Hiring Niche Talents in a Fiercely Competitive Market
Sr Software Engineering Manager at ThousandEyes
ThousandEyes was acquired by Cisco in August 2020. Post-acquisition, our organization has been fortunate enough to experience massive growth in our customer base. In order to sustain that growth, we knew that we had to invest in two areas:
- enabling our platform to scale horizontally in a cost-effective manner for the anticipated load
- building product capabilities that will support a wide range of network performance management use cases.
We converted these high level aspirations to a detailed set of deliverables and estimated what it takes to deliver them. From the top-down estimates, it was clear that we had to increase our headcount by 2x, bring in people with new skill sets and diversify the experience we had on the team. Ideally, we wanted to have ample time to ramp up the new hires to deliver on our planned initiatives before the end of the year, which meant we needed to complete our hiring by, well, yesterday! Our plan B was to hire eight engineers within the next four months, so they could be effective by the second half of our fiscal year. This meant that we had to hire at least two engineers every month, which was certainly a daunting task.
We took steps to supercharge every part of the hiring funnel, and as a hiring manager, I led that transformation.
It was very important that our top of the funnel was healthy in terms of both quantity and quality. I had an amazing recruiting team that was helping me to build this pipeline. To improve the quality of the pipeline they were generating, I had to spend time educating them on the technical requirements of the role, relevant experiences that will serve as a good proxy if we cannot find a direct skill set match, communities or organizations where the desired talent could be available, etc. As a follow-up, I would provide detailed feedback on the first few candidates they source until they feel comfortable sourcing entirely on their own. After I felt confident about the recruiter’s ability to source good candidates, I started sourcing directly.
I reached out to potential candidates within my first and second-degree connections on LinkedIn. These connections are typically more open for an exploratory conversation, which gives me an opportunity to share more context about the role. When that channel was fully explored, I moved on to cold reach-outs.
This was the phase where I spent most of my time sourcing candidates. Most hiring managers make one common mistake, especially early in their careers, which is starting with a general template. Sending those templates to 20–30 candidates every day will have a low response rate and does not produce a quality pipeline. What I did instead was understand the specific criteria that were crucial for my open roles, and reach out to candidates who just possess those criteria. Nothing more, nothing less. It is important to articulate why you think there is a strong match based on your understanding of the candidate’s profile. It is also helpful to highlight what the role offers in terms of technical or non-technical challenges that will help the candidate progress to the next level in their professional career. Curating an email with such a level of personalization will take time and you will probably be able to send 2–3 such InMails per day. However, the success rate of this approach was significantly higher than a mass outreach and far outweighs the volume benefits offered by the latter.
Technical Phone Screen
This part of the funnel can be significantly streamlined if there is data from someone who has worked closely with the candidate—someone, ideally, close enough to have reviewed their code. On the flip side, some candidates can be so convincing during their initial call you may feel tempted to skip the phone interview in the interest of time. Unfortunately, that almost always turns out to be a mistake. It either wasted the on-site interviewer’s time, or the candidate realized that the engineering culture wasn’t a good match for him or her, which could have easily been avoided with an initial phone screen. Lastly, it is a mistake to expect someone to solve problems that they would never solve on the job. Interview questions should be a very close match to the kinds of problems one solves on the job. This also saves the candidate from having to buy time for interview preparation. Lastly, any interview question should have a clearly defined set of rubrics that will help the interviewer assess the candidate objectively. The rubrics should focus on “how” the candidate solved the problem rather than “what” the candidate did to solve the problem.
We found that three ingredients are essential for an effective virtual onsite: 1. Diverse and qualified panel members; 2. Well calibrated questions with clearly defined rubrics (or success criteria); and 3. The sound technical infrastructure for a seamless interview experience.
Panel diversity is extremely important for evaluating a candidate’s technical and cultural fit from different dimensions. Depending on the candidate’s personality and background, there is a higher chance of getting a good signal on the evaluation criteria when combining feedback from all members of a diverse panel. This also gives an opportunity to highlight red flags that may not have been visible to certain members of the panel.
I made sure that everyone in the panel has a clear understanding of the questions and rubrics from other members in the panel. This enables panel members to fine-tune their questions to avoid overlap with other members on the panel. This also helps the hiring manager identify gaps in the interview questionnaire that should be filled in order to do an effective evaluation and reduce the chances of a follow-up interview.
Finally, the technical infrastructure to be used for the interview should be understood and pre-tested by both the candidate and the interviewer. This is where recruiters and interview coordinators can help straighten out.
In addition to the questions being asked in the onsite interview, each interviewer needs to be coached on extracting metadata about the candidate’s cultural fit for the organization and his or her motivation levels to join the organization. Yet, trying to extract these signals based on the questions asked by the candidate can often be misleading. This often requires reading between the lines and dynamic thinking on behalf of the interviewer to prod the candidate in the required direction. For example, If the interviewer brings up a major accomplishment made by the organization and talks with an assumption that the candidate has done their research and knows about it, then we can see if the candidate follows through or if they are unaware of the topic. This gives us a signal about the candidate’s motivation to join the organization.
We mandated a round for the hiring manager during the virtual onsite interview. The goal of this round was to ensure that the candidate is leaving the interview with all questions answered about the role and the organization. This is also an opportunity to be brutally honest and clarify what qualities make a candidate successful in the role. My experience has taught me to never try to sell the role to the candidate. Be honest about your challenges, and they will sound like interesting problems for the right candidate. The highest performers are the ones who join the organization for the right reasons and are personally motivated to take on the challenge based on all the information they have gathered. On the other hand, well-qualified candidates who were misinformed will feel demotivated when the realities offered by the job do not match their mental model.
This is the phase where all the best practices we’ve established so far should ideally come to fruition. The goal of this phase should be to extract a meaningful signal out of the interview and respectfully eliminate any derived feedback that isn’t backed by any factual evidence. I was able to accomplish this by aligning each interviewer’s feedback along with the set rubrics for their questions and to sum it up, I articulated my hiring decision by drawing a connection between the job requirements and the data gathered against the rubrics. It was also important to highlight which parts of the interview feedback did not hold any weight in the hiring decision. In addition to the rubrics, the hiring decision is also based on a certain bar, which we have always kept high enough to ensure that the candidate can land reasonably well in the role he or she is offered and will feel sufficiently challenged to advance to the next level. As a result, we could bring in high potential candidates who didn’t quite meet the bar for the published role by offering them a role that is more appropriate for their level of experience and offers room to grow to the next level.
If everything up to this stage was done right, it is best to let your recruitment partner drive the offer process, but be available to do a follow-up call with the candidate to answer any questions. I’ve occasionally pulled individual contributors from my team to jump on a call with a candidate to share first-hand experience of working on the team. It’s always advisable for the hiring managers to stay on top of the offer closing phase and follow up with the recruiting partner if things seem to be stalled. If it comes down to the numbers, as a hiring manager you understand the criticality of the role and should make the appropriate push to stretch the numbers if the business impact of delaying the hiring process is detrimental to the organization.
- For hiring managers, having a broad and qualified network will have a direct impact on the time it takes to close your open roles. Building a network does not happen overnight or even within a span of weeks. As an engineering leader, one must always be working on extending their network to strong engineers, regardless of their current hiring needs. It is also important to keep the network warm by touching base as needed.
- The requisition is not closed until an offer is actually accepted. As basic as it may sound, it is quite common to ease down on the hiring pipeline when a couple of very promising candidates reach onsite and/or even clear them. However, the market is extremely competitive and it has become quite common to lose a candidate in the offer stage.
- In a competitive market, it is a clever strategy to hire candidates that don’t necessarily have a 1-1 match with all your tech stack but have the ability to learn them on the job. However, it is important to budget the additional time they would need to ramp up on that tech stack, which depends on the team’s unique adoption of that technology and how fundamental it is to your architecture. Sometimes these become too significant to impact your hiring decision altogether.
- Ensure that the candidate understands every dimension of the opportunity ahead. Even if you feel that it will lower your chances of getting an offer accepted. More often than not, I’ve seen great candidates come on board to solve problems or challenges, rather than walk into a stable environment (rather boring). Also, chances of morale drain in the first few months are much higher when expectations don’t match reality.
Be notified about next articles from Sai Dhalli
Sr Software Engineering Manager at ThousandEyes
Connect and Learn with the Best Eng Leaders
We will send you a weekly newsletter with new mentors, circles, peer groups, content, webinars,bounties and free events.