A 30% failure is huge within any context

It is becoming apparent that many companies are failing to select the correct candidates to hire and it is getting to a stage where those involved must be made aware in order to rectify the mistakes.

A startling figure is that 70% of hires make the first year anniversary at their respective companies. That may seem impressive but one must bear in mind that that means that 30% of new hires don’t even make it to a year working in data at the company.

In what other area of the business would that percentage of failure be acceptable? None. If you failed 30% of the time in any other area of the business, it would be a catastrophe, so why is it okay in data recruitment?

I mean, of course this may be happening in other areas. Only recently we published a very popular blogpost about how only 20% of C-Suite hires are the right fit for their company. We are focusing on data recruitment as it one of our areas of expertise.

Usually companies have adapted their style of hiring for data scientists from their style of hiring for software developers, especially larger companies. This often involves tests such as brainteasers involving coding or statistics, which candidates actually practice for and which, in reality, are a waste of time for everyone involved. These clean-cut problems ignore the fact that data professionals can be working with messy data. While using guidelines from software developers is a step in the right direction from generic or vague hiring practices, it is obviously not going far enough to highlight the best candidates and their skills. If a company is working from an old model, how can it expect to attract cutting-edge candidates.

Outdated techniques are just one side of the problem. Finding prospects with the right skills is of utmost importance but how many businesses know what they are looking for in this area? Each candidate must have a mixture of skills involving business, databases and maths & stats. They should be adept at model building and should be used to using different models to solve various problems.

They must have experience in Machine Learning but if they do not, sometimes their degree may have involved modelling and they just need to brush up on the skills in order to become a valuable member of the team. Team members must be able to, firstly, find the data, and secondly, process the data. These are two separate skills and the candidate must have both. The fact that this is counterintuitive to those who are not experienced in the sector means that hiring managers can be tripped up by this and other similar eccentricities with data. And we have not even touched on the different coding languages a prospect may need to know. For example, data analysts often only work in Excel which means they are ill-suited to other data roles. It is complex and those with some or no experience within it can be challenged by it.

On top of all that, a data expert should be able to understand problems within businesses, recognise how data can be used to solve it, solve it and then translate it back for the layperson to understand. If one part of this solution is incorrect, the problem remains unsolved. They should also be able to handle pressure, learn new things and mix well with other team members.

It is clear now why so many data professionals do not make it to their anniversary in their company and a lot of the time it is not down to flaws within the candidate. Companies should take care when hiring to ensure they are hiring the right fit for their data team, prospective candidates should be encouraged to create their own portfolio to showcase their skills and someone who has expertise in evaluating these should be in charge of progressing those prospective hires to the next stage of recruitment. With more awareness, together we can get those anniversaries up to 100%.

Curious about how Zenshin Talent can help your organisation? Contact us today for a no-strings conversation about your needs and our experience.