Recruiting is, at its core, a business built on relationships. The best recruiters, particularly those recruiting tech talent, maintain a Rolodex of high-value candidates, actively trying to stay top of mind for when they might be ready to make a career change. In some cases, that may involve taking a candidate out to lunch, dinner or drinks. Which begs the question: In a network of hundreds if not thousands of candidates, whom should a recruiter take out to lunch? Ideally, the recruiter wants to take out only those individuals who will convert to a placement; i.e., she’d like to avoid wining and dining those candidates who merely want a free lunch, and instead focus on those who are likely to convert from an expense-generating candidate to a revenue-generating placement. That said, one wouldn’t suggest to candidates that your desire to take them to lunch is in part predicated on your estimation of their probability to convert. If only that probabilistic estimation was readily available and accurate …

Enter Human Predictions, a Chicago-based tech startup whose platform aggregates a plethora of data to inform recruiters of the candidates most likely to turn into successful placements, as well as other actionable information. The company was founded by a former staffing executive, Elliott Garms, who was essentially trying to answer that business lunch conundrum for his own tech-focused staffing business. He found that by applying some data to the list of available candidates, his average conversion rates skyrocketed. He continued to refine the algorithm, eventually building a platform around it and launching it as its own business.

Human Predictions is what’s known as a “people aggregator”; it scours the web for info on candidates, aggregates that info, and applies machine learning techniques to make inferences, such as predicting which candidates are most likely to accept an interview, job offer, open an email, or consider a new role. The firm is differentiated from other offerings primarily in two ways. First, it is solely focused on the tech industry, which gives the platform the nuance necessary to make better inferences. Second, while technology is very important, Garms notes that being “human-centric” is also important in recruiting, and the platform provides several features to help recruiters build better long-term “human” relationships with candidates. The product is priced on an SaaS basis that becomes successively cheaper with more users; pricing starts around $400 a month for a single user and decreases as additional users are added.

The company, launched in 2015, now counts firms such as Redbull, Groupon and Trunk Club as clients.

The Buzz

Recruiting will always be driven by the relationships fostered between recruiters and candidates. Human Predictions helps staffing firms figure out which relationships are ripe to be harvested.