Writing for the web

by | Nov 2, 2018 | Behind the Scenes | 1 comment

We’ve done a lot of work over the past 12 months to empower our colleagues from outside of the Marketing and Communications Team to contribute content to the Library website. Something that our content contributors often find tricky is producing text that works for them on a web page.

It makes sense; the way we’re all taught to write at school and university, writing up a report, the books and articles we read are all in a long-form essay format. The problem is, when we read most web pages, our brain does something different without us even knowing.

Example of the “F pattern” we see in eye-tracking experiments

Example of the “F pattern” we see in eye-tracking experiments

In order to write good content for the web, we first need to understand a bit about how people read on the web.

The crux of the matter is when we’re looking at websites, we’re trying to find something specific. The classic example of this is the “F pattern” we see in eye-tracking experiments where users are scanning the page:

Scanning and content hierarchy

The sad truth is, the majority of the content on a page won’t get read. It’s not to do with your content being boring or interesting, it’s just to do with the way the user is hunting out information that is useful to them.

We call this “scanning” and the hierarchy of importance to the user is something like this:

• The main heading
• Pictures and their captions
• Subheadings
• Paragraphs (especially the starts)
• Sentences (especially the starts and ends)
• Bulleted lists (like this one)
• Hyperlinks
• Words (eventually)

In the “F pattern” we can see that people read a little bit, then their eye drops down the page while they look for something else to latch on to, read a little of that, then continue to move down the page.

How can we use this to our advantage?

Once we realise how people are looking at a page, we just need to tailor our content to help the user along with their scanning.

We have a word for this – it’s called ‘scannability’. The easier a page is to scan, the better it will perform.

All we’re doing is making it easy for our customers to find what they’re looking for.

Begin with the conclusion

The “inverted pyramid” which aids reader comprehension.

The “inverted pyramid” which aids reader comprehension.

This sounds bizarre, but it follows a journalistic technique called the “inverted pyramid” which aids reader comprehension.

Start with the “need to know”: who, what, where, when, why and how.

Finish with the “nice to know”: supporting information and related information.


Break up your paragraphs, and put your key points at the beginning of them.

If a sentence doesn’t directly relate to the previous one, put it into a new paragraph. Paragraphs should certainly not be more than two or three sentences long.

Headings and subheadings

These are important signposts for the reader, and they should describe the content they sit above.

Try to be as succinct as possible, so that they can be read and digested quickly. Ideally, the reader can tell from a heading whether the following content is useful to them or not. If not, they can move on to the next heading, and so on.

Headings are also important for Search Engine Optimisation (SEO). The top heading (h1) on a page is what the user will see if your page comes up in a search. Search Engines themselves use headings as a key factor in deciding how relevant your page is to the search terms.


Try to keep sentences short. Ideally, sentences should be no more than 25 words. Any longer than this and the reader cannot scan them for content and will get lost.

Try to put the key points at the starts and ends of sentences, this is where the reader will look for context.

Bulleted lists

If you have very concise points, try listing them rather than stringing them into sentences or separating them with semi-colons.


Hyperlinks are important to get right for accessibility, but they’re useful markers for scanning for all users. Links should be a short summary of the page or content you’re sending them to.

This is especially true if you’ve anticipated traffic to your page in order to get to some other content buried deeper in the website.

Your hyperlinks should be readable without context; if you read the hyperlinks on your page (and only the hyperlinks) do they still make sense? If half your hyperlinks say “click here” or they’re just a URL, you’re in trouble!

Pictures and captions

Pictures are great. They make the page look a lot less intimidating, and they draw the user’s eye around the page. While we have their attention, we should try and do something with it.

If a picture attracts the reader’s eye, the next thing they will do is read an accompanying caption. The caption should serve to enhance the content on the rest of the page. Use it to pick out an important bit of information you really want people to know.


Use plain English, avoid jargon. Anything that causes the reader to stop and have to process what you mean is bad for scannability.

Using plain English also really helps with SEO. People use simple terms when they’re searching for content, so if your language matches theirs, you are more likely to show up in their search results.

Watch out for wasted words, especially adverbs like actually, quite, really, basically.


To sum up, people read differently on a web page than in a book. They need your help to find what they’re looking for.

Know your audience. If you can predict the sorts of things you think they’re looking for, make sure they aren’t hidden away in the middle of a long paragraph! Use headings, pictures and lists to make your key points stand out.

Don’t try to be too clever. Using complicated language, jargon or long sections of prose will make your page difficult to search for, and even harder to scan.

Jonathan Hogg, Digital Marketing Coordinator

1 Comment

  1. Anna Hughes

    Really interesting read Jonathan, I’ve passed it on to the AE team.


Submit a Comment

Your email address will not be published. Required fields are marked *