I remember receiving a call from a client that I had worked closely with, building their IT team from a single member to a fully functional team. One of the first developers I had placed with them, an extremely talented guy (we’ll call him Bob for the sake of argument), had handed in his notice and they needed to replace him. I asked them if there were any changes to the job spec we worked from for the developer role at that company. “No”, came the answer, “we are essentially looking for someone exactly like Bob”. This response is understandable – after all, their ongoing business plans required a team with a specific skill set and knowledge base. Therefore, if a member of that team was leaving, it would be ideal to replace them with someone as similar as possible, to minimise any disruption. However, the kicker is that it is impossible.
You will never find anyone with precisely the same skills, background, knowledge, approach to problem-solving etc. as someone else. Software development is a knowledge-based role that requires solving unstructured problems – i.e., problems where there are no known right answers, and multiple ways to solve them. This is true of any role where there is a reasonably high expectation of discretion in how the employee carries out their activities. For example, a concept designer has a huge amount of input into what designs to exploit, and which to reject, when exploring new designs for future innovations, but factory workers on a production line have very little say in how they perform their duties.
Hiring is essentially a planning activity. You are making certain predictions about the future, and what organisational resources you will need in order to be successful in that future. In this case, you are predicting what human resources you will need. You need to know what duties that human resource needs to perform, what skills they need to perform it, and how to motivate them to perform those duties. This involves placing certain “bets” on likely outcomes of the future. These bets may or may not pay off. For example, HMV’s MD famously “bet” in 2002 that online retailers and downloadable music were nothing but fads, and so failed to develop either the human or infrastructure resources to exploit this space themselves, until it was too late. However, the problem is that it is very difficult to predict how environmental factors will affect your business over time. Therefore, business owners have to make decisions on what they think will happen, and assume that those future scenarios are highly likely to occur. After all, any business is faced with an almost infinite number of possible futures, but you cannot plan for an almost infinite number of outcomes. In order to achieve anything practical, you have to pick a course, and hope for the best.
To bring this back to the company I was discussing, they assumed their ongoing business required someone as similar to Bob as possible. This is understandable, as our predictions about the future are based on our experiences of the past, and these past experiences create an anchoring effect: for example, if a company hires a graduate into a non-graduate role, and the graduate performs well, the company may well only look to hire graduates for that role in the future. This also underpins the foundational concept of the job interview; by asking candidates about their past successes, you assume that past achievements act as a realistic predictor of future achievements. Unfortunately, as many struggling companies have found out when hiring CEOs from successful businesses, only to see little-to-no impact on their fortunes, past success is a poor predictor of future performance, as performance is grounded in specific context. No two companies are precisely alike, and academics who study chaos theory within organisations assert that the likelihood of the exact same situation happening twice is so unlikely as to be impossible. Two situations may superficially appear similar to one another, but in all likelihood the people involved, the knowledge they have, and the external environment they operate in will all be different.
So, most of the work in hiring activities involves creating a set of performance criteria to assess who is ‘best’ for a role. For a sales job, the key metric – sales figures – is an obvious and intuitive choice. But, there are different ways of achieving high sales. “Cowboy” tactics, which often involve lying to candidates, and hard-selling to clients, might well result in a high sales figure. But is that the kind of person you want working for your company? An almost universal problem for job roles, including software development, is that it is virtually impossible to come up with an objective measure for performance. In software development, there is a mythos of the 10X Engineer (the software developer who is 10 times better than his colleagues). However, as Shanley Kane notes in her superb critical evaluation of the 10X myth, the whole 10X concept was based on a single, highly flawed study of developers conducted in the 1960s. Whilst companies like to talk about hiring 10X Engineers, this ignores how an individual’s performance is affected by a huge number of variables, and how the very concept of performance itself is left ill-defined. There is no objective way to determine who is the “best” person to hire, which begs the question, why are companies so obsessed with finding the best?
Finding the best depends on your own individual assumptions about the future, and what resources you will need to bring into your organisation in order to effectively exploit the future. However, those assumptions can be massively limiting. The company I was discussing had decided that the ‘best’ person for them would be as similar as possible to Bob, and this meant finding his replacement was highly problematic. A number of talented candidates were rejected because they weren’t Bob. The recruitment for the role stretched out over months, until their desperation to fill the role finally outweighed their need to find Bob 2.0.
Hiring requires a balance of flexibility and rigidity, and this requires a shift in thinking from “best” to “good enough”. “Good enough” typically carries negative baggage: ‘oh, you aren’t perfect, but you’re good enough’. However, in this case, “good enough” is a realistic acknowledgement that performance assessment can only ever be made retroactively; and, without the ability to accurately predict the future, the whole concept of “best” is meaningless and redundant. An organisation is not stable, nor predictable, nor mechanistic. It is messy, dynamic, nonlinear, and chaotic. So, hiring managers need to understand what their assumptions about the future are, and then use those predictions as guidelines for who to hire, rather than a rigid blueprint. “Good enough” is the best you are going to get, but the good news is that “good enough” may prove, over time, to be the best hiring decision you could have made.