One of the core concepts of RecruitBot is that YOU can directly influence the type of candidates that you see. RecruitBot takes YOUR specific preferences and uses machine learning to rate and sort your candidate results, tailoring your experience to save you both time and money in recruiting/sourcing.
Although ranking candidates is a straightforward process there are some tips and tricks you should be aware of when ranking candidates for your positions.
Pre-Machine Learning Vs Post-Machine Learning
When you search for candidates, you'll get an initial list of candidates matching your search criteria. This list of candidates is sorted and ranked on something we call the 'resume score', or how complete a profile is when matching your search criteria. Pre-machine learning, you CAN affect the way candidates are ranked by adding in additional 'Nice To Have' criteria, such as skills, job titles, companies, industries, etc. These 'nice to have' search criteria will boost those matching candidates higher up on the list of search results.
Must Have vs. Nice to Have vs. Doesn't Have
After the machine learning has been unlocked, and the candidates are re-sorted based on your preferences, the 'nice to have' selection criteria do NOT affect the sorting of your search results. This is due to your candidate rankings being more important and impactful than RecruitBot attempting to sort candidates based on matching criteria.
How to Rate/Rank Candidates
When you run a search and get a list of candidates for a position, take time to review the candidates profiles and determine how good of a fit they are for your positions. Some of the considerations you'll be looking at when rating candidates are:
Is the candidate a good fit based on their SKILLS and EXPERIENCE?
Does the candidate's JOB TITLE(S) match what you're looking for?
TIP: It's important you do NOT rate a candidate poorly for 'job hopping', if they match all other skill qualifications and job experience. Doing so could confuse the machine learning (essentially saying, "I like this person but they aren't a good match"). If you see RecruitBot ratings of 3-stars for EVERY candidate after unlocking the machine learning, this is an indication that your rankings affected the machine learning. To fix this, you can either remove the position and start over, OR, continue ranking candidates to improve the machine learning.
Once you determine if a candidate is either a GOOD match or a POOR match for your position, you'll rate the candidate using the star rating system.
Rating System
1-Star (⭐) - Not a good match. This candidate does absolutely not have the skills, experience or job title to match your criteria for the position.
2-Stars (⭐⭐) - Also, not a good match. There might be one or two things you like about their resume, but overall not a good match.
3-Stars (⭐⭐⭐) - Uncertain. We call this the 'I don't know' rating. Some of their skills and qualifications might be a match, but overall, it's a 'maybe'. Rating a candidate as 3 stars does NOT affect the machine learning and won't contribute either way to the rating scale.
4-Stars (⭐⭐⭐⭐) - This is a pretty good match for your position. Their combination of skills and experience makes this candidate a good fit, someone you'd want to reach out to and hire.
5-Stars (⭐⭐⭐⭐⭐) - The IDEAL candidate for the position, an exact match for the skills and experience you're looking for. The best possible person you'd love to hire.
Rating Requirements
Once you start rating candidates, you'll see an indicator appear on candidate profiles, letting you know how many more candidates you need to rate in order to 'unlock' RecruitBot's machine learning.
You'll need to provide 10 1-2 star candidates, and 10 4-5 star candidates, to unlock RecruitBot's predictions.
Unlocking the Machine Learning
Once you hit this threshold, RecruitBot's machine learning goes to work, magically resorting your candidate list in the background.
Then to see the re-sorted list of personalized, custom candidate rankings, click the 'See Unlocked Predictions' button.
Before machine learning (no customized rankings, requires a lot of time to individually review each candidate profile):
After machine learning (personalized and custom rankings for you--saving you time in reviewing each candidate):
Now, you can browse the BEST candidates using RecruitBot's recommendations, instead of wasting time reviewing hundreds of irrelevant candidates!
TIP: RecruitBot is always learning, so every additional rating you provide will update the accuracy of our predictions. Keep rating to get to best and most customized matches even faster! Additional rating of candidates can further refine the candidate ratings. Note that this happens in the background when moving to another page of candidates or viewing another search result for candidates.