Web writing workshop dec 2016

Editor's Notes

• #3: I used to be a cool scarfie. (And the Wrong Guy). But now I’m a web content nerd: And, worse, an Aucklander. Meetups! Jobs in places big (NAB/BNZ) and small (Jade) Conferences Anyone who’ll listen
• #4: You make web content, too. The content behind otago.ac.nz It’s pretty easy finding an image to represent ‘the internet’ when I’m talking to a room of people who all work on the same site.
• #5: But really, the internet looks more like this to many more people. The audience for your content may well start here. Interesting to think about things from their point of view, because a lot of things change when you do.
• #6: Like, how much “Otago” is there? Even just the “right” Otago site? Most visitors don’t start at the top and work their way through the site navigation. They’re looking for something specific. And what are they even trying to do? They have 958,000 places to go, but they’ll always be doing 1 of 2 things. **Most examples from here will be other sites – mainly other NZ universities**
• #7: FIRST THING THAT PEOPLE MIGHT DO: Find information, possibly from inferior Universities. [CLICK] like course pre-requisites, restrictions, texts, assessment methods, timetables and fees.
• #8: SECOND THING PEOPLE MIGHT DO: Interact – do stuff, Complete a task – enrol, buy something, apply for a scholarship, calculate something like how expensive Massey is, etc.
• #9: Whether interaction or for information, 2 things are true: They’d rather be doing something else. They’re in TASK MODE and just want to get something done. Your job is to help people get something done. You’re distracted by rugby kitten right now. You may not have heard what I just said. You’re here to learn something, but I’ve interrupted you with something I just wanted you to see. So that’s an example of content breaking the user’s flow. Good content is not: Whatever you feel like producing Whatever your boss wants to tell people Whatever fills the template Whatever you can get signed off before 5pm
• #10: Good content is content that helps people get stuff done. Good content is content that tells people stuff they want to know. Let’s start with telling people stuff they want to know. Something else about people on the web: They don’t read. In 2013, Chartbeat set a data scientist loose on slate.com. They looked at things like bounce rates and scroll depth. Across the site, how much of a page do people typically get through before they leave? First spike: No scrolling at all – ‘this is not the article I want’ 50th percentile: Around half the page Right-hand spike: 100% read (beyond that is comments and other post-article stuff). ‘Articles’ that fit on a single screen (eg video) count as fully viewed. People get through photo gallerie. And Slate is usually pretty interesting. This is recreational web-surfing – stuff that people choose to read. People were sharing Slate’s articles on social media without reading them all the way through. This is a Good Result. Anyway, the Chartbeat thing boomed and…
• #11: …Slate subjected themselves to it and wrote this kind of shamed article about what they learned…
• #12: …and so did Fast Company, Huffington Post Digital Trends The Verge… ..basically, everyone had the same way to tell everyone this weird news. But in 2013 this wasn’t even news.
• #13: Jakob Nielson nailed it five years earlier: When people are in task mode, they take in about a quarter of your words. I can honestly say I’ve never read the whole article that this paragraph summarises. I feel like Jakob doesn’t want me to. And it didn’t get as much attention as the ChartBeat stuff because its graphs weren’t as pretty. [CLICK]
• #14: Note the levelling off at 200-300 words. What this means for content producers: Get to the point Get the important stuff out first Be brief if you can (but the levelling off shows this isn’t priority #1) Be interesting Help people find the stuff they’re looking for People aren’t just reading from the top until they get bored. They’re looking for the interesting stuff and skipping the rest.
• #15: Going back to Chartbeat’s analysis of Slate: Remember that 1/3 or readers didn’t scroll at all. And half didn’t get past halfway. BUT look at what happened when Chartbeat graphs out the total TIME spent on Slate pages. The second half of the page, which only the minority of readers even reach, gets more minutes of eyeball time (across all visitors) than anywhere else. If everyone scrolled down at the same rate, that graph would start red and fade out to blue. (Note that the top half of the initial view is totally ignored – banner blindness.) How do people work out where to spend their time on the page?
• #16: The words people read aren’t just from the top until they get bored. Chartbeat used a heatmap to show where people scrolled to. Eye-tracking goes one better and tracks eyes – where on pages do people actually look? (Downside: artificial ‘test’ environment) Left: ‘About us’ page Centre: E-commerce product page Right: Google search results page F-shape: Across the top, down the side (when they bother scrolling), across where something looks like it might be relevant. Eyes focus on: Headings The first few words of a paragraph Things designed to stand out, like calls-to-action (links, buttons) Images (especially faces) And the first heading tells them whether they’re in the right place.
• #17: A QUICK STORY ABOUT HEADINGS The closest café to my office is in the foyer of the NZI Building. Also in the foyer is this seat. It’s just a park bench, but there’s a piece of paper stuck to it.
• #18: ‘About this seat’ Who is going to read that? It’s a seat. Who cares? Boring. Obvious, but also meaningless. Is this information: Objective, like the seat’s measurements? Thankful, e.g. someone donated it? Historical, like the seat used to be somewhere else? A message to workers, like the CEO thinks it represents humility or something? After 8 months, my coffee finally took so long that I resigned myself to reading ‘About this seat’.
• #19: You read in paragraph 2 that there’s actually something great about this seat. It’s made out of old toner cartridges! 100% recycled! And there’s more seats like this out there! That’s frickin’ amazing. No more cutting down trees to make park benches – we’ve got trash! Ugh: ‘About this seat’ Even the ellipsis – the three dots – just trail off like the writer got bored. The heading needs to summarise the best bits: THIS SEAT IS MADE OUT OF YOUR RUBBISH. -- End of my story about headings. What happens once we know what a page is about And we know it’s the page for us?
• #20: Good news! Auckland Uni might be able to help relieve migraines! Who did the research? What sort of research was it? Drug? Genetic? Physiological? New-age healing? What’s next? How long till there’s a new product I can buy? Lots of text with no sub-headings: You can imagine the F-Shape here – very difficult to pull the eye right. Difficult to scan read with few clues from the author. Paragraphs already organise ideas – LABEL them!
• #21: News articles are common offenders, but academic departments are guilty as well. I had to shrink this page a lot to fit it to a single slide.
• #22: Sub-headings are good. They break the text up into chunks, and label them. But some approaches to labelling work better than others. Read through Waikato’s international scholarships. The headings are hard work. Hard for readers to find the parts of the page that are relevant to them. Two examples: [CLICK] Two headings are distinguished by fewer than half their words [CLICK] EVERY heading starts with the same words. F-shaped skimming down the side of the page isn’t going to help.
• #23: 18 subheadings break up this long page Information is ‘chunked’, chunks are labelled. People who’ll only read a quarter of it will find the quarter they care about the most. Obviously, headings are designed to stand out, so using them like this makes a lot of sense. But LINKS are also designed to stand out – e.g. red text Just like eyes are drawn to headings, they’re drawn to links.
• #24: Remember the bounce rate on Slate (the people who didn’t even scroll)? Chances are that they followed a link they should never have clicked. But how does that happen? Every week The Economist summarises its main stories and links to each main article. So Nigeria’s recent election was important enough to lead the ‘Politics this week’ section and have two articles written about it. What are the two different angles in the articles? Which one should I read first? I have to click each link to compare the two. It takes time and effort, and I’m in task mode – time poor, don’t want to have to make an effort. Do not waste your opportunity, while you have someone’s attention on your link text, to describe what they’ll see if they click it. Common, lazy link text: ‘Find out more’; ‘More information’
• #25: “Click here” is probably the laziest. 1.85B hits, and the top two are: ‘Don’t use “click here” as link text’ (W3.org, THE internet standards people) ‘Why your links should never say “click here”’ (Smashing Mag)
• #26: Examples from Victoria: The better call to action is already there on the page! - “scholarship recipients’ adventures” describes the content being linked to - “Read student blogs” tells you what you’ll get By just using those words you improve the link text and drop the word count. Top example: Apostrophe crime. Proof of laziness?
• #27: At Canterbury, ‘Details here’ is baked into the page template for every academic profile. It was probably a quick way to deal with things, but it’s a waste of prominent text. This is Professor Jarg Pettigna’s profile. He’s in Canterbury, and he studies faultlines. Surely there’s some way to make that sound interesting? “Jarg’s latest insights from faultlines around the world”
• #28: Canterbury again: hybrid of useful and not useful. But look where the ‘click here’ crime is: THE MAIN CALL TO ACTION.
• #29: If you’re a post-grad at Auckland you get told to ‘click here’ for multiple calls to action Including the one of the most important thing you’ll do as a post-grad – submit your abstract. (Except there’s a lazy misspelling). Not only that, but the ‘click here’ parts aren’t even the links.
• #30: Good link text tells you what’s behind the link – what you’re going to get. It helps you make a decision. It saves you time. It reduces the effort of working out whether to click or not. No two links sound the same – you don’t need to check to see which one is right.
• #31: And now I come back to you, and what you do. What you write needs to work for people in task mode. It needs to get people interested enough to start looking down the page. It needs to work for skim reading and F-shaped browsing. It needs to take advantage of the design of headings and links. Who are you writing for? Who are they comparing you to? What do people want to know? You need to know your AUDIENCE and your content’s PURPOSE A good opening identifies both these things It calls to the people who you want to attract and it tells them what information you have for them, or what interaction they can take here.
• #32: So I picked a subject, physics, and went to 4 universities’ main pages for attracting new students. In most cases, a bold or sub-heading like text block. Here’s the equivalent on the Otago site. [CLICK] Kind of.
• #33: Ignoring the design of each, here’s the opening text. Remember, the audience has already found their way to the physics page, but might not have seen any other part of the site. “I’m interested in studying physics at University. But that’s a pretty big decision. What university should I choose? And is physics the right major for me?” [CLICK] This is the longest one. It hardly mentions the university, or physics. It doesn’t mention studying physics. It works hard to explain who might be interested… …but you’re only here if you’re already interested. No indication of the information or interaction that will follow if you read on.
• #34: Shorter. Says more about studying physics than the last one, in many fewer words. Hints at what you’ll learn, which is probably some of the information that people want.
• #35: This one is short, and talks about the school and its teachers. It’s good on what sort of experience you might have if you enrol – again, part of the information.
• #36: Another longer one, mainly about the department and what you learn. Good information (if you manage to read it all). == BUT None of them are particularly direct. They’re describing things to the reader, but not addressing the reader. The eventual interaction – ENROL, APPLY – isn’t there. But it’s not about stuffing things down throats, either. Going back through them, here’s my attempt to distil the message.
• #38: The audience is identified by curiosity. The purpose (hidden in the original) is to get people to want to study here. Name-dropping Sir Ernest can’t hurt.
• #40: This is a subtler re-write, recasting the good description of physics to link it to the page’s purpose – getting people to enrol.
• #42: Put the reader in the action – rather than talking about ‘our students’, talk about becoming one. So long as the page design name the school clearly enough, you can even drop the word ‘physics’ completely.
• #44: Job one: reduce word count. Drop the uni name – surely the site already has that taken care of. Drop where we’re looking for planets (people already know what’s around the sun). Simplify what’s left. Reward people who read the whole thing with a bit of human touch at the end. (Remember, we’re trying to appeal to people, not give them their 1st lecture).