How to Use AI to Find the Best Talent for Your Startup

09/03/2022

Finding the right talent for your startup is a challenging job. Whilst startup talent is in very high demand, supply of talent is relatively low. Posting your job on a job board alone is not enough to attract the right talent. Instead, you should be searching and reaching candidates, understanding their ambitions and personality, and articulating the contribution of the offered job to their career. This asks for a variety of skills like analytical skills, communication, emotional intelligence and broadcasting power. The promise of Artificial Intelligence (AI) is to make the life of the recruiter easier by supporting them in their tasks. But how much can AI really do? Profiles have to be screened and candidates have to be talked with. So how can you cut through the noise and go beyond the buzzword?

Why AI exists in startup recruitment

Since startup talent is scarce and hard to get, you have to look for candidates yourself in addition to an inbound approach like posting a job on a job board. Proactively finding and engaging candidates is often referred to as passive candidate sourcing. In passive candidate sourcing, success depends on the quality of your search, screening and outreach approach. You have to find the right candidate, screen for required skills and engage them in a way that’s convincing enough to respond. This process is very data intensive: there are more than one billion profiles across the internet to search through to find that one unicorn. AI is especially useful in this process because it can help get vast amounts of data, clean it and analyse it (filter it) so you only have to select the best talent based on the information provided. You can look at AI as a scalable way of supporting your decision making. It doesn’t replace human decision making, but it helps in making choices along the process in a variety of ways, for example, pre screening candidates, recognizing patterns in successful hires and providing prompts and suggestions to the recruiter.

How to use AI in startup recruitment

AI is primarily useful in data and analytical heavy processes and searching and screening are the most data heavy part of the startup recruitment cycle. So how can startups apply AI to find the best talent?

Your starter guide on using AI to find startup talent:

1. Know where to find your candidate

Candidates are not only to be found on LinkedIn. There are many other platforms where candidates hang out that have rich information about their skills and interests.  Take engineers for example, they usually don’t have the most up to date profiles on LinkedIn, instead showing off their code on platforms like GitHub and Stack Overflow.

Here’s a list of alternative platforms where you can find startup talent:

  1. GitHub: software engineers
  2. Stack Overflow: software engineers
  3. Reddit: any talent
  4. AngelList: engineers, sales execs and other startup talent
  5. Xing: talent based in Germany, Austria and Switzerland
  6. Kaggle: data science and Machine Learning talent
  7. Medium: talent with a voice

The advantage of using AI to support your recruitment process is that you can process large amounts of data from different places at the same time, without having to extract, parse and organize data manually. You can use a tool that does it for you.

2. Know the basics of how a matching algorithm works

Before you dive into using AI tools, it’s important to understand how searches for candidates work. Most search engines and platforms match candidates based on a combination of job title, keywords and other data, e.g. time at current position. Don’t settle with unexplainable algorithms and the story of some magic Machine Learning, most of the algorithms are still in the experimental stages so you need to understand how it provides you the matches so you can screen them properly.

The basics of an AI matching algorithm:

  • Most of them are primarily keyword based, meaning the keywords on a candidate's profile (‘Engineer’, ‘SaaS’, ‘JavaScript’, ‘Sales’...) are most important in determining the match.
  • Job title and/or headline are an important factor in determining your matches, some algorithms use it as a required variable and some as an optional variable since there are a lot of different job titles that describe the same thing.
  • Some matching algorithms use predictive analytics, like estimation of the likelihood of a candidate switching jobs.

3. Use an intelligent tool to find startup talent

There are different tools for different purposes. Which tool to use depends on how you have organized recruiting within your startup, at what speed you are recruiting, and the possibilities the AI tools bring that are within your reach. An AI talent search tool should:

  • Take into account synonyms of job titles and keywords. For example, an Engineer can also be a developer, programmer or coder. JS can be JavaScript. And ‘Sales Executive’ has around 30 job title synonyms.
  • Search different platforms in addition to LinkedIn. There are many underutilized talent pools available on the earlier mentioned platforms, your AI search tool should help you find those profiles on other platforms.
  • Enrich profiles with additional information so it does the matching based on the most recent and complete information available on the candidate.
  • Find public contact details. The advantage of most AI search tools is that it can find contact details in all the different sources that it searches, most likely faster than you.

Some tools do it all, providing a full solution for finding startup talent and their information. Other tools are specialised in part of the process and provide a single function. Examples of tools that help you in your search for the best startup talent:

  • HeroHunt.ai: AI search engine which finds tech candidates across platforms like LinkedIn, GitHub and Stack Overflow, enriches profiles and provides public contact details.
  • SourceWhale: AI powered hyper-personalised messaging engine which imports your selection of candidates and automates outreach.
  • Crystal knows: A chrome extension used to screen candidates for personality based on their LInkedIn profile.

4. Keep your human eye on the ball

Although very helpful, algorithms are not waterproof. Your human eye can see patterns that machines can not. This is why you should always look critically at the AI tools you use and test if results are indeed qualified based on your job. You can do this by reviewing the top candidates the AI tool returns to you and comparing them to a manual search that you do in your preferred platforms or through a Google X-ray search.   Yuma Heymans This is a guest blog written by Yuma Heymans. Yuma is the co-founder of HeroHunt.ai, an AI talent search engine specifically designed for finding the best tech talent on the entire web.

Share

Ready to learn more? Let's talk