It used to be hard to find talent. Before the internet, you could advertise in newspapers, hope for referrals or get involved in deep, competitive surveillance. These methods form the foundation of the sourcing part of recruiting.
As the internet matured, companies found themselves inundated with electronic applications and resume submissions. The open nature of internet communications made it easy for anyone to apply for any job, whether they were qualified or not. And they did. It wasn’t long before the problem became discovering people who hadn’t yet applied for the job.
Sourcing matured right along with the rest of the recruiting industry. By the late 1990s, finding people by using complex Boolean searches in Google became normal. The entire candidate relationship management wing of recruiting technology began as a way to harness the large volumes of prospective candidates that the sourcing team discovered.
It didn’t hurt that regulatory guidance covered people who actually applied for a job. The potential employees discovered by sourcers are not counted in EEOC tallies until asked to apply. All of a sudden, there were two databases of prospective employees.
Of course, LinkedIn’s steadily evolving presence chips away at the idea that you need proactive research beyond their vast library of profiles. As long as you are willing to work within the confines of their system, LinkedIn provides the potential to connect to almost any of their 740 million members–sort of. There are some significant defects in LinkedIn’s overall coverage.
In many circles, having a detailed profile on LinkedIn is understood as a strategy of inviting recruiters to get in touch. As a result, “in-demand” professionals tend to keep their profiles slim and difficult to penetrate. Few seriously accomplished or senior workers bother with routine maintenance of their profiles, if they even have one. In truth, the only ones who really keep their LinkedIn profiles up to date and useful are people looking to change jobs, who are the same people who used to (and sometimes still do) flood recruiters with resumes.
The gap between what’s knowable and what’s in LinkedIn is fertile ground. Over the past dozen years, tens of companies have emerged, struggled and ultimately disappeared with products that cover the delta. There have been a number of mergers and acquisitions that ultimately fizzled out.
Sourcers using complex Boolean strings definitely uncover better and more interesting candidates than LinkedIn can provide. But the work is bespoke, cumbersome, hard to measure and even harder to manage. Taken together, the twin problems in sourcing are acquiring current data about people and making the search process easier and more effective.
In today’s market, there are about 20 companies that provide some sort of sourcing automation. They range in scope from attempts to control workflow to elaborate conversational bots. They tend to get wrapped around the axle of customer requirements caused by going to market too early. Almost none of them stick to solving the real problems.
The heart of the matter will always be two complementary initiatives. The first is building a comprehensive, current, accurate, complete and well-indexed database of all available talent. The second involves developing a search interface and architecture that unearths the right prospects in the shortest amount of time. Both projects involve specialized usage of various AI techniques.
A Seattle-based company, SeekOut, is researching, building and delivering this core capability. Founded by a team of four seasoned members of the Seattle tech ecosystem, the company manages to stay cash flow-positive while building the business. Focused exclusively on the sourcing function, they compile the most comprehensive and useful profiles of people in the business.
The company uses publicly available data sets, harnessing a library of 10 dynamic sources of professional information. The real magic happens once the data enters their processing. A complex set of algorithms, models and categorization tools sort, augment and predict. The result are profiles that fill in the gaps of information LinkedIn can’t capture. It’s also updated automatically.
The combination can lead to some pretty amazing accomplishments. For instance, the tool can predict ethnicity, gender, security clearance levels and a host of other useful variables. The predictions have a very high degree of accuracy, making SeekOut a super tool for recruiting diverse, niche and/or specialty talent.
Much of the friction in the recruiting business comes from not knowing enough about a candidate before the investment begins. Current recruiting practice involves winnowing a long list to a small one and then looking to disqualify the people who remain.
It’s pretty obvious that the current method stems from an information asymmetry. Employers know what they want, but have sparse information about who is available. Even LinkedIn, with all of its reach and clout, suffers from the fact that most of its profile stock is in need of additional data.
The SeekOut target client is someone who wants a full picture of the person they are recruiting before they start. SeekOut provides the most information-rich profiles currently available. The cost savings begin with reducing the field to high-quality candidates at the beginning of the hiring process.
In its earliest days, sourcing was more like looking for needles in a haystack than ever finding the right people. SeekOut makes a very different future possible. The more specific you can be about the employees you want to find, the easier it’s becoming to identify them.