Insurance Risk
The value of address level accuracy
This example looks at how geographic information can be used to make more intelligent business decisions, how mapping out data allows you to see patterns and correlations that are not obvious from the data (that contains a geographic element such as an address) in spreadsheets or documents.
Here we will use the scenario of a business that is looking into a list of addresses, either to purchase or insure, however a list does not give the user the full picture. By looking at the postcodes and street names these do not seem to be close to each other, except maybe the two Essex Street addresses. Only when displayed spatially does the real picture begin to appear.
First of all let us look at the postcode points which the addresses are referencing over a base map
Second we can show the postcode boundaries and this shows that there is a postcode district boundary between WC2R and EC4Y running through the area, so although the postcodes on the list maybe different, they are much closer in the real world than the list implies
The red points on this map are the actual addresses from the list and several of them are clustered in the central part of the map, some in the same building or wider contiguous building 'block'. From a risk point of view, this is not good as although initially the list looks 'safe' as the buildings have different postcodes and addresses, in the area those addresses are very closely clustered
In fact the majority of the properties on that list are within 280m of each other, which is very close in terms of risk.
So what risks are there to be concerned about. Well, if there are a number of addresses in the same building, fire is a risk as all of the addresses are at risk if there is a fire in the building.
There is also the risk from an explosion, accidental or intended. If something went off on Essex Street in the centre of the area several of the addresses shown would be impacted to varying degrees.
Here however we are showing the risk of fluvial flooding where 7 of the 19 addresses fall inside the flood polygon itself.
However, it is not uncommon for the postcode that touches inside the flood polygon to be completely included in the risk model, and this increases the total affected from 7 to 11, of the 19.
Lastly we can look at a more accurate approach and use the Building Part data from OS NGD and see the specific buildings that fall inside the flood polygon which brings the addresses impacted down again from 11 to 9.
Further to this you could then look at the building heights to assess if the address you are interested in is on an upper floor, so less at risk of flood, but possibly more of a risk for fire and evacuation.
So, as set out at the start this shows that using geographic information in the analysis of properties for risk allows the user to make more informed, and ultimately better, business decisions.
Further detail about these can be found on the OS website product pages
The logo in the top left of this story map will take you to the main OS website homepage
More story map examples like this one can be found on the More Than Maps site.
