Hey there! I hope you had a great Easter and managed to take a break.
I have a lot to share today, so the newsletter is a bit on the long side. So I’ll get right to it.
To help you improve your hiring process, we’re looking at how to ask candidates for feedback. Tactics, what and how to ask, and tools to set it up and get started.
On equality, we’ll take a look at how AI is used in recruitment. And what you need to know to keep your hiring process fair.
P.S. If you have some tips on how I can improve this newsletter I’d love to hear them - just reply to this email.
There’s no better way to find out how good your hiring process is than asking the people going through it. What do they like? What do they hate? What do they think of your company?
It’s simple, right?
Not quite. It matters who you ask. Who responds. How you collect the data. How you analyse it. And whether the data you get represents all candidates - or just the few that responded.
There are also a few recruitment specific issues. Unlike getting feedback on marketing or a product, there is a different power dynamic. Some people may not feel comfortable providing feedback if they think it could impact their job prospects. Particularly if their feedback is negative.
And when you ask matters. Someone that’s just been rejected may not have anything good to say. But, give them some time, and they could be more objective. On the other hand, your new employees may be overly positive. No-one wants to start off a new job on the wrong foot.
Also, what people say they’ll do, and what they actually do, can be very different things. And this can be entirely unintentional.
If you ask me today if I plan to exercise next week I’d say yes. I honestly have every intention to. Will I? It depends. On the weather. How late I work. If I’m tired. Or feeling a little lazy.
But, when done well, asking your candidates for feedback is invaluable. There are few better ways to understand how well you’re performing, and how you can improve.
Below are some really helpful articles on how to get valuable feedback from candidates on their experience of your hiring process.
Some of the main points are:
In case you’re not sure of the value of a candidate survey, this article should convince you. Or a sceptical colleague. Five of the insights you can gain are:
If you’ve been through Heathrow security you will have seen panels with four emoji buttons to rate your experience. As this article highlights, this simple survey can be revelatory.
The company behind the device, HappyOrNot, has shown that if you make it easy, and anonymous, people will give you feedback. Even if you don’t give them a prize.
And with much better response rates, you can get a lot more data, more often. Using such a simple survey at different points in your hiring process could be a fantastic way to quickly identify problem areas. Then, once you know where to focus your attention, you can use more traditional surveys to get more specific details.
This article provides good points about how to design your survey to avoid unrepresentative results. It focuses on feedback from candidates that at least made it to an interview, but it is valid for all stages.
For example, if you only interview new employees about the hiring process you will miss the views of everyone you rejected. Which is the majority of candidates. However, if you ask your new hires about the onboarding process, you’ll have a good sample.
The article also mentions about exploring differences in results between departments. Great data to get - it just needs to be done in a way that doesn’t erode the candidate’s trust that you’ll keep their results confidential.
Although this article focuses on customers it was one of the best I found on creating surveys. It looks at proven ways to turn surveys into a source of useful insight, and covers:
😄 vs 😡: The Net Promoter Score (NPS)
As the Net Promoter Score seems to be used for everything these days I felt I couldn’t ignore it. For those of you that are unfamiliar, it measures the willingness of customers, or candidates, to recommend your company to others. This article will give you all the details.
The reasons it’s popular is that it’s simple and provides a single score that everyone understands.
Personally, I’m not so sure how good a measure it is for candidate experience. The NPS was developed as a proxy for company growth. Maybe it will indicate if you’re getting more job referrals - but it’s not clear that a high NPS will translate to better hiring results. Or a good candidate experience.
Additionally, as it is an overall score for your entire hiring process, it won’t really help you work out where you could be doing better. Something to consider before going all in on NPS.
🧰 Tools to get your survey up and running
Whatever survey tool you use should be simple, easy to use (for you and the candidate) and provide good analytics. And of course it has to look great on mobile.
If your ATS doesn’t have a survey feature, or you’re not happy with the one it does, I’d recommend:
If you missed those or would prefer an audio overview, this great podcast will bring you up to speed on how AI works and how it is affecting people.
Now it’s time to bring it all together and look at how AI is used in recruitment. And what you need to know to keep your hiring process fair.
Provides a good overview of AI use in recruitment, and the key problems it is trying to solve. It takes a positive view of the changes - highlighting the benefits to the recruiter and employer rather than what it’s like for the candidate.
I don’t agree with the claim that AI removes bias. It’s more likely just less biased than someone with a pile of CVs to review in a short amount of time. A view shared in the articles below.
One interesting anecdote was on the use of AI in video interviews (HireVue):
“Where on-demand interviewing differs is that you should also practice your facial expressions and exaggerate them — a huge smile that might seem ridiculous in person will be picked up more easily by the A.I.”
And the alternative view...
For anyone worried that the entire recruitment process will be automated, this article is a healthy reality check. It explores the risks of using AI in recruitment and provides real-life examples of how automation can have unintended consequences.
Well worth a read. And if you take only one thing away from it:
“The biggest risk of AI in recruiting (is) that it will perpetuate all the biases we’ve had.”
A good article that covers the potential and challenges of AI in recruitment. It covers tools in:
🔩 The AI behind candidate search and recommendation
Search automation wasn’t always so complex. In the beginning it was not much more than scanning CVs for keywords. Then we wanted it to better understand our keywords. If we searched for “front-end engineer”, we also wanted to see CVs for “React developers”.
Now we’re analysing the attributes of our best-performers - and trying to use AI to ‘learn’ how to match applicants’ CVs and assessments to them.
But how does this AI actually work?
We’ve been offered a glimpse by LinkedIn, who recently shared an article on the AI behind their recruiter search and recommendation systems. It’s rather technical - and dry - so I’ll do my best to summarise the key points:
The article goes through the extremely complex way that AI is used to recommend candidates. All the different data points it uses, and how feedback helps to improve the results.
But all of this work has one purpose: to optimise for people opening InMail.
Not for who is best suited to a job. Not for who is best qualified. Not for who has the most potential to succeed.
It shows just how important it is to know the end goal of any AI system. What it is aiming to achieve? What it can it actually do for us, even if it is working perfectly?
In this interview, Lindsey Zuloaga provides insight into how HireVue (video interviewing software) is using AI. And how they’re trying to reduce the bias in their AI. They seem to be taking a thoughtful approach, including using organisational psychologists to ask the right questions.
My main question is if they’re actually measuring the right thing. Can job success really be predicted by a combination of tone of voice, facial expressions and the words used?
As an employee of HireVue Lindsey is obviously biased towards the benefits of their software. So if you listen to this podcast I’d recommend listening to the following one as well for a more balanced view.
A really informative interview with Miranda Bogen from Upturn - a non-profit with a mission to promote equity and justice in the design, use and governance of digital technology.
Miranda discusses Upturn’s report: Help Wanted - an examination of hiring algorithms, equity and bias. Miranda highlights that:
👉 Get the full report here
That’s all for now - it was a rather long one! Looking forward to seeing you again in 2 weeks. But feel free to reach out before then :)
And if you know someone that would like this newsletter you can send them this link to sign up.