Do you need a data science management recruiter?
In this post, I’ll explore how to hire a data science manager and if using a data science manager recruiter (or executive search firm) would be helpful. As you likely already know, it can be very challenging to hire a data science manager. That is why many organizations use an executive search firm (or a data science manager recruiter) to hire their data science managers. Of course, not all organizations use an outside company to help hire their data science managers.
To help with this hiring challenge, below are four useful tips for hiring a data science manager. In addition, if you are looking to get a job as a data science manager, knowing these tips (on hiring a data science manager) will also help you position yourself as an appropriate candidate.
Reading these hiring tips will also help you determine if you want to use an executive search firm, or if you want to do the recruiting yourself. So, let’s get started…
1. Think about the desired skills of the data science manager
When creating a job description for a data science manager, one needs to think about the required qualifications versus the desired qualifications. As noted in my 6 Actions to Be a Better Data Science Manager post, at a high level, there are three typical profiles when hiring a data science manager.
The ideal data science manager
The most desired profile is a person who already is a data science manager who also has had hands-on experience as a data scientist. However, it can be a challenge to entice a current data science manager, especially someone who previously was a data scientist, to join your organization. Of course, maybe your organization has projects that are incredibly interesting, or perhaps your organization is willing to pay at the top of the pay scale to entice someone to join your organization.
The manager who doesn’t know data science
Another approach is to hire a person who is currently a manager for other types of projects, such as a software or IT project manager. However, managers in other fields might not understand the challenges of executing data science projects. In other words, a person with this background might not understand the unique challenges of leading and executing data science projects. A “bad scenario” with a person with this type of background is that they learn via trial and error with your projects…
The data scientist who wants to be a manager
Finally, yet a different approach is to hire a data scientist that has a desire to lead teams. In other words, another natural path is to promote a current data scientist into the manager role. This could be someone from your current team, or someone from a different organization. However, not all data scientists make great managers, and equally important, not all data scientists want to be managers. Furthermore, a person with this background would likely understand the unique characteristics of data science projects, but they might not understand how to manage projects. They might also not understand the frameworks that are appropriate for executing data science projects.
2. Know how to find potential candidates
Use job boards
One approach to find potential candidates is to post the job on sites such as LinkedIn and ZipRecruiter. These can be low-cost (or even free). However, many jobs are listed on these websites, even for a specialized role such as data science manager. For example, on LinkedIn, a recent search for “data science manager” returned 783 open jobs. This might sound like a lot, but there were 28,875 open “data scientist” jobs.
Another place to post your job is to leverage a job board within a data science-focused website. One example is Kaggle. However, similar to many other data science-focused websites, Kaggle closed its jobs board. Some websites that are still available include ai-jobs.net, which is a site directly focused on data jobs. Note that, if you look at these websites, there are far fewer jobs than on a site such as LinkedIn (and many of the jobs are listed on these websites are the same as those posted on LinkedIn.
One advantage of using these large job sites is that it is easy for people to apply for your job opening. Of course, since it’s easy to apply, many people will apply that do not have the appropriate skills… In other words, the “noise to value” is high, and because of this, many organizations skip this approach.
However, most data science managers aren’t searching LinkedIn (or any other site) for new jobs, so many candidates will not know you are looking to hire. The people that are searching for a new job will see lots of similar roles. Hence, in order for your job to be seen, you will often need to advertise your role on the job board. This can often cost a significant amount of money.
Another approach is to use an executive search firm or data science-focused recruiter.
3. Explore using an executive search firm or data science management recruiter
The previous section makes it clear why many organizations use executive search firms (or recruiters) for hiring data science managers. These firms can accelerate the process of finding candidates. One way they can accelerate the process is that they have access to their network of candidates (from previous searches). Of course, since the field of data science management is new, their network of data science managers will be small. These companies help organizations hire data science managers by also doing some of the “leg work”, such as posting jobs on public sites like LinkedIn as well as other tasks, such as the initial screening of candidates.
In general, some firms focus on certain geographies, while others companies focus on specific domains, and yet others have a broader area of focus and geographies.
One example of an executive search firm that focuses on Data & Analytics recruitment is Harnham. Brainworkinc and Smithhanley, on the other hand, are larger companies that have some domain expertise in the data science space These are just three examples. There are, of course, many executive search firms.
There are two types of searches – retained search and contingency search.
Contingency search
A contingency search is only paid once you hire a candidate that an agency has presented to you. Thus, their pay is contingent on you selecting one of their candidates. These firms will charge roughly 20-30% of the candidate’s first year salary.
Retained search
A retained search is when a search firm has a more hands-on relationship with your company. The retained search is for a specific period of time. These firms invest their time to make better understand the type of person desired. A retained search is more expensive than a contingency search. The overall fee charged for a retained search is 30-35% of the estimated first year salary of the candidate.
So, why not use an executive search firm? Mainly, for some organizations, the cost is too high.
4. Know the questions to ask during the interview
Once you have candidates identified, how can you tell if the person you are interviewing might be a good person to hire? In other words, what are some interview questions, which are specifically focused on data science?
Below are some questions that might be helpful. Note that these questions do not focus on “technical” knowledge, but rather, on data science management and how someone might lead a team.
- What is unique about data science projects? In other words, what is the difference between data science projects and other projects, such as software development projects?
- How should the team communicate and collaborate (internally and with stakeholders)?
- What are some key drivers of uncertainty in a data science project?
- How would you handle the challenge of uncertainty in data science projects?
- How should the data science team prioritize how they spend their time?
- What does agility mean in a data science context?
- How can agility be used to help deliver data science project value?
- Would you establish a repeatable team process for executing projects? If so, what would some key characteristics of the process?
- What is a data science workflow (or lifecycle) and how might one be useful for the data science team?
Key takeaways
Hiring a data science manager is not easy
Clearly, hiring a data science manager or data science team lead can be time-consuming and difficult. This is especially true for a senior manager who does not have experience managing data science projects. In this situation, knowing how to evaluate potential candidates (such as the answers to the questions noted above) can be a challenge.
Hiring an executive search firm or a data science management recruiter might be helpful
While a recruiter can help, executive search firms are not experts in the field of data science management, and hence, do not know how to evaluate and compare candidate answers. These firms will also not have the expertise to explore if hiring someone, such as a software engineering manager, would be appropriate for your organization.
