For those following Techloop in the transition to becoming a Product first company based on a SaaS platform, one of the key areas we needed to master is how to search for relevant candidates within our platform. Before talking about our latest improvements, let’s talk about what we’ve learned.
Our initial search engine was based on a simple filtering ability. This allowed our users to simply choose the tech skills needed, which would filter down profiles to become the perfect match. Simplicity is always great, but the downside of this was that users filtered results down to very few candidates with the right match, and missing out on those candidates that may have had high potential, but not the exact skills filtered.
Original filter search in 2019
So we went back to the drawing board….
The eureka moment arrived for us when we decided to completely switch a filtering type search for an Elastic based search which we wrote about earlier. This allowed us to work with different weights based on search results, providing the ability to not only show results with exact matches but also suggested candidates that could be also relevant.
We were quite confident this could meet the expectations of our users. Unfortunately, we missed the target. Despite tweaking different weights in skills, there was a larger imbalance between expectations and suggested profiles. The result was that our users had plenty of profiles to review, but the accuracy was lower, making search less time-efficient.
Search results after Elastic implementation
So how did we come to the winning formula?
In the past weeks, we’ve introduced the ability to choose your ‘Must-have’ skills. This gives our users the ability to target absolute key skills in profiles that are critical, while still having the ability to see other recommended similar skill sets related to the “Must-have” skill.
Current ability to choose ‘Must-have’
The outcome now provides much better balance between filtering your search and still utilizing the benefits of suggested results. It’s the first time we’ve seen a similar combination provided when sourcing. The initial feedback we’ve gain also proves this is the right direction for many of our users.
If you haven’t had the chance to try out the latest version of our search, let me know personally at firstname.lastname@example.org. We are keen on getting feedback from both new and existing customers. As a small token for your feedback and help, I’ll be glad to provide a special discount for those supporting us.
Stay tuned for more news coming up!