Hiring Like The Life Of Your Team Depends On It (5/5)

Published on March 29, 2018.

by Hubert Behaghel

What underpins this value is that slowing down is not an option and moving faster is a core responsibility of a team.

Hiring Process Design

After having reviewed the risks of not hiring, the risks of hiring and its pitfalls, it’s time to look at what that means for the hiring process, and the stages to put the candidate through to deliver great team benefits. What are they?

minimise the time spent or worse wasted
minimise the risks posed by team lifethreats
maximise for team fit assessment
make great hiring decisions that minimise the risk of not growing

Let’s assume it starts when someone posts a CV for an open position in your team, though at times, you should consider finding the CVs yourself. If you know of someone good who would fit well in your team, why wouldn’t you approach that person and revel in the luxury of really building your team by choosing the people to work with?

Generally, we pay someone else to do it for your team and most of the time that’s how you’ll get CVs but remember: ownership is better than diligence. It’s a bit like the parents and the childminder, the childminder is paid to change the nappies but that doesn’t mean that every once in a while, a parent couldn’t do it. After all it’s their child. And after all, it’s your team. That’s part of hiring like the life of your team depends on it.

Once you have a CV, the first stage is CV screening. CVs are a funny artefact and my advice is the following — give it two minutes, nothing more and make a call, thumbs up or down. Like any engineering decision, make sure each CV is reviewed by two people. After that, even if you move it to next stage, forget about the CV. It’s useless.

For the second stage, there is a debate. Unattended test or phone screen? The unattended test is definitely convenient as it uses almost none of your time, but exactly the opposite for the candidate. That introduces a risk in itself. We may lose good profiles for no good reason.

Less than 10% of the candidates we send a test to come back with a solution. We are not widely regarded as a tech company and being based in Osterley, I don’t think Sky is attractive enough to harmlessly impose this burden on all our candidates.

The other drawback of an unattended test is that all you’re looking at is hard skills. Whereas a phone interview lets you ask technical questions as well as my favourite, behavioural questions (have you ever been in a situation X? What did you do?). It’s precisely for this reason I strongly recommend replacing unattended tests with a 30-minute phone screen. The preparation can be optimised with a question bank that’s collaboratively maintained in a wiki. Here’s one I worked on for Sky scala community few years ago. You also need to augment it with behavioural questions to help you assess the candidate’s fitness for your team, and have a question bank for our values.

A phone screening would need 5 minutes prep time, a 30 minutes call, and another 5 minutes to collate your feedback — in total 40 minutes per candidate over the phone, versus 20 minutes compiling, executing and scanning a test solution and 5 minutes to summarise your findings. The unattended test saves you 15 minutes per candidate, but doesn’t allow you to fail fast on the biggest risk, which is team fit, while phone screening is aligned with our department’s value “Talk to your customers”. A candidate — someone who is looking for a job and wants to work in your company — is a customer of sort, and as such deserves a good experience.

If you really want to resort to unattended tests, here are few thing to keep in mind:

Make sure it’s easily doable in 2 hours or less.
Ensure you have a clear criteria, as discussed previously. Taste and randomness have no place at this stage.
Be clear on the rules that allow the test to bypass: e.g. someone who has worked in the team a year ago and was good, someone who comes with strong references from respected sources, someone who is an active OSS contributor with loads of code in the open (ask for a portfolio of code instead), etc.
Refresh your unattended test at least every year, don’t let agencies or the web spoil it.
Keep an eye on the amount of drop-off that your unattended tests generates, and react to it accordingly. Don’t put your head in the sand.

Whatever means you use for stage 2, you need to be clear on what you mean when you say a candidate “is a no”. What question are you answering? Here it is:

Is there a chance the candidate will be able to work with my team and help it deliver value?

Not “Did I like him?” or “Could she do better than me?” or even “Was those answers flawless?“. At stage 2, you’re not expected to have a deep understanding of the potential of the candidate. You’re just here to eliminate the obvious wastes of time.

On stage 3, it’s time to take it on-site and face-to-face. A pairing exercise is great. So is a whiteboard session. And again don’t get caught into pure hard skills assessment. Even tech interviews can assess soft skills and values. Assign a value to each interviewer ahead of the interview, let each cherry pick a couple of questions from the relevant question bank and wrap them into the exercise. For Improve is better than rebuild, tell the candidate that the outcome of the previous question is now deemed legacy and the requirements have changed — probe for a rewrite and observe. For Deliver business value and expertise, at each step, make the candidate think MVP and iterations, etc. And make sure you also have an additional purely soft-skills interview.

Last but not least, the debrief is to be facilitated with this question in mind:

is this candidate better than half the people in the role already?

To Conclude

I hope you enjoyed this deep-dive into hiring like the life of your team depends on it. I urge you to reflect on that within your team and to discuss this with your manager, on #disco-department or directly with me.

As a last piece of advice, do use a transparent and collaborative process to track recruitment process. And believe in measuring: measure your churn, measure the leadtime at each stage of your pipeline, have SLAs, have stats on reasons given not to hire, keep an eye on the date the last team member came on board. Without that, no continuous improvement can happen in your hiring. Own your team’s data, its life depends on it!