AI for Staffing: Get Ahead with Your Data

Michael Isinger Products & Technology

AI is becoming more prevalent within the ATS space – providing recruiters better intelligence with recommended decisions or tasks and filtering out the worst candidates or progressing the best candidates through your workflow based on algorithms against your data.

But these algorithms are only as good as the data present. The next big challenge I see is going to be around getting the right data for AI to be effective.

In my opinion, there are 4 evaluations that most recruiters make when reviewing or connecting with a candidate to determine whether they would be the best fit for a role:

  1. Hard Skills – Does the candidate have the define-able hard skills or education that are required or ideal for the role. This is the easiest evaluation because this data exists on the candidate resume without any natural language processing outside of evaluation of years of experience to determine senior vs junior roles. It’s important to start tracking scoring of applicants as they come into a system to determine their hard skill value to the role or organization and recommending or automating action from this score.
  2. Soft Skills – This is one of the more difficult evaluations to be done programmatically, to evaluate soft skills such as how are they with people, have they worked in the industry before, would they be a culture fit, etc. A recruiter can get a sense of soft skills from a resume but usually, requires a screening call or communication to determine these soft skill evaluations. One of the most interesting ways I’ve seen to have a system filter based on soft skills is to create unique weighted questionnaires around culture, career goals, etc against each job requisition and score based on defined answers. Some will send out a questionnaire ahead of their phone screen to gets answers back that can be scored. These are current options available to start scoring candidates on soft skills with closed questions but we will need to apply natural language processing to this to be able to better determine the fit and scoring on soft skills with open questions. In the meantime, recruiters should be starting to track their soft skill scores against candidates as they evaluate.
  3. Willingness or Potential to Move – How engaged are they with you as an organization and are there signs of unrest or unhappiness in their current position? This is especially important for the candidates that are passive you have sourced. For these candidates, it’s important to start engaging them and nurturing them in your talent pipeline for when they are ready to move using a candidate relationship management tool. But how do we evaluate programmatically being ready to move? There are 2 arenas – amount and type of activity recorded in the system to determine an engagement score and social media to determine the level of activity which may be a sign that a candidate may be ready to move. So it’s important to ensure all of your activity, engagement campaigns, and social media data is tracked within your system.
  4. What’s the cheapest qualified candidate I can get? – With nearly 15 million temp or contract employees being hired by staffing companies each year – the maximizing of gross margin is one of the top ways for many staffing companies to drive revenue growth. So a recruiter will evaluate among all the above 3 criteria, who can they get that is cheapest but still qualified and will ensure customer satisfaction with your clients. This means that the candidate may not be the best-scored candidate for hard skills and/or soft skills. The challenge is being able to evaluate the entire pool of candidates reviewed to determine the right fit between best qualified and cheapest of those – and how do you tell what is the right point between these? I believe this will require AI looking at a company’s historical data to determine the kind of scores needed to get a good customer satisfaction score to be able to pick the cheapest candidate among those. This is also something that your recruiters may not be able to evaluate effectively now – especially with higher volume positions – and I believe will become a defining value for Artificial Intelligence with staffing companies in the future. This final challenge and value is what drives the importance of your data – not only your current but your historical data.

In summary, I believe the key to maximizing AI effectiveness in the recruitment space as the technologies continue to develop is to ensure you start tracking the data now so it can be leveraged when that technology is available – instead of waiting for the technology to be available.

That means you want to ensure the following kind of data is being tracked in your system, either automatically or manually:

  1. Hard Skill scores from evaluations between job requisitions and resumes.
  2. Ensuring you are capturing data points around soft skills, whether driven by weighted questionnaires, open questionnaires for eventual natural language processing, or manual scoring.
  3. Recording all candidate activity in the system – whether passive or active – so engagement scores can be built to determine evaluation on a willingness to move.
  4. Tracking social media profiles against each candidate.
  5. Track net promoter scores from your clients and candidates for each placement to evaluate client satisfaction.
  6. Track desired and submitted pay rates and salaries for every candidate.

Watch  webinar on how AI Scoring can impact the recruiter: https://register.gotowebinar.com/register/4857196354018416899

If you have any questions, please send them my way: michael.isinger@jobscience.com