The Mint Blog
Paco Underhill has spent the last 20 years observing shoppers. Can his insights in Why We Buy - The Science of Shopping help make websites that sell?
When shoppers enter a store they are walking at street pace. They need to slow down before they start browsing properly. Don’t try to sell your most profitable items in this transition zone. For the same reason you shouldn’t put lots of text on your home page. Users need to slow down before they are going to read absorb anything substantial.
Don’t evaluate shop signs sitting in a boardroom. Evaluate them as a shopper would, at an angle, hurrying past, maybe in fading light. This is exactly the same with website design. Genuine users rush past your site at a 70mph blur, only pausing long enough to find the next thing to click on. Over familiarity is the problem when trying to evaluate a site you are involved in producing, so testing it in the dark on a motorway won’t help. Testing on sample users is the solution.
Shopping is a sensual activity. Good stores perform “retail judo” – taking unconscious desires and fulfilling them with a purchase. Most ecommerce is like “a warehouse club on the web”. Some sites are getting better at generating an atmosphere. In my opinion American Apparel does a great job. British e-tailer Figleaves.com aims at a premium market but feels as muchTK Maxx as Selfridges.
Paco sees plenty of scope for boutique e-tailers targeting specialist niches. For instance, he imagines an shop for tall girls who like travelling. It’s too much of a niche for the real world but would make sense online.
The butt brush
Once a couple of people have brushed past your bottom, you’ll probably leave the shop. This phenomenon probably has no online equivalent.
- Discovery USA commission Mint Digital to make Flash game to promote their hit show Overhaulin’.
- Mint redesign BusinessBricks. Matt Weston is forced to clarify his "never outsource your core skills" declaration.
- Mint get over-excited about redesigning their own website. They drop the idea after unfavourable user testing.
The Click #001
IT SOUNDS as exciting as grey cabbage, but user testing is the easiest way to improve your website.
The good news is that testing is really simple. With a bit of common sense, you can’t really do it wrong.
Sit someone (anyone: a neighbour, a toddler, a passer-by) down in front of a computer.
Get them to do something a typical user of your site would do: decide which accountant to use, decide which pie to buy, whatever.
Encourage the tester to keep talking. Ask questions like: when you clicked on that link what were you hoping to see? Which bit of the page draws your attention? What question would you liked answered now?
Scribble down anything you notice.
You can test your website at any time: before, during or after building it. My 2p is that the best times are:
- Before you start the project. Test users on 2 or 3 competitors’ websites (and your own, if you have a site you are redesigning).
- When you have designed the first draft (but before you have spent ages perfecting it).
What will the results be?
1. You’ll find that users struggle with totally different problems to the ones you worried about.
2. You will realise that there are lots of ways to use your site. ‘All web users are unique, and all web use is basically idiosyncratic’ says Steve Krug, writer of an entertaining book on web usability: Don’t Make Me Think.
3. Someone will always say they don’t like the colours. Don’t worry.
4. MOST IMPORTANTLY, you will have a list of obvious ways to improve your site. These are things that would have never occurred to you, even if you had spent a million years staring at your screen.
How many testers do you need? 5 is fantastic. 3 is good. 1 is dangerous (you might get unduly swayed by an unusual user). Check the science: http://www.useit.com/alertbox/20000319.html
Don’t be tempted to email your friends with a rough design asking for feedback. The value of testing is seeing what users do and where they get stuck.
The last words of Seth Godin’s The Big Red Fez (a great little book on web design) are: “It’s not as good as it could be, but if you test it, it will get better".
Most company website are terrible. We aim to do better. Here's how:
1. Tell the truth, but make it fascinating
40 years ago David Ogilvy said this was the key to advertising. It is doubly true on the web. You can't lie (the truth is only a few clicks away). You can't bore (reading on screens is slow and painful). Every word and every pixel must work hard to communicate honestly.
2. Stop worshipping at the golden cow of technology
One reason so many websites are rotten is that companies get sold technology when what they need is help communicating. Communication happens through words, images and design. That is what we focus on.
3. Hire brighter people
Many firms claim they only hire the best. This is an impossible brag. All we promise is that every new employee at Mint Digital will blow our socks off by doing what they do way better than we can.
4. Keep promises, whatever it costs in agony and overtime.
Every one of 25 projects we produced for Channel 4, ITV and Discovery was on time and within budget. We want keep that tradition up.
5. Create a pleasant working environment
Mint Digital employ dedicated fanatics who toil to make every pixel perfect. The best ideas happen away from the glare of a computer screen. The following are non-negotiable:
- every day we stop for lunch and share a meal round a table.
- on the last Friday of every month we take the afternoon off and do something different: sculpt clay, play softball or visit the zoo.
- every member of Mint Digital gets a two week holiday when they will not be contacted
- we will have the biggest tropical fish in the biggest tank in Oval.
6. Test, test, test
Those damn users always find a million ways to misunderstand our brilliant designs. Testing is the only way to make a site make sense.
7. Test, test, test, test, test, test.
Finishing a site is only the beginning. If you are running a website you can access much more useful statistics than a direct marketer. Very few people make use of them. We want to find clients who will let us continually refine their site in the light of web stats and AdWords.
8. Swear loyalty to the three steps
Focus your message, build trust and grow a relationship are our three key steps. We swear blind loyalty to them while searching for steps four, five and six.
9. Have fun
If you don't smile when working for Mint Digital, we'll dock your pay.
Check out Oliva & Toscani's fantastic One Minute Site Manifesto. If you are going to read one thing before getting your site redesigned, check it out. Their 3 steps:
1. Kill the sacred cow of the home page.
2. Kick the brochure idol off the throne.
3. Stop worshipping at the golden cow of technology.