The ORC Hex System
The ORC Partnership www.orcpartnership.com
All the illustrations of this document are © ORC and Crown Copyright 2002 licence no 100038018. All rights reserved, and reproduced here with the authorization of the rights holder.
Prepared by Jacques Paris under the guidance of Martin Hodder
(ORC Data Services Technical Director)
The
main objective was to establish a “permanent” geographic basis for
collecting information and processing it.
The
geographic definition of the “entities” (particularly their size) was to be
able to detect local variations of the variables with a “reasonable”
sensitivity given the main subject of inquiry.
“National”
coverage, starting with United Kingdom
Retail
market analysis.
A
grid of uniform hexagonal geographic areas measuring 500 meters across.
The
grid was created in MapInfo using Mapbasic and then clipped to the UK coastline
again using MapInfo; there
are 1.4 million hexagons (compared to 9300 postal sectors). It is held on a UK basis and it is in Native Mapinfo Format.
In
these maps, only areas with significant (in market analysis terms) data are
shown in color.
1 – Postal sectors
Width of map is about 100Km
2
– Hexagonal geography
Following are different zooms of the same map with hex data, although the second only shows a subset of information
3
– data
The
expenditure data is from a government omnibus survey known as the Family
Expenditure Survey (FES). Household locations come from ORC national Mapping
Agency – Ordnance Survey. The product known as CodePoint gives the number of
delivery points per postcode. ORC incorporates a Lifestyle Survey known as PRIZM
and this is produced by Claritas. The location of retail outlets is captured by
ORC in house data team.
Data
is allocated by FES Region. This allows for different expenditure patterns in
different parts of the country to be exposed. This is then assigned to the hex
grid taking into account the number of households in each hex and the area of
overlap between the FES regions and the hex grid. Finally a weighting is applied
to each hex to reflect the household type by income contained therein. This
information is taken from PRIZM and the actual weighted applied will depend on
the product category. For instance, the weighting may be different for
electrical goods when compared against clothing.
4
– main use
Main
tools: ORC RSPM and SiteSelect
The
RSPM (Retail Sales and Profit Model) is used to predict the sales of a town
centre or an individual retail outlet, based on the availability of spend in the
area and the competitions in the area and the drive time decay curve (that is
how visits drop off the further from the store they are). This information is
then used for working out the viability of a store or if you need more retailing
in a town to stop money “leaking” out from the area.
SiteSelect
will look at then money available and the “lack of competition” and
the drive time decay curve to identify possible new sites for a particular
retail outlet. This can be used as a first pass to see if locating a store in
the area would be possible.
The Model is a combination of MapInfo / MapBasic and MS Access or SQL server depending on the size of the model with supplemented functionality using Visual Basic and Crystal Reports.
Output
format and mapping
The output forecast is input directly into an Excel Spreadsheet where a profit and loss account can be generated, that is if a store model is being generated. All data that can be mapped is mapped. We have paper reports as well giving summaries of spend information, Census data and lifestyle information.
All
mapping output is displayed with MI and there are automatic routines for
producing images for reports and presentations. The
main mapping output is a series of maps that define the catchment’s limits
These limits are identified as Primary, Secondary and Tertiary catchment and
identify where 50, 75 and 90% of the calculated spend actually comes from.
5
–output
thematic
hex geography
(from
RSPM )
(from
Site Select)
