Why The Future of Work Does Not Include Traditional Resumes

Standardizing our understanding of skills already proven on the job could be an equitable way forward

Resumes are great for guiding conversation at an interview. But they are pretty terrible for deciding which candidates should make it to that interview. Unfortunately, resumes are still the very first thing that filters candidates out in almost all hiring efforts. We are constantly overlooking the right candidates because of a bad data source.

The other day, I spoke with some friends who are currently looking for work. They've had great careers, and are fortunate enough to be financially secure. In order to get noticed by employers, they hired a professional service to optimize their resumes and cover letters. This service cost over $1000 for each of them, if you can believe that. The results were just as surprising as the cost: unformatted blobs of text, designed specifically to get noticed by resume parsing bots.

Their experience points to three of the biggest issues with traditional resumes: an inherent lack of equity, an overwhelming amount of text for hiring professionals to sift through, and a lack of trustable data.

Resumes are inequitable. Most people don't have $1000 to spend on a resume. Some services even recommend writing a unique resume and cover letter per application, further adding to the costs. Candidates who can afford these services have a significant advantage in every single job application, and that advantage has nothing at all to do with how qualified they are for those jobs. One candidate can make it past resume hurdles, through to the interview, and get a job offer while another candidate with identical skills but a different financial situation will be automatically filtered out at the very first step.

Resumes represent tons of non-standardized, incomparable data. Consider a popular job posting that conservatively gets 250 applications. If a hiring professional spent just one minute on each of those resumes, that would be over 4 hours of their time. And for non-standardized data like the text in resumes, it's very difficult to even compare two resumes side by side in a meaningful way, much less compare one to another you read 4 hours ago. Recruiters now spend an average of just 7 seconds on a resume, quickly parsing out relevant keywords, so it is no surprise that resume bots that do a similar job became popular.

Resumes are made of the least trustworthy data in the entire hiring pipeline, but they are used to filter out the most candidates, as much as 75% of all applicants. All those decisions are based on a personal marketing tool with no vetting. The source of the data isn’t even known: it could be written by someone other than the candidate as mentioned earlier, or even very effectively by a chat bot. So…a bot talking to another bot to make a decision about a human. Sounds like a Philip K. Dick novel. This lack of trustworthy data has had the unsurprising result of leading hiring professionals to look to proxies for talent. These proxies include requirements around education, previously held job titles, and tenure. All three bypass the most important metric, which is proven proficiency in the skills required for a job. And these proxies for talent automatically filter out half of the entire qualified workforce, disproportionately affecting underrepresented groups. Removing these hurdles not only doubles your talent pool, but also naturally allows for a more representative workforce.

One possible way forward is to standardize our understanding of skills already proven on the job in a way that allows all candidates to show up equally. Fortunately, an excellent data source for this already exists in the proven track record of candidates, as observed first hand by those they have worked with. If we can harness this data that is already being generated every day on the job for both technical and non-technical skills, we can build a globally-available, dynamic skill profile for all workers.

There are a lot of moving parts to a solution like this, but it can and should be done. We’re still learning every day at Merify, but we've made a lot of progress down this road with some common sense tactics. I’ll follow this post up with details on the how our products aim to build this new skill language that is equally beneficial for candidates, employees, recruiters, employers, continuing education providers, and more.

My friends should not have had to pay to have their resumes written in a bot-friendly way, and people without the means to pay for services like that should certainly not start every job search at a disadvantage. Hiring professionals should not be forced to start their talent pipelines with unverifiable, non-standardized data. Relying on a community-built understanding of skill proficiencies based on observed, on-the-job performance could be the answer.