Imagine you’ve got a flight to catch in an hour.
You jump in a taxi and say “take me to JFK.” The driver approves with a nod and 40 minutes later he pulls into the parking lot of what appears to be an industrial estate instead of an airport.
With only twenty minutes until takeoff, you read the sign on the factory facade with dread:
Jaffa Cakes Factory
“No, I said JFK, not Jaffa Cake. We’re supposed to be at the airport!”
Your shoulders drop. It’s too late now to get there on time; and the non-refundable ticket in your wallet now serves as a costly reminder of the price of miscommunication.
Table of Contents
Natural Language Errors in Aural UI Systems
Most people in that situation would probably blame the taxi driver for incorrectly interpreting your command and causing you to miss the flight.
These kinds of errors would appear to be incompatible with successful business dealings. However, as technology-savvy folk will testify, this can actually be a common occurrence, like asking Siri “take me to Tumblr” but ending up at “tumbler.com”.
Without being told that ‘Tumblr’ is a real word, Siri goes with the most reasonable match in her dictionary – how was she to know it wasn’t what the user was really looking for?
Unlike the taxi driver analogy, it seems the burden of disambiguating the query is placed upon us, the users. This flies in the face of the goals of natural language processing systems: the user shouldn’t have to think about how to phrase what they want, they need only ask naturally.
It’s Not Our Fault
If we want to go to the website for Tumblr (or Flickr, or Svbtle) somehow we must determine how to phrase our queries, lest we get led off to somewhere else.
So do we spell out the brand name, letter by letter? Or perhaps say it phonetically, altering the original pronunciation into some grotesque malformation of its original self?
See if you can phonetically pronounce the following brand names, and think about what it would take to get a computer to recognize them:
When the only way to begin interacting with a brand is to speak their name into a microphone, it better be simple enough to parse correctly by natural language processing systems.
All the usability testing in the world won’t mean a thing if users can’t even access the website.
Choose A Usable Brand Name
Now that the UX community has recognised the concerns of designing great URLs, let’s continue to make our customers’ lives easier. Instead of forcing them to jump through hoops to engage with your brand, make it easy. Here’s how:
Pick One Standard Spelling
If 50% of your users are searching Google for CodeAcademy and the other half are (correctly) searching CodeCademy, then you know you’ve got a problem.
Aside from having to buy two domain names to catch the incorrect spellings, you’re fragmenting your audience into two separate crowds, one of which will be incorrect forever.
Beware Of International Differences
Users from Australia, New Zealand, the UK and Canada would all instinctively navigate to Technicolour.com. While only one letter different from the actual brand name, the results can be disastrous:
Don’t Be ‘Clever’
While Flickr and Tumblr are seen as hip and trendy brand names now, it’s easy to see how problematic it will be to interact with them via aural UI systems in the long term. Be sure to pick a name that both people and machines can assume the spelling.
Just like how all men named Sean have to introduce themselves as ‘my name’s Sean; that’s S-E-A-N, not S-H-A-U-N,’ you don’t want to have to introduce your company, then immediately append its spelling to your sentence.
A usable brand name is one that people can find and recognise, however they come across it.
Only Use Standard Characters
Until Unicode URLs become a thing, URLs are limited to 26 Latin letter symbols, ten numbers and some other silly things like dashes.
If you use anything other than these characters, your location on the internet will essentially be different to your name. That hurts both your credibility and discoverability.
If you have to drop the accents from your letters to fit in a standard ASCII URL, you don’t have a usable brand name. Additionally, suppose a user copies your name (special characters and all) into their browser and adding ‘.com’ to it, guess what’s going to happen?
Don’t Require Dashes
The best brands are simple and unambiguous. If you’re fighting a war against URL readability (businessshop.com vs business-shop.com) then adding a dash in your web location might seem like an easy fix.
Experts Exchange did this, but it hasn’t stemmed the mocking.
Additionally, just like having to spell your name out, necessitating the pronunciation of ‘dash’ in he middle of your brand name instantly destroys its flow.
Consider the difference between ‘A 1 dot com’ and ‘A dash one dot com.’ No longer are you a proper brand, you’ve just relinquished yourselves to the likes of ‘face-book.com’ and ‘cheap-guitars.net.’
Not to mention, a dash in your URL might get you penalised on Google.
Add these three things together and it would appear that needing a dash in your URL is a symptom of a less-than-usable brand name, and something should probably be done about it.
Summary: What Is A Usable Brand Name?
Usable brand names are strong, unique and memorable.
- They are easily pronounceable and are spelled just they way you think they are.
- They have no special characters or awkward spelling, and follow basic language rules that are consistent around the globe.
- They aren’t chosen just because the ‘normally-spelled’ URL was already taken.
- They are timeless and aren’t following a trend.
If you can meet these criteria, your brand will be accessible to all users, regardless of how they connect to you.