Context is everything

Harry Brignull’s post (Why you shouldn’t rush into a solution too quickly) about my OZIA presentation reminds me that I promised to extract a few of the key messages from the presentation and blog about them myself.

The key message of the presentation was Iterative design alone can’t save us.

As you can probably guess from the main theme, my argument was that we must ground our work in a rich understanding of the context of use, or else we run the risk of creating well meaning rubbish.

I argued that although an essential ingredient in good user centred design, iterative design alone can’t stop us from creating bad products.

To make my point, I cited the highly iterative, but decidedly non-contextual development of the UK Government’s SA80 rifle. The rifle was developed over a number of years and involved much testing at firing ranges.

The problem was that the testing was conducted under far too controlled conditions. The rifles were carefully caried out to the firing range, rather than being dragged around in the real conditions in which they would need to operate. The result being that when the product was ‘launched’ it still had major (**major**) issues. Such as:

• Couldn’t be fired from the left shoulder
• It went off when dropped
• Safety catch would break if the trigger was pulled hard
• Plastic would swell in rain and jam the safety switch on/off
• When running a heavy ammunition magazine would fall out

My argument being if we only conduct iterative design in controlled usability testing labs, using pre-defined tasks then what is to stop our projects going the same way as the SA80?

My full presentation is fairly visual rather than textual, so it may not make sense out of context (chortle).

The information about the SA80 project comes from James Meek’s excellent article, Off target, from a few years back in the Guardian.

Design without context of use

I had cause to dig out and reflect on one of my favourite bad design stories today. It is an oldie, but a goodie, so in case you haven’t heard about it…

Throughout the 70’s, 80’s and 90’s the British Government spent upwards of £470m developing a rifle (the SA80).

The project was plagued with issues, which you can read all about via an excellent article about in The Guardian.

From my perspective some of the most interesting issues are the oversights that seem to have derived from overly focussing on one or two business requirements (mainly accuracy), rather than the real requirements of the end-users (mainly reliability). Some of the highlights are:

• The design of the weapon meant the rifle couldn’t be fired from the left shoulder.
• The gun would go off when dropped.
• The ammunition cartridge would fall out of the rifle if a soldier ran with the weapon.
• Rain affected some of the plastics used in the rifle causing it to occasionally jam.
• Sand would cause the rifle to jam.

The article has a couple of choice quotes that could be taken straight from a user centred design pamphlet:

But this was only discovered after the weapon went into production, because in the early trials soldiers didn’t put full magazines in until they reached the firing range.

and:

As soon as the gun was given to real soldiers to use in real conditions, its weaknesses became apparent.

Greater understanding of the real requirements of real users, rather than the perceived requirements of business users should have avoided all of these problems. Just goes to show that information technology isn’t the only industry that can make unholy cock-ups!