Table of contents

  1. Intro
  2. Kick-off
  3. Prospecting
  4. The pipeline
  5. The Hiring Manager interview
  6. Debrief
  7. The offer
  8. Taking smart risks
  9. Common anti-patterns

The role of the Hiring Manager

Intro

This article represents my personal view and take on this role, not a policy from my current or previous employers.

The Hiring Manager (HM) is a manager’s role when hiring someone who’ll report to them. This guide describes how to act effectively as an HM for Data Science - though it mostly fits for tech and analytical positions. It differs from your previous hiring role as an interviewer applying cases. You are now accountable for hiring the right person, far from waiting for a candidate who gets “Strong approval” in every hiring case and simply concluding to make an offer. We focus on hiring Individual Contributors (ICs).

We discuss conducting an HM interview, crafting the offer, and taking smart risks when hiring.

This guide considers a reasonably large company with a Talent Acquisition team and different people interviewing the candidate. If you work in a smaller company, you will likely play more than one role.

Pre-requisite

The first thing is the budget. In a large company, it will be very structured and pre-defined, while in smaller companies, it might be decided on the fly. Regardless of your case, check it and what is the associated seniority level. Verify the flexibility if you can go for more junior or senior candidates, depending on the circumstances. However, define the target level to enable Talent Acquisition (TA), or even yourself, to search effectively. Consider the timing. If you want a hire in Q2, act on Q1. Decide or talk to other leaders to know when they expect to have this person in case you are serving other orgs.

Kick-off

Create or use a template your company offers for an alignment document to specify the profile you are looking for, skills, expectations, previous achievements, motivations, companies to prospect from, job titles for searchers, etc.

If a generalist DS or MLE would work for your position, they can move through the generalist pipeline, which speeds up hiring time. If the position is very particular, consider creating a job description and defining interviewers who can check for that knowledge.

It is a moment to decide if accepting someone from a lower seniority than the budget makes sense. A new hire is not only a marginal decision. Think about the team composition and the scope. If one or more Seniors are in the team, a mid-level can likely fit well, so it is okay to down-level. If the scope ambiguity is high, and the target level is a technical leader, you probably can’t hire a Senior candidate.

Prospecting

The TA team should prospect the correct profile if the HM can communicate all the nuances regarding the acceptable profile in the alignment document. However, the HM will likely miss important information in this process.

Do a little bit of prospecting yourself. You will realize you are using criteria to select people you did not inform the TA team of. Update them as you recognize new criteria. Keep prospecting if you have a few minutes. You don’t need to invest time in it if you are happy with the candidates in the pipeline.

Team up with TA to approach high-profile candidates directly since it makes it more likely for them to engage in the process. For example, you send that LinkedIn message with a comment linking their experience to the position you are hiring for.

The pipeline

An engaged HM is the secret to a successful hire or hiring pipeline. The HM is expected to partner with TA and stay close during this process. Partner with the TA team so they can inform you what you care most about or visit the pipeline regularly in your company’s system. Examples of action items in case of bottlenecks: pushing for internal referral, changing the candidate’s profile the team is prospecting, calibrating one interviewer who has been rejecting most of the candidates, etc.

The TA team usually requests the candidate’s compensation expectation in the first interview. They will also have a better idea about what scope the candidate is open to. There’s room for a sanity check regarding compensation and expected scope here. There is a chance the candidate is out of your company’s salary bands for that level, and there is no need or possibility to hire the person at a higher level. You can call to stop the process for the candidate; otherwise, we are wasting everyone’s time. Candidates usually expect a 20-50% salary increase in a new offer, and placing them at whatever level is needed to fulfill compensation expectations is not recommended. The Data Science market has a supply shortage, and there will be candidates who’ll be plain too expensive. Forcing their way into the team can compromise other members’ engagement and generate attrition.

The Hiring Manager interview

Your interview is going to complement the others. Do not try to assess everything. During the interview, you can do little sanity checks about dimensions measured in other cases, but at the risk of overlapping information and lacking info to make the right choice about the offer. The main info you want from it is the candidate’s seniority.

Here’s a suggested script for the HM interview. One way of approaching it is to interview the person as you would do their performance review regarding their last 6-12 months. First, You want to understand their scope and check how well they perform on it. Here’s a way to conduct it that helps achieve it.

  1. Present yourself, tell what you do at your current position;
  2. Explain how you will conduct the interview and its objectives: “This interview is to provide a picture of your day-to-day responsibilities and interactions with your team. For that, I’ll ask you to briefly explain your role and give a high-level explanation of two projects you were involved in recently. All the questions will dig into these two projects. In the end, we’ll have 5-10 minutes for your questions, and I’ll talk more about the opportunity.”
  3. Take note of the two projects to formulate good questions. The way they answer the follow-up questions is going to reveal their seniority.
  4. Provide them 5-15 minutes to interview you. The kind of questions they ask will provide further evidence about their seniority.
  5. Sell the position. Let the candidate clarify it with more questions (you can switch 4 and 5).

You might change “project” by initiative or program if they are more senior. If it is a Manager or a more senior management position, a particular initiative isn’t likely to be enough to create the picture you want, but senior management hires are out of the scope of this guide.

Notice how the follow-up questions focus on the environment and not only direct questions. You want to generate information to shape the conversation into more specific answers. We group them:

People: understand who works closely and contributes to the same initiatives, their peers, and their leaders; “In this project, who else was working with you?” “You said you were working with another engineer. How have you split the work?” “Your manager was involved in this project. What was their role?” “Who was leading this project?” “Who was the main client of the project?”

Project: the kind of project reveals the trust of their team, and it represents a specific thing to dig into indefinitely; “How has the team divided the stages of the project?” “How was the work split between you and the other members?” “How have you executed your part?”, “Why was that the chosen approach?” “How successful was the project?”, “What would you have done differently?”, “What is the team planning to do for the next version of it?”

Product: One of the main sources to understand seniority is to see how involved the person was in the reasons that made the team work on those projects (vision, strategy, problem sizing, discovery, etc). “Why did the team start working on this project?”, “I understand the team lead proposed it, but why has this person proposed it?”… go on until the person doesn’t know. It reveals how involved they are with the company strategy. The longer they can backtrace a demand, the higher their scope. “Was everyone in your team involved in this project?”, “What other members of your team were doing beyond this project?”, “How are these other projects related to the one you have worked on?”

When you surround a question with environmental facts from previous answers, it is more likely to get good answers for direct inquiries and make the interview flow more natural. After describing the team involved, their role in the project, and what they did, responding to “What have you implemented in this project?” will be way better since the candidate will instantly clarify the grey areas. “I deployed the model…I mean, the engineer did the deployment. I’ve written the Python code the engineer deployed”.

Do not shy away from insisting. When something is unclear, say, “It is still unclear to me how the team decided to prioritize this project”. You want to hit the boundaries of their scope and understanding.

Some thumb rules: The more conscious the candidate is about the end-to-end (company strategy -> OKRs-> discovery -> project -> execution -> impact -> how to do better -> what is the vision for the future of this workstream), the more senior they are; If their team demands them to define “how”, like “we need to build X to tackle Y”, it points to Senior+; If their team demands them to define “what” and “how”, like “we need to improve X, what should we do?”, it points to Lead+; If they mainly execute in a team with other DS/MLEs, the person is probably more jr; If they split work unequally with other people in a complex project and their part is more high level, it pushes to the Lead+ direction; If they have received the project description like “apply technique X in the Model Y”, they are closer to Mid-level-;

As you understand the scope of the person in the current company, they will likely have an offer to perform at the same one at your company, or if they are in the end part of it, they could be hired to perform in the following scope - just as you would have decided to promote someone internally if they describe a similar story during a performance cycle. If you are considering “promoting” the person to join your company, be consistent with the risks you take internally. You should have a couple of internal examples with similar technical skills and behaviors being promoted inside the team. Promoting a Senior DS/MLE from another company to Lead at your company will grab the attention of all the Sr. DS/MLE around that person. You should feel confident announcing it to the team as an internal promotion.

After the HM interview, you define the candidate level range. E.g., [Senior, Lead]. Perfectly, it is not a range but a single level. It only considers what makes sense for the candidate’s career regardless of their performance in the technical interviews - we will add this information later.

Debrief

The debrief is when everyone involved in a candidate’s hiring process gets together to decide. The HM should be active during it. At this point, you want to make two decisions:

  • What is the new hire’s seniority (in your company’s perspective);
  • Extend or not the offer.

However, neither might be openly discussed during the debriefing. It depends on your company culture. Some fear that interviewers won’t have the right incentives to support the leveling decision directly, and since the HM is the only one with the full context of the position, the others won’t contribute to the offer decision. Your role is to gather as much information as needed for these decisions. At the end of the debriefing, you should have the answer or guarantee you did your best to have it by then.

Both decisions are very intricate.

Before the debriefing, you will define the feasible offers you will validate during it. The viable offers are the intersection of two ranges:

  1. The position’s level range: the exercise you did at the kick-off stage.
  2. The candidate’s level range: defined in the HM interview.

If the conjunction is sole “Senior”, avoid discussing attributes that justify hiring someone as a mid-level or Lead. If, for a candidate’s career, the only thing that makes sense is a Lead role, for example, it is not helpful to spend the debrief talking about why it is a yes for a “Senior”. Skew it to talk about Lead’s scope. Even if the candidate had a “Yes” for all the cases, if they cannot perform at the Lead level at your company and it is the only thing that makes sense for them, that’s a no for an offer.

It would be great if, in your company, the interviewers could express the seniority associated with their yes/no in the candidate’s evaluation. E.g., yes for mid-level, no for Senior+. When interviewers do not specify seniority, your work during the debrief is even more critical.

There is no single script for the HM during a debrief because you consider a specific candidate and position. Here are some helpful questions that go deeper into what happened during the interviews (supposing the interviewer hasn’t covered it in the scorecard) or request the interviewer to generalize if the candidate has the needed skills for a task in the scope you are looking for:

General: “How well could the person understand the problem offered in the case? Have they made questions to clarify it?” “How did the person react when they did not know the answer?” Check on Junior/Mid-level: “This person has a CS/Engineering degree. Have they followed best practices from Programming 101?” “When you identified they were struggling in part X of case Y, have they used your tips well to get out of it?” Check on Senior: “Would you trust this person to be the only Data Scientist/MLE in a new version of [main model in the interviewer’s team or in the team the person is hired for]?” Check on Lead+: “(MLE) If we need to refactor a library, would you trust this person to lead the design? “ “Would you trust this person to define what to do to fulfill a business need?” “Have they shown any concerns beyond modeling and engineering during the case?”

The possible offers are at the intersection of the candidate’s level range (defined by the HM interview), position level range (defined by the team hiring or HM), and skills level range (defined by the hiring cases).

What if there is no intersection? If it fails because of the candidate’s level range, consider asking them if they would accept a lower scope at your company. Or, during the offer, make it clear the scope you are offering is higher than what they are operating or looking for.

The offer

This process changes from company to company, but you will likely craft the offer with the support of another person or have your manager do it (if you are in the first levels of management).

There are usually safeguards from the company to avoid doing something wrong, like salary bands and ranges for flexibility.

Compensation is the last piece of the puzzle. We will add it to the possible offers coming from the hiring process. It shouldn’t change what the team initially envisioned for the candidate, but it might change which level in the compensation range we’ll pick.

It is essential to have the possible offers without considering comp to avoid feeling tempted to force a higher scope to make an appealing offer or be conservative and downlevel it.

If the candidate showed a solid performance and a few experiences from a higher level, both of them should be present in the possible offers. The HM should avoid promotions, termination, or resignation in the first cycle of the new hire. However, it is evidently better to be wrong and promote early than terminate early. Only make the offer a promotion if it is similar to the risks your company takes when promoting someone to that level internally.

The job offer call is an essential part of the process. Decide if you want to join it in a case by case. Good situations to be there:

  • The level we’ll offer is not what the candidate is expecting;
  • We’ll offer a higher scope than what they have currently, and we want to clarify it;
  • The compensation does not match their expectation;
  • They work as a competitor for Data Science talent.

Think in advance to personalize the pitch. Now you know most of the candidate’s preferences and concerns, you can highlight what we offer that matches them and be clear about what doesn’t. Align with the TA member how you are going to structure it. A recommendation would be:

  • TA and HM congratulate on the offer;
  • HM refreshes the many pitches made during the process but as personalized as possible: the company, area, team, and the specific role the person will have in that team;
  • TA presents the offer;
  • Q&A.

In the pitch, connect the public company values and goals to what the candidate cares about the most. And progressively show how they are expressed in the role you are offering. For example, “As you know, here we are obsessed with customer satisfaction, so we treat customer service as a great differentiator. To scale it, we need to build models that offer automation without compromising our very high standards. You are going to work exactly with these models”, or “We are growing to multiple countries and offering more and more products. We have a fast-growing Data Science team. It is part of our vision to offer a platform that empowers the many users we have in the company. You will work in this platform and as your first project…”.

Another way to motivate during the call is to link their career goals to the opportunities of the position. “In this position, we expect you to lead the design of our solutions in projects such as our e-mail auto-replier, which is something you showered interest and skills during the process…”.

Reveal only public information. Be as specific as possible about the role. If possible, tell the candidate about the roadmap of their team.

What if they say Yes

Yay. Congrats. Announce their start date to the other Data Science people who’ll work with that person and their clients in their team (team lead, etc.).

What if they say No

You should know how likely the offer will be accepted during the hiring process and offer crafting stages. If we knew the chances were not very high, and you know precisely the reasons - other offers, compensation, a promise to be promoted in the current company, the common best thing is to let it go, with the assumption you did all you could to cover the possible reasons.

If it surprised the team, and they did not specify in the rejection e-mail, try to contact them to know the reason. Do not insist if they are not open. At the most, offer more clarification about the role and make minor adjustments to the offer. I’d recommend that a counter-offer should, at most, have a small change to match another offer or be backed by specific concerns - changing locations, providing to the family, etc. If the candidate’s concern is a general “the more money, the better”, that’s the wrong motivation, and they will always have a chance to leave the team for a 20-50% increase and cause future problems. Remember the Winner’s Curse: when there is competition for an asset of uncertain value, often the winner pays more for than it is actually worth. It means you shouldn’t try to cover other offers at all costs, especially if it is from a third company and not the candidate’s current company. The only justification is if you have an advantage that will make that set of skills particularly valuable, so having that person in your team has a value the other companies can’t perceive.

Taking smart risks

Being very conservative in a hiring process might be considered “having high standards” by some people, but it can be detrimental to the overall quality of your team. Conservative usually means evaluating people based on what your team already knows and setting an equal bar for every competence regardless of the candidate’s background and the position they will fill.

Efforts for diversity and inclusion can be confusing to Hiring Managers. Some will act defensively and use the traditional “I will hire the best candidate regardless of who they are”, which is hiding from being part of the solution since it will inherit the lack of diversity from the previous stages, like in the universities. On the other side of the spectrum, some will “give an opportunity, even if they failed a few cases, to bring diversity to the team”, which seems optimistic, but it can set the new hire up for failure, harm their career, and it might not be the best person to give an opportunity. People from minority groups reported to me the burden of thinking they were a “diversity hire”, so it is important to do it in a way they will realize when doing the job that it was because they are great at something valuable for the group.

Taking smart risks means assuming a risk (a future dismissal) in exchange for the chance of a compatible outcome (bringing new skills that will be impactful, increasing the team’s diversity with a strong hire, etc). The first step is to be aware of the risk and its size. However, it is also interesting to note the outcome: I don’t consider it smart to assume a high risk to bring someone who, in the best case, will bring the same knowledge and social background your team already has.

Instead of a plain evaluation, evaluate the candidates given their background, experiences, and previous opportunities. For example, if a candidate works as a Data Scientist but has a background in biology, it is expected they had no formal programming training and their coding experience is shorter than those of Computer Science and Engineering backgrounds candidates. In this case, some mistakes are acceptable in a programming case. However, having a candidate with a CS background and a couple of years of experience in a good tech company making the same mistakes might be a red flag.

Another strong trait that helps to decide on whom to take a risk is their evolution during the interviews. Candidates who learn during the interview can discuss, ask good questions, learn a new concept, apply it, and move forward.

Focus on the strengths of the candidate. Could the DS with a background in biology surprise the team on any aspect of the field? Maybe they are amazing in experiment design? Especially in large companies, Data Science professionals might have a very different challenge requiring different skills.

This approach will favor and help bring more diversity to your team, and it is an actionable way to evaluate people’s potential: when someone shows they excel in one dimension, they are more likely to excel in another when provided with the right resources and opportunities. Further, depending on the position, they will likely be more valuable to your company by offering strengths instead of a flat skill profile.

Additionally, pay attention to the team this person will work on. Not shining in a particular skill that’s highly needed to succeed in the context of the team they will join requires this team to have people with slack to mentor the new hire on that skill.

Common anti-patterns

  • Apply a simple rule over the interview results: A hiring manager wouldn’t be needed if the criterion to hire someone was having a “yes” in every interview. Be active and intentional during the process and reflect on the possible trade-offs and associated risks.

  • Hiring a Staff+ from another company to be a Senior- in yours because of their performance in the technical part: if a candidate is in a technical leadership role but they don’t have the technical level to exercise it in your company, don’t try to bring them to a previous career stage if they are not open to. Be careful about what they say since your company might be able to offer more money for a Senior role than to their current position, and they can say “yes” if questioned directly, but it is a “no” when it comes to the day to day, and they will avoid “Senior work” and try to contribute in their old scope. When the companies are clearly in different stages, and the candidate is moving from a company without much Data Science culture, they are usually more open to it and conscious they wouldn’t be offered the same position they have.

  • Hiring candidates with gaps but not setting clear expectations to overcome them, including a notion about time: when taking smart risks, you might hire someone with a great core skill for the job but lacking another important skill. The result of a hiring process is the first feedback of a new hire, and it should guide a development plan and expectations on overcoming that gap.

  • During the debrief, no one is excited to bring that person to the company: excitement is not very objective, but a debrief full of low energy and without highlights, but a series of “okays” and no red flags is usually not enough to hire candidates. The lack of a “no” in these cases might be tempting. Candidates shining in at least one dimension can easily use that strength to excite the team around about having them in the same team and investing in growing together.

  • Hire candidates with gaps to a team without someone who can teach them to cover them: not communicating the gaps and allocating people to support them will put the new hire at a higher risk of not adapting.

  • Not paying attention to what a new hire communicates to the whole team: every new hire is a message to the current team. It will show the career level in practice, which gaps are acceptable to perform at a certain level, etc. Be very conscious if the new hire communicates an inconsistency in how you treat external and internal people.

If it is your first time as an HM

Request your manager to review your requirements and desired profile, and bring them your hiring challenges to get other points of view as you calibrate. The same thing goes for the offer. Request your manager to be there. Let them do the more general pitch while you do the specific. Having someone more experienced in the first 1-2 times you do every step here will accelerate your learning.