Tips and tricks for fine-tuning your AI search

The following article is an excerpt from our latest e-book, AI Sourcing Simplified: How to Choose & Use Your New AI Assistant. You can download the full report here for all our latest data and best practices.

You may have heard the analogy that using AI in recruiting is like working with a junior recruiter. While it has the tendency to take things way too literally and it might not always get everything right the first time around, it’s an eager partner and very good at taking feedback.

But if you haven’t had the chance to experiment with this technology yet, you might be wondering exactly how this works. What type of feedback should you give your AI sourcing tool in order to get the best possible results?

How to source with AI: A few guiding principles

As with many AI tools, the quality of the input (written qualification) determines the quality of the output (profiles you see in AI Sourcing). Learning to write the best possible qualifications will take some time and iteration.

Imagine you’re training a new junior recruiter who has good intentions but does not understand nuance or subtlety. You need to be explicit and literal to avoid confusing them. What would you tell them to look for in a profile? Try to answer questions like: How would you describe your ideal match? Are there examples of what set apart the best talent in your mind? What specifically should the sourcer be looking for on a resume?

Once you’ve thought through how you’d answer those questions, here are a few additional tips to help you. 

Be as specific as possible

Try to avoid language that’s too broad, like “software development experience.” Instead, use more specific language about the length and type of experience, like “experience using React and Python on the job (not just on school projects).” 

Provide necessary information, but be concise

The longer and more complex your prompt is, the more likely it is to confuse your AI tool and lead to less relevant results. Instead of a lengthy prompt like “We need the prospect to have between 3 and 5 years of experience working as a product manager, as this is a mid to senior role that will be both customer and leadership-facing,” try to condense your prompt into something more succinct, like, “3–5 years’ experience with product management in B2B SaaS.”

Avoid subjective qualifications

Your AI tool will not necessarily know what you mean if you use subjective qualifications like “good” or “top.” Think of ways you can communicate these ideas more subjectively (in other words, things that would be easy for an AI tool to gauge from someone’s resume or LinkedIn profile). Instead of saying, “went to a good school,” for example, say, “attended a top 20 college for computer science based on US News 2024 rankings.”

Avoid vague qualifications about soft skills

Similar to the subjective qualifications that we mentioned in the previous bullet point, AI tools will struggle with terms like “entrepreneurial skills” that can be difficult to parse from a resume. Try to rewrite these qualifications in terms that will be easier to determine from someone’s profile, like “experience contributing to or leading new product initiatives.” 

Bundle similar or repetitive qualifications together

If you have a list of similar qualifications (for example, proficiency in a series of coding languages), you don’t need to list these out individually. Instead, you can create one qualification that lists all these skills together (e.g. “proficiency with Python, R, and C++”). 

Example: How to write a data analyst search

This example data analyst search shows how to put these guidelines into practice.

Let’s take a closer look at why this search works well:

Clearly defined roles and experience

The example search includes specifics like “at least 5 years of experience” and “job title contains ‘data analyst’ or ‘senior data analyst.’” These are clearly defined parameters about the relevant roles and experience that will help your AI tool filter in the right talent.

Provides necessary context but keeps it concise

For this data analyst role, using a phrase like “3+ years of experience in data analysis or a related role” is clear and concise while still providing the necessary context. You don’t need to outline all the other relevant job titles like senior data analyst, data analytics supervisor, senior manager of data analytics, data scientist and analyst, etc. because the phrase “or a related role” covers all that.

Scores similar qualifications together

Instead of creating separate entries for different data analysis tools, this job search includes all of them (SQL, Python, R, and Excel) in a single entry.

Quantifies soft skills

If you’re looking for a person who’s adaptable and comfortable with change, an AI sourcing tool won’t know how to interpret this or be able to derive this information from a candidate’s profile. Instead, consider what specific experience would be a good indication that a candidate possesses a specific soft skill. In this example, “experienced company growth through IPO phase” is a specific type of work experience that’s likely a good proxy for a candidate’s adaptability and still possible to discern from their resume. 

Looking for even more practical tips and tricks on incorporating AI in your sourcing? Grab a copy of AI Sourcing Simplified: How to Choose & Use Your New AI Assistant.