No Location

It looks as though you are using Microsoft Internet Explorer which is no longer supported. To experience this site in the way it was designed, please upgrade to Microsoft Edge

News & Blog

The Hiring Landscape, Job Board Myths - A Realistic Look at being a Fifties Female Job Seeker


The Hiring Landscape, Job Board Myths - A Realistic Look at being a Fifties Female Job Seeker

  • 11/21/2021
  • Employment, Future of work, Job search, Interviewing, Skills Building, Technical skills, Technology, New tech, Digital, Ethics, AI & Age Bias, Workplace tools
  • Rosalie Day

The author Rosalie Day is an independent policy specialist focused on issues at the intersection of public policy, science and technology; energy; education and environment. She’s also an Fellow.

Let me start by introducing myself. My name is Rosalie Day. I have 19 years job experience in the energy and environmental fields and 2 plus years on the fast track as director, senior director, to VP in the tech industry. My career highlights include a bronze medal at my first job, and a position as climate change project officer to Romania –both at the US Environmental Protection Agency (USEPA); board chair of a regional power grid administrator in Nevada as a regulatory manager for a Houston-based company; at Silicon Valley startup, APX, developing of a successful early SaaS product and the first digital renewable energy credits tracking system; and co-founder of the Texas utility, Andeler.

These experiences came between my primary masters’ in public policy in 1988 and my added graduate degree in conflict management applied to technology issues in 2016. Since then, I have been certified in data privacy in 2017 and upskilled with two curricula in data science simultaneously with my job search, 2018-20. I can program in two languages, R and Python, giving me insight on text analytics and how algorithms are developed.

Since 2018, I have completed over 500 online applications for a new job. I haven’t gotten a single interview. Not one.

So, what’s the story? How does a woman who has achieved professional success and upskilled throughout her career not get noticed? From where I sit, the online jobs application process is flawed; it’s rife with built-in biases that keep experienced women from being seen.  

Online hiring systems emerged in the mid-1990s and became the prevalent mode of advertised jobs’ applications for large organizations a decade later. Now, it is a rare occurrence for any job posting to instruct the applicant to send a resume by email. Rather jobs are posted online and an apply button looms large on the actual job description. If the post is on a third-party site, the apply button typically leads directly into a portal on the employer’s website. Often, it seamlessly connects to an outsourced hiring vendor’s system. The application portal is part of a larger Applicant Tracking System (ATS) which manages the hiring process for the employer.

Automation, combined with the artificial intelligence generally used for hiring, reflects the current hiring landscape.

The shift to online hiring created a situation in which a job posting can be viewed almost globally. The same jobs appear on the organization’s website and on multiple job boards. Application submission is all online and is relatively easy to accomplish. These two attributes result in an over-abundance of applicants (50-2000) for each position.

Automation is used to address the problem of too many applicants, and too many hopeful, yet unqualified, applicants.

What exactly makes an applicant unqualified? Software was initially built to handle this large data body while filtering out unqualified applicants. For example, career gaps, unemployment for greater than six months, having graduated more than 20 years ago were some of the general “unqualified” filters used. The specificity of the filters grew to align with job descriptions.

The uber trend toward data-based decision-making over the last decade was put to the task of deciding the qualified-unqualified question. In an attempt to optimize hiring, predictive analytics -- another term for artificial intelligence (AI) -- is frequently used. This AI is intended to identify the ultimate job candidate by matching some particular attributes, the algorithm.

The details matter: a lot

With online applications, there typically exists a few methods for entering the data, depending on the nature of the job. They include:

  • uploading a resume
  • sharing data from a LinkedIn profile
  • and filling in the blank completion which includes identification, current and past employers, titles, years, accomplishments, and educational degrees

Some of the portals allow the applicant to review and correct how the automated system has populated the fields in the application, some don’t. In the latter, it is difficult to know what information an application contains.

A recent Wall Street Journal article (9/4/2021 by K. Dill) was entitled “Companies are desperate to hire, and yet some workers still can’t seem to find jobs. Here may be one reason why: The software that sorts through applicants deletes millions of people from consideration.” As this article discusses, there are hurdles on the hiring landscape which contain demographic biases. Here are some of the challenges I’ve observed with the online hiring process in my quest to find work:

  1. Skills and knowledge discounted:
  • The automation used by the ATS to read the data entered works by weighting topics and keywords. I think of it like densities. (See a word cloud for a visual parallel of how this works.) More experienced workers (45+) frequently have a larger variety of skills and knowledge. This often counts against the applicant.
  • With too many skills in the work experience, specific individual skill topics become low density relative to the entire body of words scanned. A resume with less years of experience typical holds fewer skill topics; they comprise more of the body of the text.
  • To illustrate, a job description mentions six skills (eg: analysis using excel; cross-functional collaboration; scoping a project; presentations to senior executives; self-starter; teamwork). There are two resumes which reflect 7 and 17 years, respectively. The 17-years resume will demonstrate an expert level of the 6 skills distributed across jobs with other skills also (because typically accomplishments across 17 years also require other skill topics). The 7-years resume only has accomplishments that can be achieved with those 6 skills. Who has the more developed 6 skills? There is not enough information for that to be determined. But the text analytics will pick the 7-years resume because it cannot deal with context.
  • Some automated scans directly discount a skill or body of knowledge for the time elapsed since the job (or associated accomplishment) entailed that particular skill/knowledge set. Whether that discount is appropriate depends on the skill, the knowledge and the individual’s context.
  • I have never seen an application portal that uploads the time-consuming, carefully-worded summaries at the top of resumes, or the list of skills career coaches tout. Very few applicant portals allow the applicant to enter skills potentially acquired elsewhere, i.e., not on the job. Skills-based resumes may have other benefits, but not in automated systems.
  1. Background compared to target algorithm:

An individual applicant is compared to an algorithm, constructed to mimic the background of persons they already know to be capable, fit, or have the correct set of skills. Target algorithms for particular jobs (or particular companies, or particular disciplines) are developed from the data about a set of individuals that already do that particular job (are at that particular company, or in that particular discipline).[i] The applicant may have all the right skills and knowledge to do the job even better, but may not be selected because they are not enough alike in background on others to be a predictable fit.

  1. Non-linear career path less valued:

Some automation and (human) recruiters treat lateral changes of industry as unknown because they don’t know how it translates. Some automation and some recruiters treat self-employment as a gap. Full-time training and graduate degree-granting programs are considered gaps if they are pursued mid-career path. There exists an incredibly flawed assumption for a gap (no matter what the reason, caregiving, sabbatical, education or several years in the job search) to be interpreted as a professional knowledge and applied skills hiatus. This creates a sizeable barrier in the hiring landscape.

  1. Limited focus on narrative part of application:

If the applicant makes it through this youth biased automation, then the 6- to 7-second review by the HR professional is where the summary at the top of the resume is read. Their bandwidth is spent on the current or last job, not upskilling or midcareer training - if that addition is not while in an associated job. Furthermore, if looking to do the same job in a different industry, it’s difficult to convey knowledge of the target industry with an online application.

After three years’ experience with the online application process, it’s abundantly clear to me that getting hired for a job in a large company in your mid-50s, especially if you have let your network go stale, or intend to switch industries, is almost impossible. Without a doubt, personal contacts are even more important because of the increased sizes of applicant pools. Also, I have found time and again, that I needed to be referred to the hiring manager, and not just by any company employee. A referral from a fellow employee (formerly unknown to the hiring manager) does not necessarily get an interview. 

I recently met a former Microsoft director, a woman in her late 50s, who had accepted an offer at Google, and was joining hiking groups in the interim. She said that tech companies fill jobs at the midcareer experience-level by crowd sourcing from their own employees. (The importance of who you know!) No-one wants to refer someone they have not observed fairly extensively - which makes people met on LinkedIn in an informal interview not viable candidates for referral.

As a woman who doesn’t throw in the towel, I’ll continue to work all the angles, even if they are sharp and uncomfortable, until I find my next place. I’ll work to expand my network, tell my story, and welcome advice like the following recommended by my alumni career development coach from University of Chicago Harris School of Public Policy.

  1. Interact with hiring managers in target organizations on a non-transactional basis (if they reply) - which requires patience, follow-up, and time.
  2. Send your elevator pitch in writing to Diversity and Inclusion recruiters in target companies (although few D&I programs specifically include age among their criteria).
  3. Develop a Twitter string of posts because LinkedIn is not seen by the correct specific industry people - if they are not already in your network.

It is easy to blame the automation because it’s on the front line. However, companies develop it, outsource it, procure it, and, most relevantly, use it. In my opinion, these organizations need to recognize the potentially career crushing hiring systems they have in place and revamp them to fully include ALL candidates.

  1. [i] That set of already selected individual is divided into training data and testing data. Algorithms are built from the “training” data which are composed of the same list of factors for each individual. Factors are attributes of an individual’s background; for example, an attribute could be over 5 years of work experience, or not. The algorithms are tested on the data of the remaining individuals, not used for its training. The important point is - both the training and testing of the algorithm is the set of individual already at that particular job, company, function.



Add a comment