Menu 1

Why I like what3words

When I first came across w3w I had the same reaction many people do – I just didn’t get it.

The idea is pretty simple. Give every 9ft square on the planet three words. So right now I’m at spite.nearly.maps which is a hotel next to SFO. Or hills.boost.oldest is somewhere in England.

There are people out there for whom location is a daily problem. Firefighters trying to find a hydrant. Police trying to find one another. Delivery drivers trying to find the right door. The billions of people who don’t have an address system. For people like that, having some kind of location system makes obvious sense. It’s not hard to help them understand the value of w3w since they spend half their life looking for “the green door half way down on the left.” At 3am. In an emergency.

For most other people we only have this problem occasionally. First you live somewhere where there are addresses. Second you have a phone. But you still have situations where your friend can’t find you in a crowd. Or you go to the doctor and their address is on A Street but the entrance is on 1st Avenue. It doesn’t happen that often so you don’t think of it as a big deal.

But it actually is. If you multiply out all that wasted energy it’s going to end up as billions of dollars of wasted time and gasoline. Just think of every time the secretary has to explain to the new patient that the entrance is on a different street, every day for years. In just one doctors office.

Then in the mapping world, and the computer world, we think this is easy. It’s just latitude and longitude. Or at worst, it’s just a geohash. This is true whenever you have two computers talking about location. If we both have phones, then they can just swap location data off the GPS. Yay.

But here’s the thing. First – where humans are involved you want something simpler. And second geohashes are terrible for actual location when a human needs it.

Geohashes and lat/longs have a bunch of downsides for humans. They’re difficult to remember, bordering impossible. They’re ambiguous – if you get a number wrong then there’s no way to check. And of course everyone has their own hash system.

If you want to design a system that anyone can use and remember, three words is actually a pretty good solution.

  • You can remember three words
    • Shorter words are used near population centers to help this
    • Homonyms are removed (sail, sale)
    • Negative words are removed
  • There’s built in error correction
    • Similar sets of words (e.g. spite.nearly.maps and spit.nearly.maps) are in very different places, so it’s clear there is a problem
  • You can say it over a phone or radio
    • No “b like banana” to spell some strange code out
  • It doesn’t take a lot of cognitive function
    • It’s just three words, not some complicated looking thing that I might mistake a 1 for a 7 or something

If you look at the hashes that have existed in the past they tend to be some code for location like ‘dy32gE@3’. The technologist solution to this problem is “make an algorithm, make it open, we’re done!”

But the value is much less in just making any old algorithm and declaring the problem solved. The actual value here is first, to build a great algorithm for people, not just for PhDs in Mathematics. Second, to market it.

Because that’s where everyone kind of just gave up. There’s this fantasy that if I make something open and put it on the internet that somehow, via magic, people will use and build upon it. But people don’t. Mostly they don’t know it exists, don’t know it’s a problem and will never use your random geo hash thing.

The thing that’s useful and interesting here with w3w is that a lot of thought and time is going in to marketing and PR so people know about the problem and the solution. And if you go look, that is getting a lot of traction all over the place. Because it turns out that everyone from camping magazines to Glastonbury festival staff to people delivering mail in slums needs a location service. And they’re using w3w.

Even Ireland could do with something better than spending 27 million Euro on a code system.

To drill it home; it’s not the technology. The technology is great, and a lot of work goes in to that. Actual real linguists work on building new language versions. Other hashing algorithms I’m sure were great in their way too. But who cares if nobody uses them?

And that’s why w3w is interesting. They have truly great – some of the best I’ve seen – people who’re pushing this solution all over the place to people for whom it’s an actual problem. If you live in a western country with addresses, a functional cell phone network and you work on open mapping… that probably just isn’t you.

The other thing is, it’s a company born of a problem which means there’s a focus. It isn’t about just burning more money and figuring out what business model fad to follow this week. It’s a real problem, experienced and solved by the same people. Chris, one of the co-founders, actually had this problem coordinating deliveries of music equipment and performers to perform live events. He couldn’t get multiple trucks to show up in the same place. Even when talking to them on the phone. This is in England – a relatively developed country with a paved road network, where citizens speak a common language and road signs blanket the country. Where people are highly paid and educated.

And for some reason we consider it normal in such a modern time and place, that two people can’t locate each other occasionally. That’s actually nuts when you think about it. It’s 2016 – really this should not be a problem. And it costs us a lot of time, energy and money.

By contrast imagine you were born in a world where w3w just existed and everyone used it. You’d think your grandparents were pretty dumb to not be able to find each other at a concert. Or not be able to tell an emergency responder where you were in a field or on a big road. Or that delivery drivers actually wandered around trying to find the right door. And yet were able to somehow land rockets on the moon.

Because that’s the world we live in. And maybe by building a common language of location, for humans, we can make it a problem our grandchildren can just laugh at.


  1. weekly 297 | weekly – semanario – hebdo – săptămânal – haftalık – 週刊 – týdeník – edisi - April 1, 2016

    […] Coast ahora también trabaja en What3Words. Para ello existen sus observaciones y las de otros. What3Words intentó un sistema propietario, no libre en adición al […]

Powered by WordPress. Designed by WooThemes