Dear Readers: I wish to apologize for being absent for so long. The following post will explain why. I’ll do my utmost to resume posting on a regular basis.
A couple of months ago, a week before my 65th birthday, I was laid off from my job at a software company, along with about 200,000 other people in the volatile tech sector of the economy since January. I’ve had the usual support that comes with a layoff - a couple months of severance, Cobra, an outplacement service to assist with a new job hunt.
Throughout my career, I’ve had to resort to mass resume circulation only a couple of times: at the very beginning when had little corporate job experience, and in 2001 after the 9/11 Terrorist Attack in New York resulted in a venture capital firm pulling funding for my startup.
New jobs usually came my way out of connections made, and life has been good. This time around, much of my “network” is retired or moved on to companies and positions that are not in my field of interest of experience. So for the first time since 2001, I’m hitting the virtual bricks with resume and cover letter.
How things have changed! Never have I seen such a fine example of technology, culture and politics working harmoniously to squeeze every drop of humanity out of a deeply human process.
It has taken me two months and lots of experimentation with resume-scanning software to make a dent in my job search. For every listing in a site like LinkedIn, there are hundreds of applicants, in addition to those collected from other sources. For job seekers, technology designed to cull through the mass of applications and find the most relevant candidates has become a barrier to many qualified applicants. It’s not enough for a job seeker to be qualified; you have to learn how to work around a very dumb screening machine.
The Problem
Online employment advertising was just getting started back in 2001. In the late 90s when I was at Knight Ridder Digital, I helped architect one of the first online job boards, CareerBuilder, which was soon overtaken by Monster, and in the tech sector Dice, before LinkedIn emerged as the clear behemoth in the category, and their acquisition by Microsoft.
The technical challenge then was the same as it is now: scanning resumes created in various formats and language styles, with specific industrial vocabularies, to estimate their match against a written job description.
Tech entrepreneurs have always focused on the resume side of the problem, which poses ugly semantic challenges, especially for technical job markets. It’s easier to scan a resume for an accounting or legal job because the industrial vocabularies and job types are relatively standard and mature. But when you scan resumes for people in engineering, information technology, academia, manufacturing, construction, etc., you are working with industrial vocabularies that are changing all the time, and job titles that are harder to associate with actual skills and experience required for a job. For example, the role of a “product manager” can exist in both engineering and in marketing, each requiring a different set of skills for very different responsibilities.
When working on CareerBuilder, I realized that nobody had focused on the other side of the problem: the job description itself, and in the context of that project, we could not. Yet it remains, all these years later, at least half of the problem.
Having been a people manager at large organizations, I can tell you that most hiring managers are incapable of writing a good job description without the assistance of HR. Some get it, some don’t, and it takes experience. In the tech sector, HR may be of little use since they don’t understand the skill sets very well.
The result: Vague descriptions of job responsibilities, conflation of job requirements for professional experience and a host of other semantic issues that make it difficult to match a resume.
And so the online job board industry has never stopped growing. Today you see ads for Glassdoor, ZipRecruiter, Indeed, as well as dozens of parasitic job boards that scrape listings from other sites (many of them stale) to generate clicks for advertising.
It’s a complex ecosystem, charging employers different rates, and targeting the job market differently. The employer’s problem of finding the right people for the job, and the individual’s problem of trying to connect with a company where he can prosper, is far from solved.
Here’s my take-away so far: 35 years of applying technology to the problem hasn’t made it easier. For many jobs, it’s easier to find leads, but more difficult to get consideration than it was at the beginning of my career, when one relied on the telephone (voice) and shoe leather (presence) to connect with temp agencies to get in the door and recruiting firms for more senior positions.
The central actor in this ecosystem between employer and job seeker is Applicant Tracking Software (ATS). It scans online resumes for key words and phrases and quantifiable information that matches the language in a job description. Its use of search algorithms, semantic models and artificial intelligence (AI) has reduced the job hunt to a game of Buzzword Bingo.
Today, you don’t write a resume to accurately summarize your experience, skills and accomplishments. You write it so ATS will identify the common key words and phrases from your resume and a job description and get you past the screening software, so that you can tell your story to a real person.
Outplacement firms coach you in how to do that: Keep a Master Resume and from it tailor a resume for each job you apply to, based on the job description. My outplacement firm offers a link to JobScan, a web site that runs both your cover letter and resume through ATS to show you how well it matches up.
There are even ChatGPT-based (AI) services that will write your cover letters for you, using your resume and the job description as input. Another site offering resume templates and offers an AI feature that will write parts of your resume for you, tailored to a job description it analyzes from LinkedIn.
All for a monthly subscription, of course.
A partially free service, Teal, offers itself as the job-hunters’ AI weapon to fight through the semantic mess that ATS has created, helping you customize your resume to individual job listings.
Caption: A Teal resume assistant uses AI to re-write portions of you resume to better match a job description.
All of this to get you in front of a real person so you can tell your story. Looking for a job has always been a full-time job, but the more technology we throw at it the more miserable it gets, precisely because you never get to speak to anyone.
Buzzword Bingo
On the other side of the wall, ATS dashboards show incoming applications and ranks the probability of a good match. Beyond matching skill key words and phrases to job requirements, AI makes a lot of inferences and assumptions about what those key words mean, and what your job history means. Technology is weakest when it has to exercise human-like judgement, but that is now part of the mix.
ATS passes its data on to human screeners through a set of dashboards, like the one depicted below, who are expected to judge who should be advanced to a first-round interview.
Who’s evaluating those dashboards? Twenty-something entry-level HR people or consultants who don’t know what all those Buzzwords mean. If they see a Buzzword Bingo, they forward your application to an HR recruiter who’s been assigned to the hiring manager. They may or may not read your resume; they may just rely on the ATS report which has taken your carefully written text, flattened and summarized it for quick reading. If you’re lucky enough to get through to hiring manager who actually bothers to read your resume, you’re much more likely to get a call.
Of the dozens of calls I’ve gotten from recruiters who’ve scanned my resume in almost three months, only four so far have called me about a relevant job.
Many of those calls are from offshore search firms looking for onshore clients. Speaking with them is difficult, because you’re talking to screeners whose English is poor.
I have gotten calls from search firms asking me about jobs as a home nurse because my resume contained the phrases “health care” and “chronic disease management” taken from the following accomplishment I listed:
“Led the creation of Webex’s first health care solutions prior to the pandemic, starting with a tele-health solution integrated with Epic EHR, and later applying Webex platform APIs to engage patients in remote care or chronic disease management programs.”
This is how that statement is read by ATS and inserted into the “dashboard”:
Webex, health care, telehealth, Epic, API, remote patient care, chronic disease management.
Bingo! A nurse!
Short aside: when I told a former colleague that ATS had flagged me for a nursing job he replied, “If I saw you at the door of my hospital room with a bed pan I would pull the tubes and go out the window!”
In short, the whole process has been gamified by machine learning and completely dehumanized. It has left me wondering, “Do I really want to be in this business anymore?”
“Self-Identifying”
Something else has changed. The Equal Opportunity (EEO) disclosures at the end of every job application have gotten longer and more invasive.
Veteran status has long been included in self-disclosure. No one had a problem with it.
The madness started three decades ago with a paragraph that practically apologized for asking about one’s race: “The federal government requires us to collect information about the race and ethnicity of job applicants. You may volunteer such information below; it will not be used in evaluating our suitability for employment.”
Then the list of races and ethnicities expanded steadily over the years. They always include: Black, hispanic, native American, Asian, native Pilipino islander and white (or caucasian). It may also include Eskimos, people of southeast Asian descent and others.
I am one of those white males who never got to “check the box” until recently, when one job application included “Middle Eastern” in its list of races. I wanted to get out of my chair and cheer. Check!
No act of Congress mandates that employers ask about your sexuality, yet almost every application I have filed asks the question in different ways:
“Do you self-identify as one of the following genders?”
“Please identify your gender”
“What are your preferred pronouns?”
The first question is the least invasive. It includes male, female and “other” as choices, and mercifully doesn’t ask you to specify if you choose “other.”
The second offers as options the fake sociological categories (cis-gender, asexual, two-spirit, queer, etc.) invented by grad students in gender studies programs over the last 20 years.
The third lists the full range of fake pronouns the transgender movement has invented: SHE, HER, HER, HERS, HERSELF ; HE, HIM, HIS, HIS, HIMSELF ; zie, zim, zir, zis, zieself.
Some applications assure you that your response will not be used for selection purposes, but again, the language has changed. It used to be something like: “This information will not be used to determine your suitability for employment at our company.”
Now, it is something like this: “This information will not be used in making an employment decision.” One can almost hear the disembodied voice of a machine.
Only one application so far has asked the preferred pronoun question without giving me an opt-out. That’s not a company I would not want to work for, so I aborted the application.
Truly, I feel violated every time the question comes up. It is tantamount to asking me how I like to have sex. One wonders if they will get more specific as time goes on, in the name of diversity. After all, you don’t want a company where every “cis-gender” prefers the missionary position.
But sexuality is not the only intrusion. Thanks to the well-meaning Bush-era American with Disabilities Act (ADA), federal EEO directives now protect people with certain disabilities. The EEOC says:
“If you are applying for a job, an employer cannot ask you if you are disabled or ask about the nature or severity of your disability. An employer can ask if you can perform the duties of the job with or without reasonable accommodation. An employer can also ask you to describe or to demonstrate how, with or without reasonable accommodation, you will perform the duties of the job.”
This well-meaning protection is put into practice on the application form by asking an applicant to attest to one of the following list “impairments,” with a kind note that your prospective employer may not exclude you if you can perform the job with a “reasonable accommodation.”
Deafness
Blindness
Diabetes
Cancer
Epilepsy
Intellectual disabilities
Partial or completely missing limbs
Mobility impairments requiring the use of a wheel chair
Autism
Cerebral palsy
HIV infection
Multiple sclerosis
Muscular dystrophy
Major depressive disorder
Bipolar disorder
Post-traumatic stress disorder
Obsessive-compulsive disorder
Schizophrenia
I have bolded “impairments” that are fairly common and arguably less debilitating than the others depending on stage and treatment. While I can understand why an employer needs to know some of this, is this right right time and the right way to be having that conversation?
“Johnny, meet Jane. She is applying for the job you posted. She has epilepsy – but don’t hold it against her, OK?”
There is really no way to know whether disclosure of this information will be used to protect you or quickly exclude you. In many companies, HR coaches hiring managers and interviewers not to make written records of candidate evaluation in case of litigation.
By placing the disclosure at the beginning of the process during screening, rather than later when some determination has been made about the candidate’s suitability for the job, it is easier for employers to ignore candidates with disabilities – and less detectable.
As with other disclosures, you may “decline to state,” but that is sort of like pleading the fifth.
If you play along, at the end you are asked to certify that you are being truthful, and warned failure to be truthful could result in termination of the application process or subsequent employment (unless you are Liz Warren).
Am I being paranoid? I don’t think so. In 1994, I was recruited by a friend to apply for a Director position at a leading publisher of children’s classroom magazines. I knew people inside the company, and my interviews went swimmingly. The hiring manager was optimistic, and sent me to the HR Director for a final discussion.
I was greeted by a young woman with a mirthless smile who never rose from her desk to shake my hand, but instead told me to take a seat. She gave a two-minute lecture about how her company values diversity in the workplace and finished by saying, “The last thing we need around here is another white corporate male.”
I tried to handle it as a trick question, laugh it off, and start a real conversation. But it was clear she was not interested in having a conversation. “Thank you for your interest in (redacted),” she said when she finished, and we were done.
Similarly, the problem of age discrimination, especially in tech, is real, and HR professionals agree. And the prevailing technology enables it.
Outplacement firms will counsel you to not list job experience older than fifteen years, no matter how strong, and not to include your graduation dates on your resume or online profiles. There are two problems with this.
First, it cancels the most important thing that older applicants have to offer: experience over time, technology and change, putting all applicants on the same business experience footing.
Second, job applications require college graduation date, and that means that during the critical screening process, screeners and HR recruiters will not be blind to your age. As with disability, this makes it easier to discriminate – sight unseen.
In one of my interviews with a San Francisco startup, I had a good discussion in a Zoom room with men and women of almost every race listed on the EEO disclosure – none of them out of their thirties. During the chit-chat at the beginning of the meeting, one of the men asked if I did, in fact, graduate from college in 1980, as it said on my application and consequently, in the ATS report. “I gotta say,” he said, “you look about ten years younger.”
I tell those two stories to make this point: I encountered race, sex and age discrimination personally. It was in front of my face, which allowed me to see what was going on, deal with it directly or indirectly, and calibrate my path.
Had it occurred during behind the firewall of the ATS screening process, I would never have known.
Enter Social Media
While ATS software has been around since the '1980’s, LinkedIn was the first to apply social media to the problem of career development and job-hunting. The approach has clear advantages: the ability to build a professional network and grow it, the ability to to post your thoughts and react to others as a way of expressing yourself and connecting with peers. But as you’ll see, in many ways it suffers from the same problem as ATS technology.
Going on LinkedIn every day to look for work is a little like going to a party full of happy, kissing couples the night after you broke up with your girlfriend.
A typical LinkedIn feed is full of happy employees bragging about their latest accomplishment or training certificate; VPs and CEOs thanking their teams for their hard work, or celebrating the latest release of a product that is going to “change the world;” and loads of virtue-signaling employees celebrating Pride Week, Diversity, Animal Rights or whatever cause they think will endear themselves to the corporate community.
It’s the Facebook of the corporate world, and 50% of it is complete bullshit. And there’s a reason why: the LinkedIn algorithm boosts sentimental and cringeworthy posts, according to Rachel Garden, a social media consultant.
To help job-seekers get value out of a LinkedIn, a cottage industry of consultants, social media influencers and ghost-writers has sprouted up to help users control their visibility, the ATS-friendliness of their profile, how and when to post without appearing weird.
“Authenticity matters a lot less than being good a writing the specific type of posts that the algorithm is horny for.” (from the YouTube channel “Good Work”)
At GlassDoor, you see many of the same job postings you see at LinkedIn, but nobody goes to GlassDoor for the listings; they go there for the anonymously posted company reviews.
The site attracts more than job-seekers. Employees – current and laid-off – post scathing reviews when they are in peak depression, while HR induces employees to write positive reviews. Sometimes HR replies to negative reviews with mechanical bureaucratese: “We are sorry about your experience with Company X. We listen to a broad range of opinions and incorporate feedback to improve our employee experience.”
I used to check it out GlassDoor when considering whether to file an application, but found that almost every company had credible postings that the company culture was “toxic,” the CEO a “clown,” the work/life balance stifling, the “tech stack obsolete,” etc. Well, every company I’ve worked for has problematic leaders, silos of toxic culture, overwork, and technology is always a year or two away from obsolescence.
In the end, the problem is that the technology built around the employment market is largely based on an assumption of distrust. Employers think that applicants write their resumes to present a rosier picture of themselves, and so they use technology to objectify the applicant. Candidates think that employers do the same, and use social media to detect deception and opportunity. In fact, every company has problems; that’s why they need skillful people who are adaptive and risk-tolerant. Bringing them together is the opportunity.
Those of us who’ve done enough hiring know that resumes are written to impress, but sometimes discover a person is better than their resume when we sit down and talk to them. It’s a disservice to applicants and employers alike when technology prevents those conversations from ever taking place.
Technology could be used to enable those conversations with simple tools like video posts, blogs, Slack/Webex conversations (note to a startup - that’s a freebie, guys). But I fear that tech entrepreneurs will will propose a remedy to all this, and it will be the wrong one: Using AI to write job descriptions, LinkedIn profiles and posts – and conduct “virtual job interviews.”
Welcome to the Brave New World.
-30-
What a well-reasoned deep dive (Buzzword alert) into today’s personnel recruiting world.