Menuwhat3words & Esri Industry uses Technical benefits Explore Esri map ArcGIS Marketplace ArcGIS Setup Get in touch
what3words provides many benefits over other coordinate and address systems.
what3words is a fixed global grid of 57 trillion 3 word addresses each referring to a 3m x 3m square. You don’t have to worry about new or changing street addresses. It is fixed already and will stay that way.
Because the 3 word address system is fixed, and it’s impossible to change it, there is 100% certainty that all instances of the system running everywhere in the world will provide the same 3 word address for the same location.
A 3 word address is accurate to 3 metres, everywhere in the world (except the north pole where the squares get very slightly larger – which most people can live with).
Street addresses vary in accuracy hugely all over the word and are often pretty bad, and you never know how good or bad they are until the inaccuracies cause a problem.
All 3 word addresses will have either been converted from GPS coordinates from a known source, a device’s current location (assuming that the current location was recorded accurately), or from someone moving a map to find a precise spot.
The entire what3words system is around 10MB, small enough to fit within an app, therefore your users can use what3words when online or offline – i.e. they won’t need a data connection to find their current 3 word address or to search for a 3 word address they have been given. This 10MB is of course independent of any offline map data.
You need 16 digits, 2 characters (+/-/N/S/E/W), 2 decimal points, and a space/comma/new line to specify a location to an accuracy of 3 metres using GPS coordinates. That’s great for computers and devices when humans don’t get involved, but humans actually do get involved. Humans get involved collecting coordinates, communicating them to each other, and being the intermediary between devices, groups, social networks, interfaces, and the online-offline world.
When a human is involved, using 3 words is always easier than using those large combinations of numbers and characters. Keep coordinates at the back-end: computers and devices love coordinates, but when you interface with a human either at the search or display stage, using 3 word addresses will be far easier for the people dealing with that information.
The numbers are pretty meaningless to the humans dealing with the coordinates too (does anyone really know how far away, even approximately, 51° 30′ 30.028″ N 0° 7′ 31.796″ W is from 52° 7′ 10.571″ N 1° 38′ 59.467″ W off the top of their head)? Does anyone really know the lat/long for where they live or work?
Using 3 word addresses keeps all the communication easy for the humans, and lets the devices do all the maths at the back-end, and those devices will make light weight of the distance question above, of course.
People who deal with coordinates are plagued with errors in communicating location data – unsurprisingly as coordinates are such a mouthful. Far worse is that when you receive coordinates with errors, you will rarely know that the errors are there: after all a string of numbers, even if the wrong numbers, still checks out as a valid string. The result of this is that a lot of location databases are rife with wrong coordinates which go unnoticed until the very time somebody uses that data.
Because the what3words system has spaced similar-sounding 3 word addresses far apart (index.home.raft is in London, index.home.rafts is in Western Australia) the what3words location-weighted AutoSuggest system is very powerful. what3words’ AutoSuggest-for-typing is live in the what3words iOS & Android apps and will be launching on the website and in the API/SDK shortly.
The 3 word address for each 3m x 3m square is available in multiple languages. New languages are being launched all the time, and our goal is that everyone should be able to use a 3 word address anywhere in the world in their own language. Street addresses when travelling abroad are really hard – for everyone. With what3words you can talk about anywhere in the world, in your language.
3 word addresses are easy to use in SMS, tweets, conversations, phone calls, etc. – any and all of the ways that location is often communicated between people. If there is a computer system on the back-end which processes any of these SMS/tweets etc, the system can look out for our unique 3 word syntax (word.word.word) and auto-convert them into lat/long pairs to integrate into existing back-end systems. If the SMS/tweets are received by a human, the receiver will be able to easily input the words into a device, write them, remember them and pass them on far more easily than with any other form of address.
Because what3words’ AutoSuggest algorithm is so powerful, what3words is ideal for voice input as voice recognition errors can be easily overcome. Where voice input of street addresses into devices suffers from recognition problems using conventional technology, our location-weighted AutoSuggest algorithm provides a close-to-100% return of the user’s voice input. AutoSuggest-for-voice will be launched in early 2016 (until then, you can input using the handset’s voice input feature now of course, but it won’t yet do any clever AutoSuggest).