Friday, February 11, 2005

The use of geographic data

John Sumser in the Feb. 10 edition of the Electronic Recruiting News discussed geographical issues with web Recruiting.

Geography is highly relevant when considering recruitment. Most candidates do not wish to relocate, and research shows that this is going to become more likely as in the future. John discusses using databases and recruitment systems to identify roles within particular areas and whilst this would be a great start there is more to it than that. As he suggests, the postal code (zip code in US) is key.

Geography in recruitment strategies

Understanding the people living in your locality is hugely beneficial when planning expansion, looking at retention & predicting sustainability of workforce etc. The underlying question is 'given our catchment area is this strategy implementable?'

There are a few data sources that you should be considering. I will consider this in the UK perspective but similar data is available for all major economies.

In the UK the official labour market data is provided by Nomis, which is based at the University of Durham. The Nomis site provides lots of great information for free which makes it the perfect starting point. If we look at the area where I grew up, Sefton (link to Nomis information), you can see information about the size of the workforce, the number of people looking for work, the split between male and female, the types of jobs that they are doing etc. This gives you an idea of your total available population.

The second source you should look at is your HR database. The three bits of data you are interested in are:

Postcode of residence (home address)
Postcode of workplace
Postcode when they applied to you - shows if they moved since starting work, indicating relocation. This last bit of data might come from your Application tracking system, however it is less likely that these were linked to your main HR system.

By combining the first two you can understand how far people have to travel to work. You can use this data to understand where it is likely people are going to be living for recruitment purposes. At lower levels recruitment is highly locally based.

You can also look at retention and recruitment difficulties by geography. Given that time spent commuting is a major factor in retention it can be useful to understand the probability of someone resigning based on distance to work.

Another aspect to look at is the distribution of travel distances over time. As the local market becomes exhausted you start to see distances travelled increasing as people are sourced further away. This is a great predictor of recruitment difficulties and can even be used to determine when larger offices should be reduced in size and new centres created.

The final information source you can use is household databases used by marketing companies. This information will often give you a very detailed overview of the types of people who live in an area. If you know for example the distribution of incomes you can get a more accurate view of the relevant labour market size. This is useful when considering using location based advertising (such as posters). One large retailer found that their second most successful advertising channel for the majority of their jobs was bus shelters (after a notice in the branch). Given the marketing information it would be possible to work out which bus routes would be the most successful, or potentially even which bus shelters.

Back to online recruitment, what would help? Well my personal view is that the adoption of the HR-XML format for the resume would provide us with a more accurate way of understanding geographies. Given that the data format tells systems which bits of the resume (or CV) contains the geographic data searching by geography becomes very simple. Searching for candidates on geography becomes much easier.