How do I match the Right Tool to the right job?

Searches don’t happen real-time, they take place in databases, within search engines and under a host of algorithms. Some search engines are free of human intervention while others are a selection of hand picked links. Within search engines, some searches are guided by keywords, some are natural language driven, and yet others are advance intelligence systems.

Search engines, directories, and other search services, often allocate resources efficiently. The efficacy of search engines were proven long ago. These search services index such tremendous amounts of information to boggle the mind, internet researchers theoretically proved this efficiency under fairly stringent assumptions concerning, among other things, the depth of indexing, precision, recall and the nature of the people’s information to be produced, and the degree of competition between services. As the market changes, the differing search mechanisms have more and more efficiently aggregated information to its databases, but the ability to reach results leaves a lot to be desired.

An important insight is that most recruiters do not understand how to search the different search services efficiently yet they often evaluate a search engine as providing bad results. Lack of proper understanding is incompatible with searching efficiency.

In order to evaluate and determine which tools are a match to a job it is necessary to have a sound sourcing strategy in place. Sourcing strategy design provides a coherent framework for analyzing the great variety of search services, with a focus on the problems associated with finding information as it relates to prospective candidates.

A Sourcing strategy allows recruiters to systematically analyze and compare a broad variety of sourcing channels under various situations. By using sourcing strategy design theory, strategy design can go beyond the classical approach, and explicitly model how people are found. Through sourcing analysis, for example, different sourcing channels can be evaluated and their processes managed to deliver greater results.

Proper strategy design shows which mechanisms are optimal for different searchers, say sourcers vs. recruiters. Such insights can be used to better understand sourcing channels that are frequently observed. For example, the sourcing analysis can identify conditions under which sourcing channels maximize the expected return of investment or the conditions that cause sourcing channel failures. Analysis of sourcing mechanisms and the implementation of sourcing strategy design processes allow for detailed characterizations of results when conditions do not hold. Likewise, it enables recruiters to find solutions to the candidate shortage problems.

No resource allocation mechanism can ensure a fully efficient level of sourcing results. Sourcing strategy design permits a precise analysis of sourcing mechanisms. More generally, it can be used to analyze the viability of each individual resource through the measurement and comparison of cost vs. results and other specific recruitment objectives.

Only through the application of sound analytics can the justification be made for the allocation of resources. So what am I saying? It is human nature to believe we are better than we actually are, so rather than rely on impressions formed without merit lets form methodologies to understand the tools well, once mastery of the tools has been established then and only then can you begin to evaluate the effectiveness of each tool and the conditions under which to implement them.

Comment by Moises Lopez on August 13, 2008 at 9:48am: Hi Maren,

As far as I know there is no trusted method as to how to evaluate and use individual search engines. One of the problems is that there are nearly 500,000 search services available, there are directories, search engines, meta-search engines and hybrids and they all have different algorithms and page relevance ranking systems. To top that off, the search engines are constantly trying to change their algorithms to minimize the exposure to spammers and hackers.

How to evalute them is so subjective and based on personal experience that it makes it difficult. I am putting some articles together that address the different areas to evaluate, such as search engine ease of use; speed; flexibility; ease to modify search; relevance of results; current information; number of “dead” links, and the quality of help.

The bottom line is that we have to develop analytics and metrics that work for us and avoid developing opinions based on impressions.