What’s the 1 thing that can save even a bad interview?
A golden tip to save you on the worst of interviews
I talked in my last article about how I got my job at Facebook by showing people what I could do. This approach has also influenced our mission at LMNS, the company started this year to apply skill-based hiring to non-technical roles.
But skills-based hiring can do more. It helped me save a bad interview. In this article, I want to focus on some of the feedback I have heard about skill-based hiring, and how my own experience in getting a job invalidates that.
After 1 year in my first role at Facebook I was ready for a change. At this point, I was lucky enough to have built a network that meant getting an interview for possibly the coolest role ever (for a data scientist) was being a data scientist at Oculus.
As I had interviews already set up, one might think that showing your skills is unnecessary, they already have that.
But, I believe more is more when it comes to job apps, and given this was an extremely competitive role, I was going to do anything to make it happen. So I decided I would do a piece of data science work.
Feedback 1: Picking the project
At this point, whenever I tell people about my approach, they say:
“But how do you know what project to work on if they haven't told you?”.
For me, this an opportunity to show your prospective employer that you understand what their business problems are, which is already a key aspect of the job of a data scientist. So, with all the information that you have, what do you think is the problem they are trying to solve?
For my 1st role at Facebook, I was applying for a role in Business Integrity, what problems do you think this division is trying to solve?
2. Executing the task
As I was going to be working on a new product, I researched what had already been done in terms of insights.
Naturally, I had a huge advantage here, because being an employee I could actually read a lot of product forums, but I tend to find that desktop research does just as well. It seemed we knew a lot about who our current customers are in terms of demographics and in-app behavior, but we had no clue about what they might want.
That was it! Why don’t we identify our users' interests? Lightbulb 2, this could be an interesting clustering exercise! Let's go!
3. The time taken to prepare for this task
Another thing people often tell me is:
“that seems like a lot of work if you aren’t getting paid”.
This is a fair point. And it's true, it is a lot of time. However, I spend hours preparing for interviews. Those hours are effectively spent practicing answering questions that otherwise lead to no growth for me. Once the interview is over, I never use those skills or answers to those questions until I happen to find myself in another interview process.
On the other hand, using the time to do a skills-based assessment gives you time to practice your skills. For this particular assignment, I spent a weekend working. I would most definitely have spent that same time preparing for an interview, but this time, I spent it doing a task that would increase my chance of getting the job and I invested into my skills. Even if I didn’t get the job, I was now in a better place to get a different job in the future than if I had spent all that time preparing for an interview
4. The actual interview
Having conducted a clustering analysis over the weekend, dropping it all into a 3-slide presentation, I felt very excited about the interview. We got into the room, I started answering questions and then, then I forgot how to calculate averages in SQL.
I can’t remember when the fear and panic started to set in. But I was standing in front of a whiteboard and my mind was blank. Calculating averages in SQL is something I basically did for breakfast each day, and yet, here I was, applying for a data science role and I couldn’t get the command to fit.
Needless to say, the interviewer was not impressed. She finished the exercise early and asked if I had any questions for her.
And this is when I remembered the ace up my sleeve. I asked her a couple of questions and then, asked if I could show her my work. I talked her through what I thought about the product, how designing new features could be based on what our current users like, and even better, we could find new users by comparing the interests of current users to ones we don’t have today. I should her my clustering analysis, the beautiful dendrograms I had made, and finally the product recommendations as a result.
The whole time she looked at me and said nothing, but I could feel the air in the room had changed. Of course, this girl could calculate averages, she must know something, and more importantly, she has shown me the skills that will really count in her job, oh and actually, she has done part of the job we were most definitely going to do as a piece of this analysis, and now I don’t have to do it.
I got the job. For the almost 2 years afterwards that I spent at Oculus, I joked with her about how she let a girl that didn’t know how to calculate averages in SQL onto her team.
The lessons learnt for me are:
a) Skills-based recruiting allows you to show your actual strengths, not mitigated by your nerves on interview day
b) Skills based-assessments show the company you are applying for that you care, and you help them not waste time on interviews that poor predictors of performance.
Think this is too tough? Or too much? Well, that’s exactly why LMNS exists. We want you to spend time showcasing your skills, not constructing the problem and research people.
To join our challenges, grow your personal brand and win some cash, sign up here.
