Writing for the web

Some day this will just be about writing

Chris Lawson

CUPE National

Workshop outline

  • Who are you, where are you from what would you like to get out of this workshop?
  • Why a workshop on web writing?
  • Why write for the web?
  • What is good about writing for print?
  • What do we need to change?
  • How do we write for the web?
    • writing vs repurposing
    • Writing to be understood
      • good web content
        • structure: T vs Inverted pyramid
        • break up content
        • use lists
        • add hyperlinks
        • avoid rhetoric
    • Writing to be found
      • Search engine results
      • Getting ranked
      • Passive voice makes a comeback
    • Exercise: writing scannable, searchable headlines

Why a workshop on writing for the web?

Most people who've been writing for popular consumption since before the advent of the world wide web (that would be for more than 12 years) probably think they're really well-equipped already. But we're not.

At a certain point, I'd like to think that this will just become a workshop on "how we write" but for now it's about writing differently.

Ever browse the web and see -30- or how about For more information see our website? We need to write for the web because print content transplanted to the web doesn't work.

Why write for the web?

It's a different medium

Look at the screen. This is it. This is the web. It's all you get. For most of us who've grown up with both writing and computers, looking at text on a screen suggests incompleteness or raw material. It's not final until it's printed. Some people out there cannot edit copy on screen. However more often than not these days, the final product of writing is on the internet. And with web pages that's almost always the case. Some people still print out web pages, but it's a step down from the web content, done to make reading more comfortable and diffuse information to places where there aren't internet connected computers.

It works differently

I pause here to mention that there are different kinds of writing for the web. And as internet communication technology changes, more formats will no doubt evolve. I will mostly be talking about organizational content written for public consumption. Blogs, tweets and other forms of user-generated content are a different matter and play by a different set of rules.

For the most part, content on the sites you run has a public (as opposed to intimate) voice and your visitors are looking for answers, not literature.

By and large, these people don't read on the web. They scan. People also scan print documents, I hasten to add, but reading a computer screen is still harder than reading a book. So even if they're reading continuously, your users will benefit enormously from copy written for the web.

Eyeball tracking experiments have shown that site visitors' eyes follow an 'F' pattern as they scroll down your page. They read across the headline and then follow a straight line down the left margin, heading over toward the right side again when they get half way, then continue on down the left margin.

They don't want to be on your web page. They need to be there because they are looking for an answer to a question.

There are other dimensions to your content. It's also a tool for finding more content.

You care that your website succeeds

It often seems enough of an achievement just to post something to the website at all, to say nothing of tuning it so it matches the medium. Running a website is work. Wouldn't you rather it paid off somehow? Content created for or changed to follow some fairly widely accepted conventions for web writing is much more usable than that which has merely been copied from a leaflet and pasted onto a web page without so much as a pause to remove the page footers.

The Jacob Nielsen study took two documents off an information technology company's website and edited them. They asked two groups of the company's employees to find information contained in the documents and later quizzed them on the whole document. The first group worked with the original documents, the second with the documents edited for the web. The edited version scored 159% better overall including:

  • reduced time required to complete tasks;
  • improvements in retention of information;
  • fewer mistakes made

What of print can we keep?

Or, rather, how we're supposed to write for print.

  • Correct spelling
  • Proper grammar
  • Simple sentences
  • Active voice
  • Some blabbermouth from CUPE led the workshop vs The workshop was led by some blabbermouth from CUPE

  • Minimal jargon
  • No clichés
  • Strong verbs:
    • Decide vs Make a decision

    • Survey vs Conduct a survey

Writing in simple, active sentences with jargon-free correctly spelled words will help you along the way. But you're not there yet.

What do we need to change?

  • Writing more
  • Jacob Nielsen estimate that reading text on a computer monitor (a CRT) slows reading speed by 25 per cent. Crawford Kilian concludes then, that web writing should be 25 per cent shorter than writing for print. Nielsen himself argues it should be half the length - or shorter.

  • Paragraphs

  • We're taught in high school that paragraphs are supposed to be a complete argument - three, four or more sentences. On the web that ensures that readers will skip over everything but the first line of the first sentence.
  • Every piece is a compendium

  • Print writing assumes that the only information the reader has at the tip of their hands is that which we have provided them. We include background, asides, parenthetical or boilerplate information that is easily and properly replaced with a hyperlink.
  • Assume start-to-finish reading

  • In journalism class, students learn they lose 25 per cent of their readership per paragraph. That's grim, but it assumes start to finish reading. On the web we cannot assume even that. Web surfers scan.
  • Inverted pyramid is a poor way to explain things

  • To ensure that the busy print reader gets the news before their eyes wander off, journalism students learn the inverted pyramid style, where information is conveyed in descending order of importance. But a busy web reader's eyes are nibbling at the start of each paragraph, trying to determine if they should continue reading. Sentences need to be ordered in some sort of logical order, like begin from the beginning, or even alphabetical order, to help visitors determine if they're getting closer to the answer they seek.
  • Rhetoric

  • Enough is enough. We're not going to take it any more. If you try to write empty text, you're going to have a fight on your hands, because when all is said and done, in the final analysis, it ain't over 'til it's over.
  • Wall of words

  • Web surfers need visual cues to see where they should be looking. Subheads, hyperlinks, bullet lists. Each signals potential new chunk of information which may or may not be what they seek.

How do we write for the web?

As the web becomes our primary means of communication and the internet the final destination for our content, these issues of repurposing will go away. As a web editor, your problems should go away as soon as the web becomes your organization's primary medium and the print editor starts having to inherit your content hand-me-downs.

To any print editors in the room I cackle loudly and scream, "welcome to my life." Seriously, I think it will be easier to repurpose web for print, than the other way around.

The main reason for this is that web text is more concise than print text, and, assuming you take out the click heres and the references to menus, web text should withstand print scrutiny. Alas the reverse is not true.

Writing vs repurposing

Writing vs repurposing

Are most people inheriting stuff from the print world? I suspect so. Too bad. Unfortunately content created for print comes with a lot of baggage and we don't have time or authority to unpack it and set it up so that it works on the web.

Writing to be well-scanned

  • Structure: T vs Inverted pyramid

  • The ideal first paragraph of a web page tells the whole story: who, what, why, where and when. In many ways it's the equivalent of a standard news lead. It's allowed to be a bit longer if it means it's more complete and there are no wasted words.
  • The 'smorgasbord' lead can also be used as a blurb on your home page or other landing page to help improve your information scent

  • Classic inverted pyramid orders content arbitrarily according to the author's idea of what's important so that if the person doing layout needed to cut the article so it would fit, they knew the could cut from the bottom without risking the piece's integrity.
  • But that often takes information out of a logical order, the sort literacy advocates espouse.
  • break up content

  • subheads - meaningful ones, naturally - ones that give an indication of what's to come - act as rungs on the visual ladder of your page.
  • chunking - paragraphs are shorter and independent of each other, and divided up in to chunks of two or three paragraphs. Killian:
    Narrative order, which relies on chronology, imposes a sequence that web users may not want or need. Logical argument tends to be too long and sequential for impatient visitors.
  • use bullet lists

  • same reason as subheads. Things like:
    • contract gains;
    • concession demands
    • recommendations to governmenet
    • nicknames for the night shift supervisor
  • can all be turned into lists
  • add hyperlinks

    • instead of yet again telling everyone that your union represents X number of workers at X employer, just turn the first mention of your union's name into a hyperlink to your about page.
    • instead of spending a sentence or two describing what nefarious government or employer deed you're denouncing today, hyperlink to the government/employer press release, or a media report on it.
    • instead of explaining some obscure work process, legal procedure, hyperlink to a page explaining it. Wikipedia is great for this.
  • avoid rhetoric

  • It's possible that there's a segment of our readership that really likes to hear the talk, but they're likely reading your website to find out what you're going to do about it.
  • Write positive

  • Say what you are going to do, not what you're not going to do.

An example: Repurposing A News Release

Writing to be found

In print we write cute titles, clever teasers that hint at content. The stated purpose of this approach is to "draw the reader in," by "whetting their appetite," or "sparking their interest" or any other cliché you care to add here.

But in print, you, the author, find your audience. You send your newsletter to their homes or stick it in their hands on the shop floor or whatever. They are, to a degree, a captive audience.

On the internet, by contrast, they mostly find you. Of course, as communicators, we all seek to have repeat traffic to our sites. But most people come to your website because they've found you using a search engine. Most likely Google.

People don't search for clichés, adjectives or rhetoric. And when they're looking at search hits they need to be able to tell if the hit they're looking at is the one for them.

Search engine optimization

Search engine optimization is a huge industry composed of equal parts secrecy and chicanery.

On the one hand, the search engine people want to keep their indexing algorithms secret to protect the quality of their results, and maintain the value of their advertising services. On the other hand the SEO gurus aren't about to give up their intel on how the search engines work. Not for free anyway.

However there is publicly available information on how Google works that can improve how well our content does in searches without polluting the index. See: http://www.google.com/webmasters/

Many aspects of Search Engine Optimization are beyond the scope of this workshop: inbound links, domain naming, markup, etc etc. But content does play a critical role in how high up the result page you appear and whether they choose on the link to your page from among the thousands of hits.

Keywords: where and how

A keyword isn't necessarily a keyword. It's more likely (and more effective) if it's actually a phrase or a few words. Mostly nouns. Mostly nouns that explain the who what where and why of your website.

The cognoscienti of SEO are forever discussing Google's constantly morphing search index algorithm, aka what makes Google rate page X over page Y for searches for keyword Z.

Time was when they thought it was all about keyword density: a measure of how many bytes of content are the keywords searched versus how many bytes of information are in the page in total.

Lately though, a lot of them are abandoning this theory. This is a relief to the writers in the crowd, because writing to boost keyword density would make your average copy editor scream.

Instead, current thinking suggests it's more important to Google to find the keywords in a variety of places: headlines, subheads, the page title, image file names, image captions, directory names, etc.

The process of researching and picking your keywords is a complicated one, fraught with much Googling and politics. Imagine how short your life would be if you argued that the union should start calling child care workers 'baby sitters' instead because that would help your child care pages rank better with Google.

But it does mean you should use old words, not buzz words, slang or your own jargon. It's what people search for.

Search engine result pages

 <title>This will be the header 
for the search result</title>
 <meta name="description" 
content="This will be the blurb text underneath" />

Search engine results

In any given google search listing you'll notice some very neat listings which have meaningful titles, and good summaries beneath of what you can expect to find if you click on the link. In others, not so much.

How do they do that? Two hypertext tags: title and meta description.

Both of these tags are in the head of a web page, and not visible to humans. So we tend to forget about them. If they're not present, Google will make up the search result using whatever it can find by way of a headline, and it will gather the text around the headline to make a description. Sometimes this works, sometimes it doesn't. Indexed PDFs and Word documents produce the worst result listings.

A caveat. Google won't always use your carefully crafted meta description. If the keyword match isn't part of or close in meaning to anything in your meta description, it may choose to highlight the area in your page where it found a keyword match.

Passive voice makes a comeback

Web surfers scan search results too. Say you have a page about grievances. On the page you have chunks of content about carriage, arbitration, and representation. Surfers will find your content more usable if your headings - both the invisible, search engine result kind and the visible human scannable kind - put the keyword as the first word.

And this means (or can mean) using... ulp... passive voice. For example:

Arbitration is where many grievances are resolved by the union

Normally, I would tell you to write this:

The union resolves many grievances at arbitration

Link text

<a href="http://cupe.ca/privatization">CUPE's privatization page</a>

"CUPE's privatization page" is the link text

Link text is the text that people click on when they want to visit a given page.

In hypertext:

<a href="http://cupe.ca/privatization">CUPE's privatization page</a>

"CUPE's privatization page" is the link text. A web site visitor will see it as underlined, and in a different colour.

Use the link text to tell visitors what they ge when they do click. "Click here" is not useful. Well-designed websites use underlining and colour changes to indicate a hyperlink. Web page visitors scanning through your content will be drawn to the differently-styled text and will read it to see if it has any information scent. If it's just "click here" then they have to scan back and then forward to see just what it is they will get if they do decide to "click here".

Meaningful link text will also help improve the target document's relevancy rating in search engines. Google parses any link text it finds in a document, and compares the link text to the target document. If the keywords in the link text are also present in the target document, that target document gets a higher rating for searches for the keywords found in the link text.

Writing scannable, searchable headlines

The most plain, direct, descriptive title possible. Website visitors need to know if the link they're about to click contains the information they seek.

The first gives you no idea of what the news is, nor where it's happening.

The second is better because at least we know what hospital we're talking about. Location is a keyword and putting keywords into your headline is good.

The third drops the clichés and the rhetoric and gets to the actual news while maintaining the location keyword and adding two more: strike and vote.

Think of who, what, where, and why. When isn't so useful. Write the words down and string them together to make a headline.

Real life example

What does the McGuinty government have against home care workers? Deliberate exclusion from new temp worker Bill continues low wages and exploitation of home care workers 168 characters

The first example was a bit a straw dummy insofaras I made it up and gave myself a really easy way to make it brief. The above is a real live headline about a more complex problem.

Pass 1: pick out keywords

That leaves us with:

McGuinty government home care workers exclusion temp worker bill low wages

But that makes no sense.

Pass 2: make it english

McGuinty government's temp worker bill excludes home care workers, guarantees, low wages (88 characters)

Exclusion becomes a verb. You need those. But it's long.

Pass 3: sand it down

Ontario temp worker bill excludes home care, guarantees low wages (68 characters)

Only need to say 'worker' once. Replace 'McGuinty government's' with 'Ontario' because it gives you another keyword and uses fewer characters.

Exercise: concise headline

Web catchy

Further reading

My delicious bookmarks on writing:
http://delicious.com/cmkl/writing