re:Worker story: Starting small by bringing data to hiring
I work for Tribeca Technology Group, a managed service provider delivering specialist IT support to clients in the financial industry with offices in London, New York, and Hong Kong. In our small company of thirty-six people, there is only one HR person — me. Like HR administrators at other small organizations, I handle everything from hiring to benefits to disciplinary processes.
I wound up working in human resources after an apprenticeship (like an internship) in my current company. When I began my full-time role as an HR Administrator, I set out to find resources to help me in my new role.
I signed up for a course on “Human Resource Practice.” It covered the basics but was almost exclusively focused on compliance. So much of HR is about not doing anything wrong rather than trying to do things better. For example, when we covered interviewing, the lesson was to use interview process forms to ensure you have good records to cover you in discrimination lawsuits rather than talking about how to find and assess good candidates. Staying compliant is important (very important!), but I wanted to know what I could practically do to make smart, not just compliant, decisions for the people at our company.
It was during this search for more answers that I stumbled across the re:Work website, and I read every guide, case study, and blog on the website. Twice. Inspired by all the examples of how others had used data to improve their people processes, I decided to experiment within my organization.
The first project I undertook sought to understand the pain points in our hiring process. I started simply, by logging the job applications we received in a humble spreadsheet (a homebrewed applicant tracking system, or ATS). I built a hiring dataset by tracking answers to questions I often found myself wondering about our candidates: How many times had this person applied? Where did they apply for the role? What roles were they applying for? When were they applying? Why were they applying?
Next, I looked at our annual recruiting budget, calculated how we spent to fill a role, and figured out the different costs of each sourcing channel. I found that we put the least amount of resources on our most effective channel — from which we hired 62% of applicants — and we put the most focus on a channel from which we hired just 6% of applicants. So we reprioritized our approach and in a matter of months, we’ve been able to reduce the time it takes to hire a first line engineer to four weeks, down from three months.
Another experiment I ran was inspired by the re:Work guide for writing job descriptions. Our usual job adverts led with details about the role, but when talking to candidates, I realised what got them excited about interviewing with us was our clear commitment to make work a happy place for our staff. To test whether the wording in our job adverts made a difference, I ran a simple A-B test. I placed our usual advert for a senior developer role on our website and let it run for two weeks and then reworded the advert, shifting the focus from the daily tasks of the role to the bigger picture of our culture. I let that new advert run for two weeks, and we received an increase of 20% in number of applications.
This whole process has opened my eyes to the power of data and helped me realize that an organization doesn’t need a huge people operation staffed with PhDs to start using data to make better decisions. Anyone with a spreadsheet, an eye for detail, and a set of curious questions can start testing hypotheses.
This blog is part of our re:Work series celebrating the millions of small and mighty businesses who are innovating and driving the economy around the world. Read more: