Sutro’s Algorithms and the Beauty of Scoring

on 04 | 09 | 2015

Data today is becoming increasingly valuable in practically all consumable shapes and forms–and the world of recruiting is no exception. For Riviera, we use Sutro as our own powerful data aggregation platform, continually collecting and classifying data from external and internal databases. The data empowers our recruiters to make better matching decisions by presenting real-time, data-driven profile insights. But aggregation is only part of the story. As inbound data is repeatedly fed to Sutro, our powerful algorithms are digesting it all and feeding back useful findings to our team.

Utilizing the algorithms initially developed by our friends at eAlchemy, we’ve put together a workflow where candidate and company profiles in our system are automatically scored against sets of distinctive scales. We utilize a variety of other algorithms that score based on activities and trends, and as we obtain more data, we’re refining and building new algorithms.

While this may seem trivial at first glance, these algorithms incorporate meticulous methodologies using each data point collected to score individualized profiles at a macro level. This is hugely important because all of this information used to be compiled manually on a case-by-case basis; now, we are automating some of the most labor-intensive pieces of the sourcing process. The time saved by our recruiters is then spent curating candidate profiles and contributing a level of insight that a machine simply can’t.

One of the most important factors in our scoring efforts is the interactions we capture throughout the recruiting life cycle. Information that cannot be gleaned through analyzing online profiles and company databases is gathered through direct communication between recruiters, candidates and companies. As such, our goal is to capture every encounter we have with both companies and candidates so that they can be incorporated into our profile scoring.

The benefit of scoring is twofold. First, it allows us to quickly identify standout candidates, regardless of specific role. We also use it as a filtering mechanism for specific roles. Using these scores, we can sort query results based on different factors. For example, if a client requires their next leader to have been involved in a company that achieved a specific outcome (e.g. IPO), we can instantly narrow the candidates down to those who have been involved in those types of outcomes at prior companies. If demonstrated ability to scale teams is another key criteria, we can also filter based on that dynamic.

Because we’ve incorporated the lift from our algorithmic scoring formulas into Sutro, we have seen increased leverage in our business, which has led to better service and experiences for our customers. This has been substantiated by the feedback we are seeing from the customer base. Now, our recruiters are spending more time interacting with clients and candidates rather than simply searching through profiles and manually matching based on what they might think is best. This also provides the benefit of allowing all recruiters and–by extension–all clients and candidates to benefit from the data that we are aggregating and assessing. And of course, our scoring algorithms are a critical component in our matching algorithm, which is the final technology-based step in finding the best fit between clients and candidates.