I want to…

write the history of

“I want to…”

“I want to…” – a reasonably common piece of UX furniture that provides a little bit of guidance for a user looking for what to do next.

It’s one of those things that’s seemingly been around online since time immemorial – which, in Internet years, is roughly since 1999.

But I believe it does have a history, and I thought it might be interesting to note it down, before it’s lost forever to digital posterity— because there are a handful of useful lessons that can be learnt from its speedy rise, precipitous fall, and slow resurrection into its current form.

And also because I had a hand in its genesis, and feel somewhat responsible for this small part of the UX language we take for granted today.

I know that someone, somewhere is going to prove me wrong – that there were precursors, and I should take no credit (or deserve no blame) – but this is how I remember it. 

Microsoft on the Web

Microsoft launched its first website in 1994.

Amusing ironists that they are, they’ve kept a version of the page live to this day.

As you can see, it took the whole “cyberspace” thing to heart – in a really 1994 kind of a way:

There was no digital style-guide, and no constraints on what you could post to Microsoft.com — the Microsoft website was what they refer to these days as “an ungoverned space”.

Which made the whole site a bit of a mess.

This couldn’t last, and within a year or two, Microsoft had put in place “Pubwiz” – a central tool to publish to Microsoft.com, which acted as a gateway for the code that could be posted to the site.

If you didn’t follow the rules, and include the right code, you couldn’t get your content up there.

Toolbar 1

But to get something published, all you had to do was include a small piece of code that presented a little black strip of options at the very top of the page.

This had a standardised form – with some variation allowed for different countries – and created a modicum of uniformity across the site.

But that was it – one little black strip aside – you could make your pages look like anything you wanted.

And people did.

Microsoft in the UK.

Microsoft UK was one of AKQA’s first major clients in the late 90’s – and their very first work for them duly had Toolbar 1.

An example Microsoft UK Home Page design from January 1998:

This followed the brand guidelines, so was notionally on-brand.

It was the 90’s — let’s just leave it at that.

Microsoft UK site design

Then, as now, AKQA were no slackers, and they knew that this lack of a digital style-guide created online design chaos – no two pages had to look even slightly alike.

So they suggested a consistent page design for the entire Microsoft UK site – a single structure with flexible options which could be adapted to the different content, products and audiences that Microsoft wished to address in the UK.

And it worked pretty well – here’s the only screenshot I can find of what this looked like, from mid-1998 (thanks Wayback machine!):

All of which was solid digital design thinking, and which helped secure AKQA’s appointment as strategic lead digital agency for Microsoft UK.

Toolbar 2

AKQA had shown that a flexible but consistent UI would work for Microsoft, and that the entire site would benefit from more consistency.

And, slowly but surely, Microsoft Corp took this on board and began to take back the UI of Microsoft.com.

After having ensured that every site had Toolbar 1 – they now came up with Toolbar 2 – a header for the page, this time including a small graphic bounded by two sets of navigation.

The format and options of Toolbar 1 had been determined by Corp, and that remained the case for top right nav in Toolbar 2 – but the bottom navigation options were up for grabs, and you could put whatever you wanted in there.

What’s more, these new navigation options had – can you believe it! – DHTML drop-downs, meaning you could have several options appear underneath one of these headings.

Licensing

I started at AKQA on the 28th of October 1998 as their first strategist, assigned to the Microsoft UK account.

And the first big project I worked on was the roll-out of Toolbar 2 – I had to figure out what to do with this new format, and how it should be applied to the UK site.

This was well before the discipline of UI or UX design had been formalized in digital agencies, and well before anyone at AKQA had a title like “User Experience Designer” or “Information Architect”.

So, we picked a reasonably innocuous part of the site – that concerning Microsoft’s Software Licensing options – and had a jolly good think about how this new Toolbar could be applied to it.

The fact that there were drop-downs in Toolbar 2 meant three things:

1. We could put all the site navigation options under drop-down headings

So

2. That there was no need for a left-hand navigation if we didn’t want one

But

3. We couldn’t have too many headings, as we’d run out of space on the right-hand side of the Toolbar

So, the first thing I did was try to fit all the site content under a set of headings that would work as drop-downs.

And I could not make it work – the nature of the site’s content meant that there were always a few bits left over that didn’t fit too neatly anywhere.

And I couldn’t just keep the list of headings going on – we were still notionally designing for 640x480 – so I needed to compress several options under one short, sweet, generic heading.

Something that could cover a multitude of sins.

“I want to…”

Which is how “I want to…” was born – I needed a place to put a grab-bag of stuff, and that seemed like a nice “customer-focused” way to do it.

Conor Gunn, now Head of Ecommerce for TJX Europe, was then AKQA Account Lead on the Microsoft UK account. He and I toyed with a couple of different options for the phrasing – “I want to…”, “I’d like to…”, “I need to…”.

We thought “I want to…” was simplest, softest and not too far away from Microsoft’s tag line “Where do you want to go today?”

“How kind of you to ask, I want to…”

So that was it.

Necessity mothered invention, it fit with the brand, and had a nice air of “customer-focused” about it.

So the UK Licensing site was the first to adopt this new Toolbar design – with this equally new “I want to…” drop-down navigation option – around February 1999.

Truth be told, the overall navigation scheme did not work too well – the DHTML drop-downs didn’t work in Netscape, people weren’t familiar with them as an interaction method, and you couldn’t immediately see what was under the options.

“I want to…” two

Which meant that the UX needed a little re-jigging.

The left-hand navigation would present most of the site-specific navigation options – making them reasonably clear and visible.

The Toolbar 2 drop-down would then retain a short set of common options, one of which would be the “I want to…” drop-down – since it was still felt to be a smart way to meet customers’ needs – gathering up all the cool stuff under one heading.

So every page on Microsoft UK had the “I want to…” drop down, with the content tailored to the sub-site it was on.

Here’s how it looked on the updated Licensing site:

“I want to…” too

And we thought not too much more about it.

Until Jason Warnes, then Account Manager for BMW at AKQA, now MD at TH_NK, came back from a presentation that BMW had given where they’d cited this “I want to…” drop-down as a customer-focused innovation that they thought highly of.

Which was nice.

And then we noticed it popping up everywhere online.

At this remove it’s hard to recall precisely where, but I’m pretty sure it was a key part of a Mobile company’s site design (Vodafone maybe?), and a bank’s (Barclays?), plus a handful of other sites.

And site owners at Microsoft Corp went *crazy* for it – it started appearing all over Microsoft.com– on the Technet and MSDN sites, their intranet, all over the shop.

Though we’d certainly not planned it, we’d created a little bit of “a thing”.

“I want to…” three

So we kept at it – even when the next version of Toolbar 2 came out in 2001, which restricted the options in the lower drop-down, and added left-hand navigation, and a page footer.

We moved “I want to…” to the main page content, with a lot of other Microsoft.com site owners following our lead.

But the cracks were starting to show.

I don’t want to anymore…

The problem was twofold.

First, as a drop-down, it was impossible to know what was under this heading; how could you be sure it would have what you needed?

Second, since there were no restrictions on what you could place in the drop-down, and since people might be coming to your site for all sorts of reasons, site owners would put just about anything, and sometimes almost everything, in it.

Which meant that when you clicked on the drop-down you’d get an unstructured list of 30 things to do (all in very small font).

We went over to Corp to speak to them about site design, and their UI researchers were pretty clear that, as implemented, the design simply wasn’t working – they’d done a lot of testing, and people weren’t getting what they needed from “I want to…”

Consequently, it would not be supported as Microsoft Corp finally began to roll-out the full-page site design templates c.2003 that everyone would have to use.

So, “I want to…” disappeared from the Microsoft website.

And it appears, the same problems dogged implementations elsewhere, and this little digital navigation trick more or less disappeared from the Internet.

“I want to…” once again

So “I want to…” was gone.

But was not necessarily forgotten, because there was and is something compelling about the first person voice as a way to help navigate options and make choices.

The problem with the design as originally implemented was that its options were hidden, and there were often simply too many of them.

Make the options visible, and limit them to the small handful of things that people actually want to do, and the method might have some utility.

And, indeed, that was what happened – in a few places here and there, site designers took the concept, made it visible, short and to the point, and, by doing so, made it work.

Bank of America, for example, has long included an “I want to…” section as part of its online banking design.

And other brands often use this as a key part of a site’s home page design and overall navigation.

For example, Tork, the tissue, paper towel and soap company.

Or the automotive data company Eurotax Glass:

And Volkswagen UK has long used a variation on this format as the main form of home page navigation:

Even Microsoft couldn’t resist – as recently 2010, they made “I want to…” part of the design of their Office Home page:

“I want to…” wrap it up

So, “I want to…” does have a niche and can be a helpful part of site design, even some decade and a half after its first inception and application – because that first person voice does help frame the decision process.

The trick was to make it visible and clear, and its options few and simple.

Which is consistent with what we know about people – they are intensely visual, and they prefer fewer choices.

By which we might appreciate that we should:

1. Not be afraid to just see what sticks – We were in a bind, and “I want to…” helped us out of it. The idea was part-user-led-design and part-post-rationalisation; in truth we didn’t know if or how it would work, so we just gave it a try and it turned out to have some legs

2. Stay true to your intent – We conceived this as customer-focused, and we should have ensured that it stayed so, by setting limits on what you could add in there – too many options diluted the customer-focused nature of the proposition

3. Change up the format – As a drop-down it didn’t work, but make it visible with few options, and you’re not just giving people choice, you’re helping them choose

4. Recycle – Stuck? Take a look back and see if there is something you can re-use or recycle in a new context, or in a new way. There is, after all, nothing new under the sun

5. Understand choice – I.e. really understand how people think, not just how you think they might. Study and internalize the psychology of choice – there’s a lot of research out there which should be able to help you shortcut parts of the design process and, unlike us, leave a little less to chance

So that’s it – a little bit of history about a little bit of UX design and a little bit of food for thought about the interaction design process.

I want to…

Go and have a cup of tea… and wait for the emails telling me how wrong I am, and that someone else came up with all this first…