By Chimni Project Lead - Nigel Walley

sdfI recently signed up to a new online ‘community’ web site - a UK version of a recent US start-up. Going through the data entry process I put my postcode into the address page. It quickly found a match and then, in the process of identifying the community to allocate me, it proudly declared ‘you live in West Askew’! 

To support this assertion, it produced a Google map showing an area, defined by a coloured overlay, that included my street, and with the name ‘West Askew’ emblazoned over it.   The only problem is that I have lived in this part of London all my life and I have literally never heard the phrase ‘West Askew’ before?  There is no such place.

There is an Askew Road a little way off to the east of our road but the idea of a ‘place’ called West Askew is risible.   Not even the estate agent community have come up with that one, so why should the tech/data community suddenly decide to impose it on me? I suspect that, apart of the lack of basic human understanding that can characterise so many tech start-ups, this site was struggling with something affecting lots of us working with local community web sites - a lack of support for ‘local’ or ‘communities’ from the address/database supplier.  I assume that this site saw no defined local 'place' relevant to my home address and so felt empowered to make one up.

I actually live in a part of West London called Stamford Brook. We locals know that Stamford Brook is a ‘place’ because we have all the attributes to be one.  We have an Underground station - Stamford Brook – and we have a park - Stamford Brook Common – which our local council describes as ‘defining the character of the area’. We have a road and an avenue named after Stamford Brook; we have a Stamford Brook Residents Association and we are a conservation area – the Stamford Brook Conservation Area.  However, according to the Post Office, we don’t exist.  This is because Stamford Brook is not a geographic area that maps neatly onto one of thier postcodes.

There are areas of London that do map neatly onto postcodes.  For instance ‘Bethnal Green’ which is the covered by the E2 postcode district. But Stamford Brook sits at the intersection of three postcode areas, W4, W6 and W12, and is covered by two separate London Boroughs. This means that, when auto-filling addresses online with systems based on the Post Office Address File database (PAF), we are either left with empty fields, or the fields auto-fill to include an adjacent town (Chiswick, Shepherds Bush or Hammersmith) which, as local users, we then have to delete. 

A key component of the Chimni system is delivering web templates for community groups and residents associations (you can see the Stamford Brook version here).  Many people taking part in product research for the web template project told us that they find the 'address fill-out' process so annoying, they prefer to just type in their addresses manually when entering in websites with an auto-look up. The Chimni systems are therefore being built so that they default to a manual-entry screen and merely offer an auto-look up as a secondary validation, not the other way round.

However, we are struggling with the lack of recognition of locally defined areas in the automated systems.  The broader problem for web services being designed around our homes is that, if an area doesn’t have recognition from the Post Office (or anyone who currently runs an address system based on the PAF database) then it doesn’t exist on the web or in any web services that use those databases.  Stamford Brook doesn’t appear on Google maps and, for data purposes, do not exist online.  This means my home, in its current digital guise, is not where I (the homeowner) believes it to be, where my neighbours believe it to be, or where any of us would like it to be described as being in our use of local web services.   This is the kind of nonsense that Chimni was established to resolve. 

In the early Chimni research we found dissatisfaction with the basic, off-the-shelf sign-up and membership systems that local groups had been trying to use – and the problems were mainly around the addressing. Residents associations think about their local area in ways that are not recognised by the Post Office or by software companies.  They like to cut roads by odd vs even number, group them in strange ways and allocate resources to them based on local distinctions.  Most of all, as with Stamford Brook, they want their software to recognise the area they think they represent.

As the data suppliers often say “PAF is not always correct in the eyes of the householder”.  This phenomenon is recognised the Post Office but arrogantly dismissed by the addressing community as a ‘vanity address’ culture. The idea that if a homeowner is unhappy with the way the PAF presents their home, it must be vanity (the example often cited being people in Windsor who don’t want a Slough postal address).   

The PAF and its related data sets are full of anti-consumer nonsense that make no sense to the average homeowner or, like UPRNs and UPDRNs are invisible to them and therefore unusable by them.  Chimni is aiming to democratise these data sets and return control of them to their rightful owner (the UPRN number for my house should be mine).  Even the obvious stuff is made obscure.  In the PAF, there is no address field for ‘City’ even though many people living in Sheffield, Edinburgh or Cardiff and the other great cities of the UK might expect to see one in their address.  (In PAF language a ‘city’ is captured by the cover-all field of ‘Post Town’).  There are also two entities that describe the local town you live in – the amazingly named ‘dependent entity’ and the even better ‘double dependent entity’. But these entities are wholly based on postal routes, not the cultures or communities that are becoming important on the web. This is a triumph of the faceless technical bureaucrat over common sense and the home owner.  Its a world view defined by the needs of a post office sorting machine, not the needs of locals trying to build web services. 

This postal worldview is now being imposed without question on the world of web 3.0, Smart Homes and online ‘home’ services and it is not fit for purpose. Companies like Chimni have to work around the PAF when developing systems, rather than building on it. For developers building websites and services targeted at local groups and communities, this is a problem.  We could call it the ‘West Askew problem’.  Quite frankly, as our homes become the focal point of the next wave of wave innovation, my home (and Stamford Brook) deserve better.

The second problem for web 3.0 companies looking to work with home addresses is that current address look up systems are designed on a one size fits all basis.  In online systems, every homeowner is presented with the same address fields to complete, irrespective of whether they live in a terraced house in a village, or a 25th floor flat in a city tower block. The Chimni system is working on presenting distinct field structures to owners of different home types.  Our system asks them to fill in address fields that make sense to them based on their home type.  Our first question is about that home type – so we can make allowances for address structure, not the other way round. But this focus on home type, and its link to address structure, has to become more widespread.  The ONS breaks this information out in its home surveys, but it is currently unconnected to the address database community. 

Clearly for the emerging property tech start-up community  to take on the might of the Post Office and Ordnance survey would be reckless.  But we are not being properly served by those people currently in control of the datasets being created around our homes and our communities. 

The core principle of the Chimni culture is to take control of my home’s digital life’. To achieve this we have to recognise that the world has changed, and the Post Office is just one stakeholder in an increasingly complex mapping/location/addressing requirement building up on the web in support of our homes.  The underlying data sets that we need to use, have to reflect the way that consumers and homeowners define their homes and their communities, not the other way round.