Monday, May 4, 2015

4 billion people without addresses ….balderdash!

When it comes to (postal/street) addresses there’s a figure that’s bandied about: 4 billion of the Earth’s inhabitants do not have a proper address. I don’t know where this figure comes from, but it’s everywhere.  It’s one of those figures that has been repeated often enough to be accepted as fact and published and republished without further thought of checking. 

Here, for example. And here. And here.

What I do know, though, is that it’s highly inaccurate.  Let’s break this myth and find a better figure.

4 billion. It’s one of those digestible numbers that gives the impression that it’s been plucked from thin air, a number that’s high enough to be newsworthy and difficult to check up on.  Like newspaper reports that 4 billion people watch the final of the world cup. We know this figure hasn’t been based on any measurements, and when you think about it, you know it can’t be true. We know that when a company claims they made $ 1 billion profit, that’s not the exact number. Their computers have the exact number, down to the last cent, even if they don’t tell their shareholders or (too often) the tax authorities. I much prefer more accurate numbers – they’re not only likely to be more accurate, they look more realistic too!

What’s a proper address? Whether you like it or not, “the yellow house opposite the bus stop in the street of the ladies of the night, Lagos” is a perfectly valid address. You don’t need street names or building numbers to have a “proper” address – just ask anybody in countries like Japan and South Korea, where streets are generally not named but where addresses exist, based on areas and number sequences.  And if addresses need to be valid and fit for purpose … for whom? And for use by whom?  The address in Lagos above will work fine to help people find the building concerned, but might be less efficient if used for emergency services, utilities or the tax authorities.  I’m not by any means trying to negate the idea that everybody needs addresses – I’ve been working with addresses for over 20 years and am evangelical about them – but what wouldn’t work as an address for one person might be perfectly understandable and usable for another. If we’re going to quote numbers about people without usable addresses, one first needs to define what an address is and how, and for whom, it becomes usable.

To do that, though, we need to find a basis figure. Let’s try to find a better number to work with. if 4 billion’s not right, how many is it? It’s actually a hard number to pin down.

Well, let’s make some assumptions.  A hefty number of people live in shanty towns and unplanned slum areas in and near large urban areas.  Because these are unplanned they are largely (though by no means completely) address-less because, even in countries with (postal/street) address systems, the authorities haven’t (always) introduced infrastructure to those areas.  This is a sweeping generalisation – many shanty areas are well enough established to have street names and building numbers.  But let’s go with this. The best figure I can find for the number of people living in these shanty towns is that it is one sixth of the world’s population, which at the time of writing is estimated at 7 310 125 276. So, if we’re generous and assume not one of those people has a usable postal street address, that’s 1 218 354 213 people without an address.

1.218 billion.

Next, let’s look at countries without street postal deliveries – something I know plenty about. Unlike this reporter, you can’t assume there are 4 billion people without a postman.  There are countries (about 22) without street-level postal deliveries. Not having street-level postal deliveries does not mean that there are no street addresses in a country. In fact, countries with highly developed address systems, such as the United States and Canada have no street-level deliveries in certain areas. Not having postal services may be indicative of problems within a country, other than its address system. But If we add up the populations of all the countries without any street-level deliveries, we get 509 650 000 (rounded slightly to save my calculator finger some work).

0.509 billion.

Not 4 billion. 

Some of these people will already have been counted as shanty town dwellers, but let’s be generous and include them all.  Add these to the shanty town dwellers and we get:

1 728 004 213.

Let’s say 1.8 billion. Let’s assume I’m missing a whole bunch of people without addresses somewhere and you want a more digestible number – let’s make it 2 billion.  We’re being generous.

It’s still not 4 billion.

Let’s work the other way. If I start with the world’s population and start subtracting the populations of countries which I know how full street address systems, it takes no time at all to get well below that 4 billion mark even excluding the one sixth in some countries we assume live in shanties without addresses.


Yes, I approve heartily of the campaigns to provide addressing for everyone. Read some of the marketing material of companies creating universal geospatial coding systems, and you’d believe that we all have an address already.  But I am absolutely not a fan of these coding systems – people require a human readable and understandable street address, not a code. Let’s get everybody an address, but let’s work on the basis of accurate numbers. Let’s get rid of this 4 billion myth once and for all.  2 billion I can live with, and it’s more than enough to warrant our efforts to work towards an address for everybody.

Friday, January 30, 2015

Partial address validation can be easy, cheap and effective!

I was surprised to receive a package today (late ...) from a company in the UK which included in my address the province as NIERDEROSTERREICH (sic), a province of Austria, instead of NIEDERSACHSEN, the province of Germany where I live.

There's no purpose to including a province in a German (or Austrian) address - I'm flummoxed as to why so many German websites ask for it - but the UK company concerned has my province name (correctly) stored in their records. To me this looked like a software error, where the sender had been allowed to choose a province outside the country to which the package was being sent when entering details for the courier service. Parcelforce, though, let me know via Twitter that their software allows free form entry of addresses outside the UK, so the error lies with the sender; and I suppose it's not completely out of the question that somebody in the UK with an unusually high knowledge of European province names had typed the name in incorrectly.

To give them their due, Parcelforce answered all my tweets.When I suggested that they introduce basic validation into their address capture software, they suggested that the expense for validating every address in the world would be prohibitive.

It wouldn't even be possible. But you don't have to go the whole hog, from no validation to full validation. Too many businesses think that way. There is a lot that you can do to test an address which is very easy and very cheap and very effective.  You could, for example, test the basic validity of a postal code - length, allowed digits and characters, format.  You could only allow the entry of a province which is in the country of the addressee.  All this information is easy to obtain online, and easy to program. Partial validation is easy, cheap and effective - no organisation should be scared of trying it.

Parcelforce's parting shot was that responsibility for the collection of the correct address lies with the sender, not with them.  Very likely. But Parcelforce is owned by Royal Mail, a national postal authority, who must understand the importance and value of address validation. Fobbing the blame off onto each small business using their service is a bit lame.

The package arrived (late, as I said, because GLS claimed not to have been able to find my address first attempt and didn't want to bother contacting me to help them - I suppose if you're in a van with eyes front and not wanting to slow down to check building numbers it's as good an excuse as any), but it did highlight again how simple actions to verify basic address elements can be in everybody's interest and not the great drag on resources too many people imagine it to be.