How to Lead a Predictive Recruitment Project

Everyone is hearing more and more about predictive recruitment. This new type of recruitment is allowing businesses everywhere to anticipate the capabilities of their candidates and their team members so that they can maximize their efficiency on the job.

At AssessFirst, we have a firm belief that the combination of Big Data and human resources is not a new wave to surf, but rather an opportunity for businesses to bring intelligence to HR data with a real return on investment and may be the key to long term success.

It’s a beautiful idea, but how can it operate? Well, as the leader in the predictive recruitment segment in France, we take advantage of this opportunity by offering tailored predictive recruitment projects.

Today, AssessFirst is able to offer predictive models on success in a given position and adapts the models based on the specificities and challenges that face your business, while also respecting the diverse personalities of your employees, past and present. Simply, “show me who you are, and I will show you how to succeed!”

Let us be clear: by success we mean the ability of an individual to be efficient in their function, while also feeling like they are in the right place.


How to lead the charge?

The success of a predictive recruitment project rests on three pillars:

  • A correctly targeted problem
  • A well-defined study population (generally of a function)
  • A set of effective and relevant data


How do you define your problem?

You simply begin by asking yourself what you hope to change! Perhaps it is to have fewer people jump ship or to have team members perform better? Perhaps (surely) the two resonate with you. And if you realize that it is the most efficient team members that are departing the fastest, what do you do? Having a problem clearly defined allows us to recognize what you really want to address and to offer you the best solution. For example, we have helped businesses wanting to optimize the cost and duration of their recruitment process by applying a filter first that screens candidates for a specific position. Others seek to recruit people with the highest potential to succeed in the office, or simply to identify the candidates that could correspond the best to the enterprise culture and remain in the job longer than average employees through “traditional” recruitment.


How do you define the study population?

By its volume, its responsibilities, the associated indicators of success…

Why must a population be well-defined? Because we are going to conduct statistical studies, and this requires a minimum sample population. If you have 5,000 people in a position, studying 200 is not good enough. If you’re comprised of 250, studying 200 is excellent!

Once we know what motivates your request, we will collect two types of data:

  • A layer of HR data: your data, over a period ranging from several months to several years, such as operational performance indicators (sales revenue, % of objectives achieved,…), or some indicators more HR-oriented (turnover, absenteeism,…). And it’s here that the third pillar, a set of efficient data, intervenes. This point is intimately bound to the first two, knowing of a problem and having a well-defined study population. If these two premises are respected, then it allows us to be familiar with your priorities and define what indicators we use.
  • A layer of behavioral data: the data of your office team members, that is to say, their answers to our questionnaires (personality, motivations, and aptitudes)

We are then going to study the connections that the two types of data present between them and who is going to make a positive – or negative – impact on the success of the team already on the job. What are the personality traits and characteristics that the employees display the most? What aspects of their motivations characterize the veteran employees? To what extent are reasoning skills predictive of their potential?


“This seems interesting, but how much time does it take?”

We aren’t going to hide it, the collection of data is clearly the most time-consuming step and directly depends on the fluidity of our collaboration. In regards to the analysis and the treatment of the data itself, this can range from 2 weeks to 1 month of work!


What does it make?

Once the study phase is completed, the conclusions that emerge are translated into a usable program available directly via the online AssessFirst platform: you will immediately be able to use it as a part of your recruitment process. Specifically, you can pre-select targeted candidates based on the model, which applies a lens that allows you to instantly distinguish those who have the highest probability of success and the highest probability of failure for a certain position. You finally have easily exploitable data regarding whether a particular person has the natural predispositions needed to succeed at your organization!


What ROI to expect

Let’s talk about return on investment. On average, we see a 25% improvement in the performance of new recruits, a decrease in turnover of 50%, and a decrease of 20% in recruitment costs.

Additionally, because you will know what to focus on during recruitment (the model having first made a filter), you can expect a reduction in the cost and duration of recruitment (e.g. through predictive recruitment, BHV Marais has reduced the duration of the recruitment process by 30% and decreased its turnover).


That’s the financial aspect, but we could also talk about HR impact…in a future article, why not? 😉


And then we come back!

Now you say: “Ok, the behavior of an individual has a lot of influence on their success in their position, but there must also be an impact outside the context of the position or the position of the business, right?” We think so, and that’s why an update of the program is proposed retrospectively, for two purposes:

  • To compare, after a year,  the results obtained by employees recruited before the introduction of predictive recruitment with those issued by newly hired employees with the program. #ROI
  • To incorporate this new data to our algorithms so that the device is always accurate and appropriate to the situation, the economic context of your company, and in a world that is constantly changing! #UPDATE


In conclusion

Predictive recruitment is a tool at the service of the recruiter (or the manager). It has neither the pretension or the capacity to replace him or her. Rather it will allow them to focus on what really matters: more time for more qualified candidates, propose evolutionary perspectives to current employees, and identify future potential.
The right person in the right place at the right time. This is the philosophy of predictive recruitment and the key to having employees that stick to the corporate culture. Achieving employees that are happy and more fulfilled each day is the recipe for a more efficient company, not only financially, but humanly.



Christophe HOAREAU

AssessFirst Chief of Science and Innovation Projects




Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.

%d bloggers like this: