hyperlinked text. If you entered the same words both in the hyperlink and the bookmark, you should be moved straight from the contents to the matching section of text.

If you have a picture of some other graphic element which you would like to use, such as a company logo or a photo of someone which you have stored on your computer as a .gif or .jpg file, click on the place in the document where you would like it to appear. Then click on Insert, then on Picture or on the equivalent icon (outline of mountains to the left of the open book). In the section "Picture", enter under "Image source" the name of the file. Also, under "Alternative text," enter the explanatory words that you would like to be presented to people who do not have graphics capability (including the blind). Then move to the "Options" section, and select the height and width you want, and, clicking the down arrow next to "Default," decide whether you want the graphic to be centered or aligned on the left or right. When finished, click "OK." (When you upload your pages to the site of your service provider, be sure to upload the associated graphics files to the same directory.)

When you have finished your page, save it with a name that ends with the suffix ".htm" and as type "HTML document". (When it comes time to move your pages to your service provider's machine, doublecheck to determine their naming procedures. In many cases, you'll want to give each of your ".htm" files a suffix of ".html" on the server. This is a minor nuisance -- Word,. even with Windows 95, only recognizes the three character suffix .htm, and many UNIX systems, commonly used for Web servers, only recognize the four character suffix .html. So be sure the addresses you enter for hypertext links among your pages use the form of the name that you will be using on the server. In my case, I name all the pages .htm on my PC and rename them all .html on the server, except in unusual circumstance. So links to my own pages are normally in the form .html)

To see how your page will look to readers on the Web, click on the eyeglasses at the far left of the toolbar. To return to edit mode, click on the pencil icon that then appears in the same space where the glasses were. To test your hypertext links, first connect to the Internet as you normally do, then open your Web page in Word again and click on the links. Word will then launch your Web browser and take you straight from Word to the foreign Web address in the link. As an alternative, connect to the Internet, launch your browser, open your page as a local file and give it a test drive (With Netscape, click on File, then on Open File, rather than Open Location, and enter the name of the file, including it's directory.) That way you can experience your Web page just as readers would -- except for the relative links among your own pages -- before making it available to the public. (Your browser wouldn't understand the relative links because your current directory is that of your local file, not the directory on your Web server.)

NB -- When making edits to an existing page, to see the changes you have entered clear your disk "cache." Your browser "remembers" the pages it has been to recently to save time in loading them. If you click to go to an address it has on file, it serves the information to you from your hard disk rather than going back to the actual Web site. Often even hitting "Reload" won't get you to the new version, unless you erase your "cache". With the Netscape Navigator, you do this by clicking on Options, then Network Preferences, then under "Cache" click on "Clear Memory Cache Now" and "Clear Disk Cache Now", then "OK."

Once you have mastered these basic techniques, you may want to experiment and try to take advantage of features you have admired at other sites. I recommend spending lots of time and effort on your content first and doing all you can with words alone. But when the temptation to move beyond reaches the critical stage, go to the page you'd like to emulate and save it as "Source" with a suffix of .htm. Then from Word, open that file, edit, cut, and paste, and move things around, and click on the glasses to see what difference the changes make. If graphics files are missing, click on View, then on HTML Source, and try to decipher the name of the missing files; then in your browser go to the specific file you want and save that as well. You can learn a lot looking at the markup codes in the Source, and by taking pages apart and putting them back together again -- like working on an old jalopy.

To add bizarre new markup code to a page of your own directly from Word, click on View, Insert, then on HTML Source. Then you'll see the background code that makes up your page. You can add and edit code directly in that mode. When you are done, click on View again, and then on "return to edit mode."

If you want to add Metatags (to tell search engines like AltaVista how you would like your page to be described and what related keywords for that page you'd like added to their index), you can do so either in the HTML Source (as above), or by click on "i" (where you entered the HTML title) and then on Advanced and then on Meta.

When your page is done and you have posted it at your site, be sure to go to AltaVista (www.altavista.com) and at the bottom of that page click on "Add a Page", then in the form at the bottom of the next page enter the URL for the particular page that you just create. Then the full content of your page will be added to the AltaVista index within a day or two, the world will be able to find you. (You also might want to check my article on how to publicize a Web site over the Internet for free www.seltzerbooks.com/public.html).

If you are stumped by some of the fancy new effects you are trying to add, use AltaVista to search for a free on-line tutorial on the topic you are curious about. Or as a last resort, buy a book or take a course.

But when you get to that stage, remember that your investment is probably for the entertainment of playing with new technology, because those extras really aren't needed to deliver useful content to readers on the Web.


Search Engine Considerations

Full-text search engines, like AltaVista, are an essential structural element in the Social Web. These search engines depend on robot programs, often called "web crawlers," that automatically find and retrieve content on the Web that will then be processed and added to the search engine's index. Tens of millions of people a day go to AltaVista alone, enter queries, and decide where to go next based on the results lists they receive. If you want to use your Web pages to get in contact with other people, you want to have your content fully indexed in the search engines.

But web crawlers are blind. They do not see and do not retrieve graphics or photos. They go after text, all text, and nothing but text.

If your information is contained in one, big, beautiful picture -- great artwork, with the words embedded inside it, in some graphical format like .gif or .jpg or .bmp -- the crawler won't see anything at all, and the words won't be indexed, and nobody will know that your image is there. That's like designing a beautiful poster and hanging it in a closet.

Also, if the top of your page is strewn with graphical elements -- each of them beautiful in and of itself and each of them associated with a word or phrase (this is a style that is very popular with large corporate Web sites) -- the crawler will retrieve only the words. And the search engine will take the first couple lines of text -- whatever text it was able to glean -- and will use those words as the "description" of your Web page. With a graphic-oriented page, those words could add up to a totally random hodge-podge. But if someone makes a query, and your page is a good match, those are the words your prospective visitor will see when trying to decide which page to go to.

Yes, there are ways to get around these problems, but the simplest solution is to have your Web page clearly state what it is about and provide useful and interesting information.

If you have some compelling reason to include graphics and photos, then enter Alternative Text (as described above, or with the <ALT> command if you are doing the HTML by hand). For people using the standard graphical Web browsers, those words won't appear. But they will appear for people who regularly turn off their graphics and those who use text-only browsers -- like the blind (who use text-to-voice converters to "see" Web pages) and people with slow Internet connections and also web crawlers of search engines.

Also, if you include images on your pages, give those images logical names -- let the name say what it is, because those names will be indexed, although the images themselves will not be. For instance, at AltaVista, I can do a search from image:mars and get a list of all the indexed images on the public Web that are named "mars."

If you have some compelling reason why you don't want to state what your page is about in the first couple lines of text, you can add a description Metatag to your page. If you are using the Internet Assistant for Word, as described above, you click on File, then on HTML Document Information (or click on the icon that looks like the letter "i" on a sheet of paper), and then click on Advanced and on Meta. There you should enter <META name = "description" content="your text"> If you are doing this by hand or with some other Web authoring tool, just enter that line between <HEAD> and </HEAD> and after <TITLE> </TITLE> Replace "your text" with the very words that you would like to appear in the results lists of search engines. Limit yourself to just a couple lines -- that's all the search engines will take. And check AltaVista a day or two after you submit your URL to see how your entry looks. If you decide to make changes, be sure to submit your URL again, and then check again.

When writing the text of your pages, you should be sure to include the words and phrases that people looking for your kind of information would be likely to use as a query in search engines. If for some reason those words and phrases don't appear anywhere in the normal text or perhaps you want to be sure to cover synonyms, you can add a keyword Metatag to each of your pages. Such a tag line is created the same way and goes in the same area of your page as a description Metatag. Here you enter <META name="keywords" content="your words one after the other without punctuation"> Once again, limit yourself to a couple of lines -- that's all the search engines will pay attention to.

Also, resist the temptation to try to fool the search engines -- such behavior is likely to backfire on you. For instance, when AltaVista first went on-line many clever Webmasters thought they could drive traffic to their pages by repeating over and over again words and phrases that were important to their target audience. Sometimes they put the repetition at the bottom of the page, included it as wallpaper style background behind the real text. This accomplished nothing, because the designers of AltaVista, not wanting to encourage such behavior, which would have led to lots of useless clutter on the Web, set up their ranking algorithm (the formula for determining which pages appear where on search results lists) so that repetition makes very little difference. AltaVista counts to two and no further. A single repetition makes a slight difference in ranking; but beyond that, repetition is useless.

Other Webmasters figured that they could bring about the same kind of effect without annoying their visitors by putting the repetition in comments. Comments -- a series of words placed between the markers <COMMENT> and </COMMENT> or after an exclamation mark, such as <!this is a comment> -- are used by page designers as private remarks and reminders intended for their personal use, for instance to help them when it's time to update a page. But because comments are meant to be private, AltaVista does not index them at all.

Once keyword Metatags were introduced, other Webmasters tried to use those to stuff the ballot box -- repeating the word or phrase that's most important to them over and over. But the folks who run the search engines do not take kindly to such behavior. Trying to maintain the integrity of their indexes for the benefit of their users, they sometimes penalize pages that abuse the keyword Metatag with excessive repetition.

The most important elements for ranking at AltaVista are 1) the HTML title and 2) the first couple lines of text. After that come the description and keyword Metatags, if they exist. Hence pages that clearly state what they are about in the HTML title and first couple lines of text are likely to get better treatment than those that rely on Metatags.

And many of the fanciest effects, the kinds of things that are difficult and expensive to do, would prevent your information from being indexed. Naturally, audio and video files are not included. But also any information that is behind a login/registration procedure is not indexed, even if there is no charge for registering (for instance, if you set up the process just to gather information about your users). Basically, a web crawler is just a dumb robot. It cannot fill in a form; it cannot answer questions; hence it stops short whenever input is required. That same obstacle prevents crawlers from accessing information in databases.

Search engine web crawlers also cannot handle information inside frames. Frames make a page look and feel like two pages. The frame itself typically remains constant throughout a Web site, providing corporate branding information, site navigation links, and sometimes advertising. And the window inside the frame is where the real content appears as you click from link to link within the site. For users with small screens, this approach can be a major nuisance, greatly reducing the space usable for real content, perhaps to flash an annoying ad at you over and over again. In any case, web crawlers only see the frame, not the content in the window; and hence for all their efforts, the sites that use this technique do not get their real content indexed (unless they provide a non-frames alternative version of their pages) and hence miss out on lots of good search-engine-generated traffic.

Many commercial sites also go out of their way to make their pages "dynamic" and "personalized." Some use "cookies" (a piece of software built into some browsers that provides Web sites with a way to keep track of what pages particular users have looked at) to "recognize" users and provide them with unique and different experiences each time they come back. Some use database-style applications which assemble new Web pages on the fly from pre-stored elements in response to user requests or behavior. But the content on such dynamic pages cannot be indexed, because each URL is long, complex and unique to that one visit. A web crawler visiting such a site faces the prospect of retrieving an infinite number of custom-created pages, and halts immediately.

Many sites also use Java applets, both for fancy graphic effects and also as a way to present information. But once again, the information that appears as a result of a Java program is not retrieved by AltaVista or other search engines and is not indexed.

In other words, not only are simple plain-text or mainly-text pages easier and far less expensive to design and edit and update and maintain, they are also far more effective in the search-engine-dominant Social Web. This is a major reason why individuals and small companies can compete for attention with the world's largest corporations, why a personal Web site that costs next to nothing can compete with sites that cost $1 million or more per year to operate. The big guys try to win with money and glitz. You can match them with hard work, good content and focus on the basic dynamics of the Social Web.

In the next chapter, we'll talk about content -- what kinds of material you are likely to have available that are likely to produce good results.


Redesigning A Company Web Site

This article was heard on the radio program "The Computer Report," which is broadcast live on WCAP in Lowell, Mass., and is syndicated on WBNW in Boston and WPLM in Plymouth,

A friend who just got a new job wrote to me saying, "My first project is to create order out of the mess they are using as an Internet. Could you kindly share your experience with me? What's your advice regarding Web usability and navigation?"

First keep in mind that you have two separate and not always compatible goals:

If you are perchance battling corporate branding rules that legislate design techniques that get in the way of search engines finding your content (e.g., frames, dynamic content, content generated by Java script, etc.), you may want to consider creating two separate sets of pages: The text of the search engine pages should be the same as that on the main site, but should be plain static HTML, with the most important information appearing in the HTML title and the first couple lines of text.

Forget about Meta Tags -- they are worthless. Focusing on Meta Tags would probably lead to your missing the most important factor: the actual text content of your pages.

Each of your mirror/search engine pages should have links to the home page of your main site, recommending that users go that way to get the optimal experience. There's no need for links from there back to the search engine pages.

Every page of yours should have some text that makes the context clear (what is this site? what's its purpose?) and links for easy navigation within the site (including Help and the site map). Presume that each and every page is a potential entry point for visitors and make sure they won't be confused when they arrive.

Create a "site map" page that in plain static HTML has a hyperlinked list of all pages at your site, and include a link to that site map on every one of your pages. That is the page that you should point search engines to. That makes it easy for crawlers to find all your pages.

Keep your Web addresses simple -- the fewer the levels of directories the better. Search engines like AltaVista presume that the higher a page is in the directory hierarchy, the more important it is; and some will simply halt at about the third or fourth directory level.

Also keep in mind that search engines are inclined to give more value to large pages as opposed to small ones. So don't divide long articles into a series of linked short pages each of which is no bigger than a screen or two. The longer the better (up to about 64K of text). If corporate design rules or someone's notion of the ideal user experience forces you to design that way, then create mirror pages where the full content is all together. That's not only good for search engines, it also makes it far easier for visitors to print your content.

Avoid search engine optimization and submission services. Submit yourself to AltaVista, Excite, Hotbot, Lycos, Google, Fastsearch, and others you know of and like -- using your site map page, not your home page. Also, be sure to submit to the Open Directory and Yahoo. Submit to the LookSmart directory, too, if you are willing to spend a couple hundred dollars.

For optimum results and fast indexing service at AltaVista, you should submit each and every page individually, whenever you create a new page and whenever you make a significant change to a page. You pages should then appear in the AltaVista index within a few days, or a week or two at worst, while some of the other search engines will take a month or two, or even three months.

Do not change the URL of existing pages. Some companies automatically move aged content to an archive area where it has a new URL. Don't . That screws up both search engines and also bookmarks and links that friends may have created to your pages.

Also, never delete a Web page that contains significant content. If the information is old or the product is discontinued, add text at the beginning that explains the situation and links visitors to the latest and greatest related information and newer products.



No Frills Web Pages for Community Service Organizations


If you belong to a community service organization and have an Internet account that provides you with space for your own Web pages, you should consider creating pages for that organization.

Enter brochures, booklets, newsletters, meeting schedules, contact information, etc. Provide a place for members to share experiences and information with one another. Keep it simple. Go with plain text (no graphics). (For details on how to create such pages as easily as doing a word processing document, see "Low-Tech Web-Page Design" )

If you have the time and inclination, take active steps to let people know that your page exists. 

My wife and I took this approach with Prescription Parents, a support group for parents of children born with cleft lip/cleft palate. We posted the basic organization info, several booklets, and pointers to related organizations. We get about 50 visits per month to those pages, mostly from people who need the information and who otherwise would have had to request booklets by snail mail (if they were lucky enough to find the right address), and would have had to wait a week or two to get them. (Keep in mind that it costs us nothing when folks retrieve these booklets online, while otherwise it would cost us time and money to reprint and send them)

Every month, several of those visitors have unique specific questions that they send us by email, and we do our best to answer. Since ours is a local Boston-area organization and the Web reaches the world, our replies are sometimes pointers to related organizations.

All in all, at no cost, and for a minimum investment of time, we're able to extend the benefits of our organization to a much broader audience.
___________________________________________________________________________________________________________________________________________________

Web Sites for Bed and Breakfasts

This article is a transcript of a segment on the radio program "The Computer Report," which was broadcast live on WOTW 900 AM, Nashua, NH 12-2 PM Sundays.

When shopping for a bed and breakfast for a weekend getaway, I noticed that many such establishments depend on umbrella sites which group dozens or even hundreds of them from the same geography, all using the same basic template and style. For instance, Massachusetts Lodging Directory of Bed and Breakfast Country Inns & Small Hotels (a subset of virtualcities.com), The Massachusetts Bed and Breakfast Inns Directory (under bedandbreakfast.com), and Rimstar International. The typical bed and breakfast page at such sites shows you photos of the interior of the rooms for rent, lets you know the prices, and gives you the address and phone number. Typically, the proprietors know nothing about the Web. They just submit their information, pay their fees, and wait to be contacted by customers. The sparsity of the information and the similarity of style often makes it very difficult to distinguish one from another and decide which to stay at.

It's a very rare case where you can see online whether there are vacancies on particular dates, much less have the ability to make and pay for reservations online. Considering these businesses typically have half-a-dozen or fewer rooms, such sophistication is probably too much to expect.

But there are a number of things that such a small business could do to increase its Web traffic and hence bring in more customers. Ideas of that kind started occurring to me when I got a request for advice from my uncle who has a bed and breakfast called Edgewater in Mahone Bay in Nova Scotia and currently has a minimal listing at bbcanada.com.

First, I suggested that he keep his current page, that the new site he wants to build should be in addition to that, not as a replacement; so as not to lose the business he's getting already.

Second, when building his new site, he should keep in mind that, unlike his listing at bbcanada.com, what he posts will appear without context. He will need to build both the geographic and the business context.

Geographically, he should have maps (Eastern Canada, Nova Scotia, and Mahone, the town he lives in). He should also have driving directions from likely starting points in the US and Canada.

He should also include lots of photos -- not just of the rooms that he rents, but of his town, showing his house in the context of the block that it is on, and showing everything of interest within walking distance and within a reasonable driving distance, plus photos of himself and his family.

Business-wise, he should have a page that clearly states the prices and the business terms, and that clarifies the currency/exchange situation and what that means in purchasing power for visitors from the US. (Bed and breakfasts in New England should likewise have such explanations for visitors from Canada). He should detail the standard terms like check-in and check-out times and procedures, and anything else that he has found that outsiders are sometimes confused or surprised by.  He should indicate the differences in price with season; and also indicate when he has openings.

The overall design for this Web site should be very simple -- no fancy effects, just text and photos, in plain static HTML -- so he could easily and quickly update your pages. He should be able to change the online indications of room availability in a couple minutes by hand, without the need for any fancy custom code.

Once he has taken care of those basics, he should focus on the cultural environment, providing as much information as possible about his house, his town, Nova Scotia, and Eastern Canada -- history, places of interest, entertainment,
literature, everything.

He should include blurbs about and extracts from popular novels and movies that are set in his piece of the world, e.g., The Shipping News and Longfellow's Evangeline. He should include whatever he can by authors who were born or lived in Eastern Canada, e.g., Lucy Maude Montgomery, author of Anne of Avonlea, etc. -- she was from Prince Edward Island; there's a tourist trap devoted to her there; and all her works are in the public domain and readily available on the Internet. He could also include the full text of public domain books about Canadian history, which are available through the Gutenberg Project on the Web and on CD from my little publishing company.

Once he gets going, if he follows some simple design rules (like writing unique HTML titles for each page and making sure that the first couple lines of text of each page are meaningful), all this text should bring traffic to his site by way of search engines, many of whom should be people who are fascinated with his part of the world and plan to visit there.

seltzer@seltzerbooks.com  privacy statement

Google
  Webseltzerbooks.com