Why The Resume Won't Die. Or Conform.
2024 Trends in Hiring
AI and skills-based hiring are the two biggest trends impacting hiring in 2024.
Employers already use AI to scan thousands of resumes and screen for quality candidates. Job seekers use AI to write resumes and auto-apply to thousands of jobs in minutes. These tools lower the cost of applying for jobseekers and screening for employers to the point where not using AI to find and apply to jobs or help sift through the glut of resumes becomes a massive disadvantage.
Skills-based hiring is a growing shift away from favoring candidates whose degrees are from prestigious institutions or who have previous job titles at impressive companies to candidates whose skills align with the job description. The focus is on specific skills a role requires and finding candidates whose skills align, regardless of past experience, which leads to a better job fit.
On the surface, a shift to skills-based hiring plays to the strengths of machine learning algorithms. After all, pattern recognition is at the heart of artificial intelligence, something that is easier when things are broken down into component parts, such as skills. The algorithm can look for a unique combination of skills rather than a specific previous job title or work history, expanding the candidate pool and surfacing candidates with diverse backgrounds that could contribute to otherwise unexplored innovations for their new employer.
The Promise of Technology
In a dream scenario, for both candidates and employers, a global hiring system that matches jobs and candidates based on skills, without the need for candidates to submit their information from scratch for each new application would be ideal. Employers would no longer have to rely on recruiters or post their jobs multiple times on multiple job boards. Instead, they would simply post it once and instantly identify all candidates that meet the qualifications. In a perfect world, job seekers complete one profile listing all their skills, and that one profile becomes their application to any and all jobs.
So why doesn’t this dream scenario exist? Aside from concerns about creating a monopoly from a legal point of view, there are also concerns about differences in industry requirements, specifically around the context in which skills are used or will be used.
At the moment, most candidate-sourcing tools operate at the level of keyword fidelity. Indeed and LinkedIn will allow you to search for terms like “engineer” which will return civil engineers, mechanical engineers, industrial engineers, environmental engineers, and software engineers, all in the same search. Searching for soft skills will also turn up an overwhelming number of matches. After all, 98% of candidates consider themselves “team players” despite the lack of clarity or consensus around what that actually means.
Systems that pull skills from resumes or profiles and list them or allow searches on them don’t yet have the ability to consider the surrounding words and their collective meaning. An AI tasked with flagging every candidate with the term “statistical analysis” in their resume makes it hard for an employer to understand just how a candidate used that skill in context. Turns out, context matters. The stakes and conventions for baseball statistics vary greatly from the standards required for statistical evidence in research papers, or to make predictions for risk analysis, all of which fall under statistical analysis.
Resumes Provide Context
So while AI can help screen, it can’t yet perform the critical task of evaluating the level of expertise one has regarding a skill, what contexts they have applied that skill in, and how successful they are at bridging that skill with other skills they have to complete problems requiring multiple skills used in tandem. A good resume can include enough details for hiring managers to understand past performance and the ability to accomplish certain tasks even if the “keyword” skills were not directly mentioned.
For example
Recommended by LinkedIn
This single bullet point from a real resume implies several skills an AI wouldn’t have recognized such as “process optimization”, “stakeholder management”, “writing”, “editing”, “regulatory compliance”, “data analysis” and others which would all have been used to accomplish this one outcome, none of which would appear as a searchable keyword for an AI query.
The resume remains our best tool for putting these skills in context to better evaluate if a candidate should move to an interview stage or not.
Can Context be Standardized?
If resumes are the best solution we currently have to put all these skills into context, why can’t we standardize resumes so they are more easily searchable by AI and machine learning tools? Why not move to a single profile that allows for such context?
LinkedIn was built to serve this purpose, to provide a standardized profile that would contain all of your work history and be searchable to employers and recruiters (for a fee). So why hasn’t LinkedIn, as popular as it is, replaced the resume as the most common way to apply for a job?
The answer, in part, is the over-abundance of data on a single profile. Hardly anyone’s career is linear, and many change drastically. Someone reviewing a LinkedIn profile may see jobs in various industries and different job titles, and fairly ask themselves, “Wow, this person has done so much, but I can’t for the life of me figure out why they want to work here? What’s the connection?” This can be especially true of career changers and those with unemployment gaps.
The resume is how a person can showcase only the relevant parts of their experience in a few pages that provide the information a hiring manager needs to evaluate their fit for the specific role. The resume removes the burden of search and connection-making from the recruiter or talent team so that the most relevant criteria are easy to identify, making their job easier.
If the format is fixed and rigid, it becomes harder for candidates to showcase only the relevant information, and more importantly, it becomes harder to put that information into a context that will be easily understood within that specific industry or role. It is the very flexibility within the accepted norms that allows for great candidates to stand out, and similarly, for candidates who have less relevant skills, or, sadly, are less adept at highlighting relevant information, to be passed over.
A Hybrid Approach
As is so often the case, the best approach is a combination of existing tools. AI will not go away as a tool for sourcing and screening candidates. But neither will the resume. Working together, AI can reduce the number of resumes that a hiring manager needs to review, but it’s not yet able to make intelligent evaluations beyond a limited set of objective and tangible criteria such as degree level and prior job titles.
Some companies, like Scismic, are improving the limited keyword matching technology by using ontological mapping expanding the ability of algorithms to connect parent/child skills, commonly accepted synonyms, and alternative spelling of skills. All of this helps improve the quality of candidates that match and apply to jobs, but it will not replace the need for resumes, human screeners, or interviewers. At least not anytime soon.