Words: Ashley Rigg
Published: 18th September 2008
Globaledge Search Index
There has always been a lack of reliable data on consumer demand for overseas property. Property portals publish insightful statistics but it is very easy for these results to be influenced by the unique search engine positions of the portal publishing the results.
In
July this year, Google began publishing actual monthly keyword volumes which we believe make it possible (with a combination of hard work and a consistent approach to mining the data) to produce something more valid and reliable than what is currently available.
The objective
Our aim is to create reliable time series data that allows easy comparison of consumer demand for overseas property in different locations across the world.
Launch dates
The Globaledge Search Index will be launched in full in November 2008. We will be publishing interesting insights and snippets before the launch to help generate interest and publicity.
Methodology
Gathering our data set
The Google Keyword Tool is only as good as the data that is entered into it. To get valid results, you need to enter the right search terms and locations. If you miss out “Paris Property” for example there is no guarantee that Google will suggest it when you enter “France property”. This means you need a comprehensive list of locations and property search phrases for every country and region you analyse. To this end, we have had a team of five people working for over six weeks manually gathering information from a number of sources which include:
- Lonely Planet Guides (names of towns, cities, lakes, coasts, tourist destinations)
- Property portals (Kyero search options are a good starting point for Spain, for example)
- RSS feeds from consumer property and real estate newspapers and blogs across the world (good for development names and new launches)
We now have a foundation list of over 20,000 location names which include countries, regions, towns, lakes, coasts, tourist destinations and development names across the world.
Creating phrases to query Google Keyword Tool
Each location is combined with the word “property” and the word “real estate” in the following combinations.
- <Location> properties
- property for sale <Location>
- <Location> real estate
- properties for sale <Location>
- real estate <Location>
- properties for sale in <Location>
- Location> property sales
- Location> property
- property <Location>
- property sale <Location>
- property sales <Location>
- property in <Location>
- properties in <Location>
Querying the Google Keyword Tool
Our system queries the Keyword Tool for search queries above in all countries outside the country being queried ensuring that we only extract exact match phrases. The Google Keyword Tool then comes back with hundreds of keyword phrase suggestions. For example, [spain property] returns [property in spain], [property sales spain], [property spain], [spain real estate] [villas in spain] etc.
Aggregating the data
The monthly search volumes for each phrase are then aggregated into the location categories that you see in the results tables. Our system only chooses phrases that demonstrate some intent to buy overseas property. For example, we include [villas for sale Spain] but not [villas Spain] as many people searching for the latter are looking for vacation rentals. The system also de-duplicates all duplicates.
If two locations exist in the search phrase, the volume is added to the lower location category. For example, [property denia spain] would be added to Denia. When locations exist in multiple countries, like San Pedro (Spain and almost every country in South America), then the results are marked with an * to indicate the possibility of double counting.
Time series data
The process is repeated every month in an identical manner to ensure reliability. We plan to cover 11 regions of the globe.