Recruiters need flexibility for software auto matching applicants to clients. However the recruiting software database must contain complete and accurate information on applicants and positions for auto matching to work.
Before we can begin to discuss recruiting software auto-matching, I think it best to define two types of contingency recruitment.
One camp of successful recruiters relies on the ability to start from ‘ground zero’. This recruiting model follows the method taught to me by my mentor and the basics of that model have stayed with me for 25 years. When I think about it, it may be a part of my foundation stones when I write code for our recruitment software application. By the way, my mentor was a very successful recruiter. He would say to me at least once a week, “You know when you can rightfully call yourself a recruiter when you can take a phone book and roll of dimes, go to a pay phone and find the person your client has asked you to find”.
I am not sure if that holds in today’s world of the internet and globalization of the workforce. But I do know that my mentor could still find people with a phone and phone book. The only thing that may change would be to substitute the roll of dimes for a roll of quarters and perhaps give him a laptop.
The other camp of successful recruiters relies more on an established network of contacts in a specific field or profession. These recruiters spend most of their time taking care of their ‘flock of candidates’. They are always adding quality people and dropping off the ones that turn out to be average or below average. Remember, recruiters are not paid handsome commissions to produce the ‘average’ professional.
Auto matching is a tool more for the recruiter who starts from scratch. But as I write this article it comes to me that auto matching is useful in another recruitment model. Some recruiters fill professional positions that are a ‘plug-in’ for the business process. Don’t get me wrong, these individuals are still highly skilled professionals. It is just that their skills are so precisely adapted to a particular function and they are almost interchangeable. Nurses, teachers, airline pilots fit this definition. Most professional nurses, teachers and pilots will probably retire as a nurse, teacher or pilot.
When all is said and done auto matching is a good tool for recruiters in some situations but best be left unused in others. Now that we have somewhat defined the area for the auto matching tool we can describe how it should work. I think auto matching first deals with the fundamental principal of good recruiting, which is to have a good description of the position you are trying to fill. You must also know the personality of the client. If you have a complete, well written job order and your recruiting software database contains accurate information on clients and candidates, good auto matching software should work well when matching candidates to the position. If your recruiting software database contains missing or inaccurate information on clients and applicants then your auto matching software will not work well no matter how good it is.
In order that I may pontificate on auto matching I must make a couple of assumptions.
• The recruiter relies on having access to large bodies of applicants and their resumes.
• The job description contains a great deal of information and details about the position and the skills and qualifications of the person needed to fill it.
Ok, now we are ready to begin as we sit in front of a computer staring at the source code. First and foremost the program must be flexible. It must have the ability to be tweaked and tuned by the recruiter for the particular recruiter, recruiting model and position sought. In other words, auto matching must have many options to change the matching algorithm by changing matching factors and the weight or importance of each factor. Here is a list of matching factors scheduled for use in our recruiting software.
• Job title
• Years of experience
• Keywords and phrases • Education
• Years in one position and years in one company
• History of companies worked for
• Style of company
• Budget responsibility
• Number of people responsible for
• Year of last degree
• Rate of promotion and salary increase
Things you can do with matching factors
• Establish a rank factor or eliminate the factor from the algorithm.
• Establish a plus or minus variance for any or all factors.
• Establish an overall score and rank accordingly.
• Compute keyword density.
I am sure anyone reading this article can think of other factors and weights. The ability of an auto matching feature to add, change and delete factors and weights is critical to a good auto matching feature. I would take it one step further and allow the recruiter some control over how these factors should be identified.