Language is a virus (from outer space)

September 18, 2008

Hey – still not posting often enough, sorry… Anyway – here’s a little off-the-wall musing (do I ever do any other type of musing?)…

Techniques for “Writing for the web” are long established and well proven.

It’s not hard to do and a few points cover it: write short paragraphs with key words at the beginning of them, use headers and bullet point lists, avoid jargon, abbreviations, hyperbole and buzzwords… that’s about it.

It’s also our most frequently ignored best practice intranet guideline. Everyone thinks they know how to write, and ‘communicators’ seem to be unwilling to change their writing to adapt to the medium of the web – so they write with a style that assumes you are reading on paper, or a glossy brochure.

On our intranet we’ve got very well written guidelines about web writing. Most intranets do.

So why do – otherwise competent – writers not write well? I think the style guide is competing against something more pervasive, the house style is defined by a long tradition of previous writing – something can’t be important if it doesn’t have the whiff of officialese, if it doesn’t feel like previous writings. Bad writing is infectious.

Ironically some of the best writing in our intranet is emerging from social media, from less trained writers. Social media doesn’t encourage such a formal style and it doesn’t have the weight of history. But writing style is still infectious and the style adopted in a social media will spread. It’s best to make sure the good kind of writing spreads, so I suggest one of the first things to add to a wiki is a writing style guide. You may not ‘infect’ all the contributors, but you’ve set the tone and enabled the wiki gardners to do their work.


Its been a while since I did a quiz, so here’s a new one:

It’s about voices overs and so kind of fits with the theme about applying the right style to the context. Comments always welcome on the quizzes, so boast about your high scores in the comment box.

Pop music

My music choice for this post is from Laurie Anderson, indeed, the idea of describing web writing as a communicable disease comes from her song that gives this posting it’s title. I love Laurie Anderson, her Big Science album especially, but if you get the chance to see her you should do so, her performance at the Glasgow Concert hall (mid 90s? ) was one of the best things I’ve ever seen.

Sometime the you tube embed doesn’t work so here’s a link

Content, the once and future king

June 12, 2008

A few years ago (or several hundred in internet time) there was a saying that “content was king”.

The phrase was used in discussions during the web 1.0 boom where to get market share more and more sites were using eye-candy rather than optimising the content ( is the poster child for this). I think history shows that indeed content was king.

A recurring theme of current commentators is that ‘collaboration is king’, google the phrase to find it popping up all over the place.

It’s a misleading phrase though, setting up collaboration in opposition to content (there is only one king!)  misses what is the most valuable aspect of collaboration; it produces better content. That’s certainly true about sites where the collaboration is the content – such as focused social networking sites like Ravelry, or  many wiki’s.

Which leads to looking at the collabaration process from the other side, not how many people are taking part, but how good is the content that it produces? We’ve got a decade’s worth of research that tell us how to assess how good content is, we should apply those techniques to collaboration sites and it will tell us how good the engagement / collaboration is. By that standard, I still say Ravelry is the social media site with all the answers.

The usual, rather sketchy, pop music analogy.

Collaborations between musicians seems to fall into two camps, amazing and terrible. As with websites it’s the outcome (the “content”) that defines how well the collaboration worked. This collaboration is, um, not likely to be an anybody’s list of easy listening favourites. (I did initially look for Bonnie Prince Billy’s brilliant collaboration with Shooglenifty, but its not on Youtube), so here’s Sonic Youth and Lydia lunch making a terrifying racket.

Update: for some reason the video isn’t working, here’s the URL


The Style Council

May 13, 2008

governanceI’ve been thinking quite a bit more about my “governance of the crowd” idea.

I think I differ from most folks who comment on this because I still think social media needs governance. Governance is wrongly perceived as a device to slow down and limit something. We seem to have forgotten why governance became an important issue in the enterprise, it was needed because the free for all that characterised intranets in the late 90s was not efficient and various problems of duplicated, abandoned, orphaned, contradictory, or just badly written sites proliferated.

It may have a radically different publishing model, but we can still judge a wiki in the same way we judge any other web site – is it providing me with the info I need.

But it’s a barrier to entry to force people to go through all the training needed to become a web writing expert, and people are too busy with their day jobs to do it anyway. Much as I would like to, we can’t make everybody read Jakob Neilsen – this isn’t to say that Jakob is usually spot on with what he says.

There is a good model about how this can work already, wikipedia, but most commentators focus in on the amount of content contributors and fail to notice wikipedia has a very well defined set of standards and a good mechanism for allowing the users to flag pages that don’t meet this standards.

The standards can be arbitrary as long as they are consistent. There is no right answer to what font to use on a web page, but picking one and getting everybody to use it helps with the consistency.

With that in mind, here is a quiz on the small details of style.

do it with style quiz

A couple of the answers folks might disagree with (and if you do so, thats what blog comments are for!) the point is its a style guide (in quiz form) that most folks can understand.

And – not to be obvious or anything, but todays pop music link is Paul Wellers post Jam group The Style Council

Now rather overlooked, and often willfully, annoyingly mannered they did some cracking tunes (and a lot of dross). The best tune they did was this one, a stomping Curtis Mayfield homage to breaking down barriers.




Can intranet managers survive on their good looks alone?

April 8, 2008

In the middle of my posting about the ‘boring’ problem, I suggested that difficulty with boring websites in an intranet was that a common solution was to add in some fancy graphics.

However a more common scenario in a large company intranet is to get the design group, or an external agency designer to put the fancy graphics in first, fill the pages with “placeholder text” and ask for sign off / approval.

Usually the design is in powerpoint slides, or static images in pdf format and the approval often boils down to ‘does the sponsoring department think this looks cool’

That’s a bad way to do things.

How to assess a design properly

The important aspects of a design are (in order)

  • Is it functional and efficient
  • Is it accessible
  • Is it extendable
  • Is it maintainable

None of these features are easy to determine from looking at slideware. Often the people originating the design won’t consider these factors, and we end up with a cool looking site which is difficult to maintain looks shabby and ill-fitting after a few tweaks.

Tenuous pop music reference

For the pop music link, here’s an interesting example. Jobraith was a mid 70s glam performer most famous for being a total failure, launched with a huge publicity campaign, the audience stayed away in droves.

It’s not hard to hear why, but it is a fun video for people fond of 70s kitch.

What The Thompson Twins can tell us about intranet management

March 18, 2008

…or how to cheat when trying to assess stuff from design agencies.

In the late 70s – early 80s, I was living in Edinburgh and was a serious music fan, in between occasional studenting I went to see a lot of bands, maybe 3- 4 every week. Many of them were amazing, local bands like Josef K, Fire Engines, Scars, The Associates, The Visitors, The Freeze… and also touring bands like Joy Division, Magazine, Pop Group, Fad Gadget, Birthday Party, The Cramps… I could go on…

But one thing I noticed is that I was useless at spotting which bands would later have success. I once saw a band called The Young Marble Giants who I thought were just gorgeously life-affirmingly special… a band that never troubled the compilers of books of chart placings. They were supported by a band I thought of as wretched, obvious and talentless; The Thompson Twins, soon to be Smash Hits and TOTP regulars.

I’ve been going to see bands for well over 30 years now and I still can’t spot a winner.

I’m better at spotting good intranet designs, but only because I cheat by having a short mental checklist of things to avoid. All intranet managers should have such a checklist for rejecting things that are probably a bad idea. Quite often you will be asked to judge a site based on sketchy details from a design agency. A pdf of a screenshot or a few pages of ‘greeked’ text so that means you have to cut corners and play the odds.

Here’s another comparison: movies – I’m useless at spotting box office success too – if a movie stars Chuck Norris, it’s a safe bet I won’t like it, my checklist then -my heuristic- is that it’s not a good movie if Chuck is in it. Now maybe I’ll be wrong one time in 100 – and I’m not saying that this tells us anything about that movie’s popularity – but it’s a good rule when there are too many good movies and not enough time.

Here then is my checklist:

  1. Will it be easy if I need to add or remove something?
  2. Is there a page heading about one inch down, one inch in?
  3. If I screw up my eyes so I can hardly see, does it look balanced (define balance how you like).

And that’s it really. There are other things I’ll look for in a finished design, valid HTML, semantic markup, headers in sentence case, alt attributes in images etc, but they can be fixed.

The first point is maintainability. You will have to change that design… A real red flag is Flash used as navigation – I know its possible to do this well, just as theoretically it’s possible that someday Chuck Norris makes a good movie. But the flash navigation sites I’ve seen are usually a maintenance headache and an accessibility nightmare.

The page heading thing is basically to cover what I think of as predictability. Every time you click something the most common thought nearly everyone has is ‘ did I click the right thing’ confirm that (by a page title that explains the page) and you’ve won half the battle.

And the screwed up eye thing, well – not sure what the point is, but it seems to work.

Could I have used a checklist to tell if other people would like a Chuck Norris movie – or indeed the Thompson Twins?

Yes and no, I’ve weeded out the dodgy stuff, but I need to now pick the best stuff. The next thing to do is define a “persona” of a person that likes movies with Chuck Norris in them and work out how typical this persona is to my overall audience. For an intranet site I need to think about what typical users know, what their goals are, what else they use. A lot of this stage is less about overall design and more about arrangement and organisation and content creation. Define your personas, check they are accurate representations of your audience, check your personas can achieve their goals. Oh and don’t give them silly names.

But personas and their users stories aren’t a magic bullet see this discussion (via Phillie Casablanca, esp Kerry’s comment – cheers!) you eventually need real people giving you real information. I mentioned the Thompson Twins, I first saw them as a raggle taggle sub scritti politti band… but a year or so later I saw them again, supporting David Bowie. They still sucked, but I could see the rest of the crowd liked them, they had radically changed into a glossy electro-pop confection, they were getting very good (if you like that sort of thing).

The key point here is they evolved after they launched… I don’t know how difficult it is to work out you need to change if even the 50 or so people who saw you supporting Young Marble Giants one rainy night in Dunfermline, and who look like the sort of people to go see almost anything, still think you suck.  But they did work out they should change, and did. Intranet site ‘hand offs’ happen too quickly too often, the site is launched the same day the development guys finish, but you are going to get the most, and best, user information in the first couple of weeks you launch, especially if you don’t have a public beta have a development period that covers the period after launch. Use that period to improve your site, improve you personas and re-write your user stories.

Summary, in case you can’t be bothered reading all that stuff about CHuck Norris and obscure welsh post punk groups.

  1. Assess design quickly based on maintainability, predictability and the screwed up eye thing.
  2. Define user personas and their associated stories, optimise the site for them.
  3. Get it out there and be prepared to fine tune it often during the post-live launch period.

No chance of a Thompson Twins link – sorry, but here’s the Young Marble Giants.
Despite having little success in their lifetime the were pretty influential, Kurt Cobain liked them a lot and Boards of Canada / Ghost Box owe quite a lot to them. This is ‘Final Day’ the sweetest song ever about the end of the world…

Give me a reason to love you

March 16, 2008

Agile software development is a collection of principles about the development of software.

It’s much discussed and has many enthusiasts and detractors, I’m fairly keen, though I am concerned about the usability of the output; that’s been much discussed too.

A long time ago I came across a book which suggested that the principles of software engineering (at the time mainly a linear process called “the waterfall model”) could be applied to website design, and suggested that this be called “media engineering” – I liked the book a lot, but the concept never really caught on.

I guess it didn’t catch on because the majority of people involved in internet / intranet production come from a different background and didn’t want to go through the rigours of formal software engineering processes just to get some web pages created. But also because formal software engineering methods didn’t have a good reputation, they were slow, bureaucratic and failed to deliver, so suggesting they be applied to web site design felt a bit like suggesting in 1969 to the newly formed Led Zeppelin that they wouldn’t be successful unless they followed the techniques of the previous years best selling artist – Engelbert Humperdink.

However software engineering has changed since then and there are many great ideas for ways of working that are ripe for re-use by intranet managers.

Here is my first stab at suggesting a few things an agile intranet manager should do:

Agile intranet management

What will agile intranet management look like?

1. Much shorter design iterations. Have reviews of content every month.

It need not be – indeed it should not be – an extensive analysis, merely a trawl through available information to find out if anyone can improve anything. You want to answer the question: “What problems are people having and how can we fix them?”

2. Don’t talk to users, observe and measure them.

Agile methods use many techniques to remove themselves from the old ‘statement of requirements’ built-in-stone binding agreement that is out of date before the software is finished. You need to measure (automatically if you can) how successfully people can achieve their goal when using your site. You only get that information once you have something out there and working, so get something out there quickly and lead the folks using your site to help you improve it.

For instance, if you have a ‘help’ or a ‘FAQ’ or a ‘feedback’ or a ‘contact us’ link, get rid of them. (A lot of people like FAQs, but I think they are an admission of defeat – the goal is not having people asking questions frequently!) Instead have a link saying ‘Can’t find what you want? Tell us’. This is very similar to the agile software development discussion about not listening to users (via JermolineFND ). Never ask your users (and especially not ‘key stakeholders’ “What do you think?”.

Agile intranet management is going to be important, as more and more social and collaborative techniques make it into the enterprise. The reason for this is that these technologies require the user to actively want to use the features or the system fails, being obsessed with end users and delighting them is the only way they can work. When it comes to “social/web2.0/collaborative” platforms agile software development isn’t agile enough.

Which leads to the choosen pop music to go with this post. The new Portishead album P3 is fantastic, but the title of this posting is a quote from Glory Box from their first album ‘Dummy’, and if you’re searching for a manifesto for the agile intranet manager, then forget the cluetrain, all you need are the lyrics of Glory Box.

“we’re all looking at a different picture,  through this new frame of mind, a thousand flowers could bloom, move over and give us some room”

Give me a reason to love you indeed.

What Althea and Donna can teach us about developing websites.

February 16, 2008

Recently, watching an old clip of the late 70s one-hit wonder ‘Up Town Ranking’, I said to my partner that I preferred Althea to Donna. I just thought she had a nicer voice, especially the ‘ooo’ bit as they repeat the title. My partner usually ignores my pontification about pop music, but this did bring an exasperated retort; “No one else in the world, when hearing Althea and Donna, try to analyse which one is the best, and anyway what do you know about singing?”

It’s true I can’t sing a note, but just as I was framing a reply in my head to prove my knowledge – quite extensive knowledge actually- of 1970s female reggae vocals, (take sides! Janet Kay or Susan Caddogan?), I wisely perceived that this was a debate I would lose by the mere fact I wanted to have it and my partner didn’t.

Asking people what they think of a website leads to a debate you can’t win either.

Within an organisation you are usually asking a senior manager, or key stakeholder ‘what do you think’ – in other words you are asking someone who won’t be an intended user an unstructured question about which, really, they aren’t skilled enough to answer. If you were deploying air conditioning would you show them the blueprints and say ‘well, what do you think?’, or a complex legal policy document and ask them is its OK?

Maybe it’s because people are so familiar with websites they think they are experts.. how hard can it be?

So my only advice is to never get into a position where you are asking an open ended question – sure, ultimately you’ll have to make your decision makers happy, and you should find out what makes them happy and then swallow your pride and make them happy, but as a professional you need to lay the groundwork, explain how the site is brilliantly optimised to deliver business benefits and so that cool gimmick they saw on another site isn’t appropriate. You need to do that before you ask them what they think.

Oh and the next song on the TV was The Proclaimers… I kept quiet.