How I screen CVs

Introduction

Screening CVs is a critical step in the hiring process, used to determine which candidates progress further in the hiring funnel. While the approach to CV screening can vary by company and industry, the primary goal remains the same: to identify candidates with the potential to succeed in subsequent interview stages.

In my experience, CVs serve a crucial role in the initial screening process of candidates, determining who progresses further in the hiring funnel. However, the way CVs are used can vary significantly depending on the hiring scenario. For instance, hiring managers at high-profile companies like OpenAI or SpaceX may receive an overwhelming number of applications from highly qualified individuals. In such cases, they might discard the majority of CVs by default, only considering those that stand out. Conversely, hiring managers at average companies might use CVs to filter out the bottom 60-80% of applicants, advancing the top 20-40% to the next stages. Ultimately, regardless of the initial screening, candidates’ progression in the hiring funnel and their final outcomes largely depend on their performance during interviews, not solely on their CVs.

Funnel

The macro view of my hiring funnels is as follows: my recruiter, with whom I've worked for years and is well-calibrated to our needs, sends me batches of CVs to review. My job at this stage is to select the CVs that will advance into the hiring funnel, where I screen out the bottom 60-80% of applicants, allowing the remaining 20-40% to move forward. Specifically, in Data Science hiring, the funnel typically has four steps (60 minutes each), with an approximate conversion rate of 1 in 3 at each step, resulting in an overall conversion rate of 1 in 81. Taking into account the CV screening step itself, I review about 250 CVs for each hire, a process I consider quite selective. Over the past four years at my current company, I've hired around 20-25 people using this method, and I'm satisfied with the strength, performance and culture of the team.

Signals

When screening CVs, I look for a clear, concise summary of the candidate’s work experience that sends a compelling message within 30-60 seconds. A well-crafted CV should be easy to read and parse, much like technical writing. Here are the key factors that lead me to discard a CV:

1. Typos: If you can't proofread your 1-2 page CV, how can I trust you to produce accurate Analytics and Data Science work?
2. Poor Formatting: Similar to typos, poor formatting indicates a lack of attention to detail.
3. Inconsistent Formatting: Consistency is crucial in technical documents, and is a signal of quality of subsequent technical work.
4. Excessive Length: A CV longer than 3 pages often shows a lack of understanding of its purpose.
5. Unbelievable Experience Claims: Listing superficial experience with many technologies can be a red flag.
6. Short Tenures: Frequent job changes, with a median tenure of less than 18 months across several jobs, can be concerning.
7. Incompatible Technology Stack: If a candidate’s experience is with a different tech stack, they might not be the best fit (eg. Microsoft stack vs. open source stacks).
8. Mismatched Seniority Level: A senior role requires more than five years of experience.
9. Irrelevant Experience: Experience that doesn't add value to the team or duplicates existing roles (e.g., certain contractors) is less appealing.
10. Technology-Focused Descriptions: Emphasis on technologies used rather than the impact or outcomes achieved is not impressive.
11. Irrelevant Low-Level Experience: Skills like MS Excel shouldn't be listed on a CV for a Data Scientist.
12. Lack of College Education: While not an immediate disqualifier, candidates need good work experience and an online presence (Github, blogs) to compensate.

It’s worth noting that I don’t automatically reject a CV based on a single factor. Instead, I assess the overall impression and make a final judgment call.

Factors I do not consider when discarding CVs include:

1. Obvious discriminatory factors such as gender, race, nationality, etc.
2. Attending an unknown college or university: Many people, including myself, didn’t attend top universities, so I don’t hold this against candidates.
3. Working at unknown companies: The same logic applies as with education.
4. Working gaps: Taking a few years off, for whatever personal reason, is completely fine. More power to you!
5. Non-native English sentences: As long as the CV is typo-free, minor language issues are fine (but not great, since many free tools are available to help).

Conclusion

In conclusion, for most companies and candidates, a CV is not intended to secure a job outright. Instead, its purpose is to clear the initial stage of the hiring process and demonstrate that the candidate has the potential to succeed in the interview stages. A well-crafted CV should present a clear, concise, and relevant summary of the candidate’s qualifications and experiences, showcasing their suitability for the role.