Use an HTML validator
One of my previous tips admonished all Webmasters to use correct HTML.
It's easy to write incorrect HTML because most browsers will make sense
of it and let you get away with all sorts of egregious errors. Even
worse, some erroneous HTML generates nifty document effects.
Unfortunately, bad HTML can lead to all sorts of problems. New tools
expect your HTML to be in pretty good shape, and those nifty effects may
work great in one browser and fail miserably in another. Most
importantly, bad HTML is usually a horror to maintain, and the person
who succeeds you in your job may be hunting you down once they get a
glimpse of your documents.
You can avoid all of these problems by running all of your documents
through an HTML validator. HTML validators will examine your HTML
document and point out every error, problem, and potential problem in
your document. They will find all sorts of obvious problems, like
incorrect tag nesting, invalid attribute values, and missing end tags.
They'll also point out a lot of minor things, like inappropriate use of
quoted values, nonstandard attribute values, and other niggling problems
with your document. Some will even go so far as to check the validity
of your links to ensure that nothing is dangling or broken.
There are two kinds of HTML validators: those that run locally on your
server and those that run remotely across the Web. Local validators
can be further divided into freeware and commercial software categories;
remote validators are almost always free.
You can find a nice list of validation tools on the
page, with examples of both HTML and link validators in the free and
commercial categories. Visit the page for the complete list; three you
should definitely check out are:
CSE 3310 Validator
This commercial tool runs on your Windows 95 desktop and can validate
one or more documents at a time.
- W3C HTML Validation
This online resource is provided by the W3C and is
intended to check documents against the new HTML 4.0 standard.
free Perl-based validator runs on your server (preferably Unix-based) and has
been around for years.
With all of these resources available, there is no excuse not to be
using correct HTML in all of your documents. Start checking (and
In the original list of ten tips, I urged everyone to explicitly
copyright their work by including appropriate wording and copyright
indications on every document you place on the Web. I was taken to task
by several readers who pointed out that this kind of explicit labeling
is not needed to ensure that your copyright is secured and intact.
Before we go any further, let's be clear about one thing. I am not a
lawyer, nor should I play one on the Web. I did once take an interesting
trip through small claims court that ended up in the Third Circuit
Court of Appeals for the state of Florida, but that, thankfully, had
nothing to do with copyright violations.
Those who wrote regarding that original tip are correct. Copyright is
acquired the moment you create something, even if you never publish or
otherwise make it available to the public. Labeling your work as
copyrighted is helpful to your audience, but it is not necessary for the
purposes of obtaining the copyright.
The point of this tip goes beyond that first suggestion, to insist that
all serious Webmasters acquire a conversational knowledge of copyright
law. The flip side of copyright law determines how you can use someone
else's work, what is appropriate and inappropriate, and how you can keep
from infringing on someone else's copyrights. A little knowledge will
keep your site on the up-and-up and keep you out of court. You should also
acquire similar knowledge about trademarks and service marks.
And, in case you were wondering, I won my case.
Never forget that your Web site exists to do something. Gone are the
days when sites existed simply to exist -- a fun place where you could
experiment on the Web. The Web is very much a formal media now, and you
need to make sure you are using it correctly.
Every Webmaster should be able to answer a few basic questions about
- Why does this site exist?
- Who should be viewing this site?
- How do we pay for this site?
- How will we know when this site is a success?
The first question isn't too hard. Are you trying to create market
awareness? Are you selling products? Are you trying to deliver
information or support to a group of customers? Why are you on the
Web? If you can't give a good answer, don't even bother with the rest of
the questions, or, most likely, the rest of this column!
Assuming you know why you exist, who should be viewing the results of
your hard work? Existing customers? Your competition? Women between
ages eighteen and forty-five? Once you figure our your target demographic, you need
to answer one more question: how does this group know how to find your
site? Are you registered in all the search sites? Are you
advertising? Do you couple your URL into legacy media like print and
Once you have a reason for being and a customer to acknowledge the same,
you need to pay the bills. Many sites are funded as a cost of doing
business and subsidized by the sale of products. Other sites must be
self-supporting through online advertising or similar tactics. As your
site grows, the cost of maintaining it will as well. If you define
early on how to pay for your site, it will be much easier to get extra
funds down the road. Site funding that is hidden in other budgets will
suddenly become a thorn in someone's side. Because your salary may be
carried as part of the site cost, you may have a vested interest in
ensuring long-term funding for your site.
Finally, define your success. A previous boss of mine used to ask,
"How do get an A?" When will you consider your site a winner? Will
it have to drive a certain amount of revenue? Get a certain number of
hits? Service a certain number of customers? Only you know. By
defining success correctly, you set expectations among your management
and can ensure that you are perceived as an asset to the company.
Define and enforce content standards
Nothing is more irritating than a site that is inconsistent, confusing,
and difficult to understand. Often, these problems are caused by a lack
of content standards for the site. Content standards define, in clear
terms, what is acceptable and unacceptable on your site. They should
cover everything from color and imagery to link styles, page layout, and
Content standards are especially important when a site is managed by a
team of contributors. It is all too common for different authors to put
their own twist on a page -- even in slight and subtle ways.
Unfortunately, these slight twists clash as a visitor moves from page to
page on your site. If one author writes in the passive voice and
another in the active voice, your visitors will notice. If some pages have
navigation tools at the top and others at the bottom, your site will be
confusing. Content standards, along with enforcement of those
standards, can keep these kinds of errors from happening.
Even if you are a solo Webmaster, creating content standards is still a
good idea. It forces you to think through your site design and make
decisions in advance. If you decide to design a page months later, it
may help to review those standards to see what you had in mind when you
first conceived the site. Most importantly, when your site becomes so
successful that you must hire an assistant, those standards will make
it much easier for that person to understand and support your site.
Keep and use statistics
Numbers are wonderful things. It only takes a few to prove almost
anything you can imagine, and the right chart or graph can make the
difference between making your point and being ignored.
Your Web site is generating numbers constantly: hit rates, browser
usage, data transfer rates, and all sorts of other metrics. These
statistics can be divided into two groups, both of which you need to
understand and exploit.
The first group of site statistics includes all the measurements of your
site's health. These include basic hit rates, data transfer numbers,
referrer logs, and agent logs. Using this data you can make sure that
your site is up and running effectively. At the very least, your hit
rate will tell you that people are visiting your site. If you probe
deeper and examine hit rates as they change throughout the day and week,
you'll learn more about your customer base, when they like to visit, and
when lulls occur during which you might choose to update your site.
Data transfer numbers give you an idea of how many bytes you are pushing
through your Internet connection. Are you bandwidth bound? It may be
that your usage spikes a few times a day, but that the rest of the day
you are running at fairly low volume. In these cases, you could switch
to a cheaper, slower connection without impacting most of your users.
If your site is pushing the limits of your connection all day long, it's
time to upgrade that link.
Referrer and agent logs let you know where people are coming from when
they visit your site, and which browser they are using to visit. You
can use this data to see who is linking to your site (perhaps a
reciprocal link agreement is in order) and which browser is most popular
among your users (warranting some content change to better exploit that
The second group of statistics helps you improve your content and prove
the worth of your site. By more closely analyzing your hit rates and
patterns, you can see which pages are most popular. These pages should
be tweaked and perfected because they are seen by your largest groups of
visitors. You can also see which pages are ignored, possibly warranting
elimination from your site. Finally, you can see how people jump from
page to page on your site, letting you know if your navigation tools
Once you know how people visit your site, and in what volume, you can
begin to prove that your site is being used correctly, justifying
further investment and growth. That is, of course, if you followed the
previous tip and correctly defined your site's metrics for success.