In SmashFlyX, instead of focusing solely on the candidate profile and then querying jobs at runtime, the platform analyzes all jobs in the platform as well as all contact profile data to create clusters that share commonalities and have a high degree of relevance. Every recommendation generated by the platform is given a relevancy score to that contact, and there are several filtering and weighting systems in place to adjust what is considered a good match based on factors like Location, Area of Interest, and Job Title. Each contact has a relevancy weight which is applied globally to all matching processes and if a recommendation falls below that level, it is not included in the recommendation alert. If no jobs match the contact, rather then sending a generic no jobs email, the contact will not receive a message and will wait until the next configured send to generate messages.
More specifically:
- Job Alerts are branded emails sent to contacts with a set of open job recommendations
- Open jobs are recommended to the contacts based on the following criteria:
- Area of Interest is mapped to a category; if the data is not present, this is ignored
- Location Range is configurable. The contact's location and the job location are then compared to see if they fall within this range, based on zip code
- Work History job titles are considered in job recommendations
- Frequency of job alerts is configurable
- Resume information is a large source of keywords and matching, especially job titles.
The more information a contact provides in their profile, the more accurate will be the job alerts received. There is an AI that considers all the text in a profile - job title, resume, preferences, and anything else that is there. Based on that, the contact is placed into a group of job types. When a job is opened, all contacts of that type are connected. The process looks for a percentage match of data to come up with top contacts who should receive an alert.
In SmashFlyX as stated above, there are many configurable elements concerning who gets job alerts, location range (default is 800 miles), frequency of alerts (default is every 15 days), and relevancy (default is 0.4 on scale of 0.1 to 0.9). A higher relevancy setting means more accurate and fewer job alerts. A lower relevancy setting allows great flexibility in the accuracy of the alert with more job alerts sent.
