Have you ever wondered if employers track hiring success? If you wonder how they do it, here is an overview of essential metrics that would help track the effectiveness of hiring and onboarding.
Coming back to hiring success – you'd probably be surprised to learn that 76% of employers reported having hired the wrong person for the job, according to Glassdoor.
Well, to err is human. However, is there anything you could do to improve the hiring process quality? According to the research-based book by Daniel Kahneman, Olivier Sibony, and Cass R. Sunstein entitled "Noise: A Flaw in Human Judgement", there is a way to identify and reduce noise in decision-making. Even when it comes to hiring.
Here is the case in point – a software company is looking for a Product Manager to lead the new product launch. The hiring team is presented with a pool of 7 candidates for a Product Manager position. For privacy reasons, we've taken out the names of the candidates. However, we know that Candidate 1 is the least experienced applicant, and Candidate 7 is the most experienced one.
Each of the 12 hiring team members (HR, Founder, VP of Sales, VP of Marketing, Data Scientist, and others) evaluates each candidate independently. A hiring team member then assigns a score to a candidate on a scale of 1 to 7, where 7 is the top-notch score.
In Figure 1 below you can see the results of such an assessment. Now let's have a look at what's so intriguing about it.
Figure 1
Having a closer look at the column Mean Rating per Assessor, we observe that some assessors (decision-makers in the hiring team) tend to provide lower ratings to the candidates. Have a look at the average rating provided by the Founder (3.0) and Back-End Developer (5.0).
Kahneman, Olivier, and Sunstein argue that the tendency to over-and underrate the candidates stems from the noise in the decision-making. According to the co-authors of "Noise", such kind of variability in the assessment (3.0 vs 5.0) relates to the fact that some assessors are generally stricter than others; the co-authors coined the term "level noise" to describe such kind of variability.
Level-noise alone, however, is not the only component of the noisy decision-making by the hiring committee. Let's see why that is so. For a moment, let's assume that level-noise is the only component of the noise and have a look at Candidate 6.
Note, that the average (arithmetic mean) rating received by Candidate 6 is 3.58. Now, if level-noise (the tendency of some assessors to be stricter than others) was the only component of the noise, you would predict that the rating, assigned to Candidate 6 by a particular assessor (let's say by a Back-End Developer) is calculated as:
Candidate 6 Mean Rating + (Grand Mean Rating - Mean Rating by Back-End Developer)
plugging in numbers from the Figure above: 3.58 + (3.94 - 5.0) = 2.52.
However, we see that the Back-End Developer rated Candidate 6 with a very high rating of 7.0. We therefore can suspect that there could be something else that adds to the noise, apart from level-noise. What is it then?
SYSTEM NOISE = LEVEL NOISE + (?)
The co-authors of "Noise" further argue that there is another component to the system noise (or "total" noise in decision-making). Have a look at the standard deviation (amount of the variation of the set of values from their mean) of the rating for Candidate 6 – it is 2.11. Yet, level noise – the standard deviation of ratings performed by the hiring team – explains only 0.58 of the system noise.
In the prompt to the case in point, we've mentioned that candidates were ordered by their experience (e.g. Candidate 1 is the least experienced, and Candidate 7 is the most). In a noise-free world, each candidate would get a similar rating from every assessor. It means the Back-End Developer and the Founder (in fact, anyone) would give the same rating to Candidate 6 (for example, 4.0). All 12 hiring team members would give a similar rating to Candidate 6, there will be no variance in the rankings, and the assessment column for Candidate 6 look like that:
Figure 2:
Now recall, that the Back-End Developer has given ratings of 5.0 to Candidate 1 (the least experienced) and 3.0 to Candidate 4 (more experienced), refer to Figure 1.
The inconsistency of ratings given to Candidate 6 by the hiring team is called "pattern noise", another term coined by Kahneman, Sunstein, and Sibony. Had everyone in the hiring team been completely free of noise "pattern noise", we would expect similar ratings assigned to each candidate across all the assessors.
Rephrasing Kahneman and his co-authors, in a perfect world, the candidates would face unbiased recruitment, however in reality they face a noisy system. The candidates suffer from two levels of noise:
Level noise – the tendency of some hiring team members to be "harsher" (e.g. give a lower ranking on average) than other team members
Pattern noise – the variability of each hiring team member's ranking assigned to a particular candidate
Conclusion
Hiring is a complex decision-making process, where decision-makers are prompt to noise. As a result of the noise, your company faces a higher risk of miss-hiring. While it is impossible to eliminate noise completely, there is a way to decrease the level of noise by performing a noise audit.
There are two ways how to deal with the consequences of noise in hiring. The proactive approach is described in detail in this Harvard Business Review article. The reactive approach is to work around the consequences of noisy decision-making – by helping new joiners adapt to the new place and get outstanding Onboarding experience.
Want to Create an Oustanding Onboarding?
At Corpedios, we put together the best features of text, images, and videos to create an exceptional onboarding experience for employees at your company. Try onboarding new talent via a conversation that we tailor to your needs in our text message-based onboarding solution. Your coworkers and new hires text – would it be terrible to put it to good use?
Comments