Is Job Search Really Difficult?

Is Job Search Really Difficult?

Maybe it is when the tools are monetized

·

13 min read

I have been around for a while. Back in the day, job post sites (when sites became a thing) were here and there, and then Monster came along. Add Dice, and maybe a few others, later Craigslist with jobs and gigs, HotJobs, a few others. Real money and energy put into these things, never quite meshing with the employer side of the equation. How many times do you need to submit your CV as a Word document so they can re-parse it into their own database, how often was your contact info stripped away so some headhunter could take their cut, how often were we re-filling the tired Add New Experience... items over and over and over.

In about 2002 or so, I found myself living in Phoenix, after my main stint in Silicon Valley. I was looking for jobs post-9/11, which was obviously not planned, and the job market (particularly for remote consulting) was all but evaporated. Non-stop posting of some form of job application via these various job sites was a painful and almost certainly fruitless experience, because who knows what happened to my submission when I hit the form button here, there, this site, that site. The same may be true on the hiring side, though in an employer’s market with professional recruiters paid to do their jobs, the risk is less. Still, the message, the brand, and the emotional attachment on either side are filtered out in the piles of job listings and CVs and bad algorithms.

The cover image for this post is likely not sized correctly. That is a topic for my next post, perhaps, but for now some detail: Indeed and SimplyHired jumped into the mix back then, and where SimplyHired was visually busy with obvious effort going into the look and feel, Indeed took the Google-ish approach and kept it simple. Both were scraping job sites and jobs listings on company websites, making them searchable from one (or two) portals. That was a nice addition to the job seeker’s toolkit, surfacing job listings where one might not have looked, and certainly making the chance of spotting an interesting match possibility greater if that one cool position was not listed in The Usual Places.

Long-since lost to the ether, I had written a blog post somewhere comparing the two sites, in greater detail contemporaneously than I have done here, and I came down on the side of Indeed for simplicity and utility. SimplyHired was not bad by any means, but it was not a tie in my mind. This lead to an email from an investor in SimplyHired, who had seen my blog post and wanted to learn more about my views. Dave McClure was sliding into my inbox.

Given the various good and substantially bad things that have come about under one of any of 500 hats, I'll leave the specifics of that discussion for the ages, but I have always had a special place for Indeed, even subscribing to their RSS feeds for my searches to keep a finger on the pulse of the job market wherever I was. The take-home from the cover image should be that on the chart of job seeker utility sites, the trend from SimplyHired to Indeed is still downward. I do not know as I type this if my image will include the upper-right side of the chart, where we have a tool presently called X, for the name and in fact its very existence is unknown at this time.

I should point out that there are only more and more places to post jobs, CVs, videos (Tik Tok rolled that out, I think, so everybody else did the week after? Something like that?), and all manner of engagement attempts from any side of the work market. I’m touching on these two here because of that closer look so long ago, though I have to say that SimplyHired is only more interesting at the moment because their templates for CV output (after the same old item-based form entry loop) are somewhat attractive. Nothing you can’t do with some css or whatever you like to use to create such a thing. These two tools are prompting my rant, they are not the subject of it.

What is X?

I eventually found a job that brought me back to Silicon Valley for round two of my sentence, and with so many posts to so many places during my search I continued to get headhunter emails for weeks and months after. By the way, the job in question was found and secured by way of a colleague with whom I had worked at my very first job out of college. Something to think about.

One day I received an email from someone who was very clearly driving their numbers up rather than actually considering roles and candidates, with a very long job description in the message that eventually lead me to determine that the keywords Software Architect and Electrical Engineering (from different parts of my CV) where used to match me with a role working with/for PG&E as an architect. As in, someone who draws designs for buildings and other structures.

There had to be a better way. Still true.

My First Pass At It

I actually got wrapped up with someone who was working on their dream recruiter tool, HireMe.tv. Video CVs for the win? I am on the fence about this whole notion, as there is a time and a place for video or audio or other non-textual evaluation of candidates, but it is a minefield in my opinion. Let us stick with the 80% that would not require such complication.

Taking something from my lengthy and disappointing job search while in Phoenix and that hapless headhunter email for an architect role at PG&E (and countless other nuggets of observation through my years and travels), I came up with my own pet job site concept that I actually purchased a domain for, called, Resumedar.

If I may paraphrase my former colleague, the late Phil Karlton, naming this is hard. You may like that name, you might even go snap it up, it is probably not the best choice, but at the time it was good enough to focus my attention for a few minutes.

More important for some than the name, here are the features I was aiming for:

  • Seeker-centric. People looking for a job need the tools. People looking to hire have resources, the tools are built for job seekers. They are useful to recruiters, headhunters, etc but this is about the people looking for work.
  • Keep your content. While hosting a document are elements in a database could be an extended feature, the core feature works with a CV or portfolio or whatever that is published by the job seeker. Visitor analytics, layout, structure, visual elements, and in general your life story are yours, this is what the W3C spent all that time on HTML and CSS for, so use them.
  • Keyword steering. This was before Schema.org and various other structured data trends that have since become useful and maybe even popular, the goal here was to help to avoid that PG&E scenario I mentioned above and to enable the job seeker to adjust how they are matched with particular job reqs. See below for the modern take on this and the rest.
  • Searcher pays. A purpose-built platform for job seekers should not be and would not be funded by people who are looking for work. If your job is to find and recruit contractors and employees, you foot the bill. That has not changed, nor will it.

Those were the base notions that went into this ideal platform, from my perspective. Wasting time on keyword matches that are controlled by an unseen hand and that recruiter for PG&E, for example, is bad for everybody on all sides.

Modern Take

That first pass at Resumedar was based on less experience and insight than I have now, and most definitely on a younger internet. Quite a bit has become of our online life, though I personally think that too much of it is handed to others. The self-publish element of the X project (since I can already tell without you saying anything that you do not like Resumedar) is actually pervasive across several of the things I work on lately. In a near-term future Web 3.0 world where everything is decentralized into different blockchains or whatever shakes out of that, self-ownership is going to be a thing anyway, so this is almost a given.

Massaging the feature set a bit, how about the X platform has some of these:

  • Platform API. This may not apply to all job seekers, but surely there are those who would come up with interesting ways to help the non-coders out there (or add a no-code interface?). Standardized data structures would figure into this automatically.
  • Standard Data. Strangely, Schema.org has a schema for a JobPosting but nothing for a CV (correct me, I will happily admit my error). There have been XML schema defined (I think it was HR Schema?), there was hResume in the microformat days ahead of Schema.org, there is a yaml-resume project, and there are likely others, but more often than not they end up becoming a DocBook-style structure to feed into a template for a pretty PDF to send to a recruiter to parse into their database for mismatching you against architect jobs. The time has come to fix this.
  • Portable Data. In addition to being standard (which might imply portable, but hear me out), making the JobListing and CV (and related, like Portfolio and other elements that often do have Schema.org definitions) follow their open and portable forms, in such a way as, when it comes time to actually submit my CV for a job listing, or when I want to aggregate job listings for my students, or whatever, I am not re-inventing a parsing engine to figure out where you placed which elements, which date format you used, etc.
  • Machine Tags. That one might require the WABAC machine, but if you were a Flickr user back in the day you remember those. Also known as Triple Tags) (scroll down a bit as there is no anchor to the TT), these are interesting in that they can namespace the hashtag keywords AND associate some useful value with the keyword. the geo:lat=1.23 example is common, but how about software:architect as a simple throw-back to my recurring nightmare above? Use a rich context:keyword vocabulary to tune the matches, and where useful a value can indicate experience or some other quantified input to the match.
  • Years Must Go. There are a few places where Years are a useful to stipulate. Are you purchasing alcohol, retiring from Intel, or running for President? Otherwise, my 15 years of Python might be the same as 1 year if I’m up against Guido, so qualifying and filtering on explicit years of experience is an anachronism that needs to go. In some cases a skills assessment score might work, or a self-assessment from 0-10, or some other specific, useful qualifier. This applies on the searcher side as well, where the match keys and appropriate augmented keys with a value range can be explicitly constructed and the results scored directly and transparently (both sides should see that, at least anonymized).
  • This is about matching. Trying to own the relationship in some way (messages flow through the Indeed email gateway, for example) gets in the way. If I can program my terms and get feedback on matches, I can see where I need to steer my configuration to find (or be found by) closer-fitting potential roles. Do I need to be concerned with the legalese during this search? Do I need to read through paragraphs or re-answer the same questions about self-identifying my gender, right to work in a particular region, and so on? These matches can be completely automated and both sides of the job hunt can find each other without the mystery of The Algo under-serving everybody.

The source material on both sides is self-hosted, ideally. If the owner of a piece of content (req, CV, etc) wishes, their content can be fanned out like RSS feeds, but a link back to the origin is required. A copy of my old CV in a database somewhere that gets me a headhunter email for obsolete or incorrect info is a waste of time for all, so let us keep origin links intact.

Once there is a match that seems interesting, check out the origin! Make sure you have the latest and greatest version from the person who owns that CV, or that the job req is still open, etc. Flip through the portfolio content, read the actual CV, scrub through the videos, read the req, etc etc etc. Does the excessive white space on your CV make it more than the number of pages I am comfortable reading at first glance? Match on keywords, not on what your PDF looks like.

I could go on, but by now you are either picking up what I’m putting down, or you are not. The question I have, now in 2021 after 20 years of Indeed (and others, but I pick on Indeed here) not changing the game in any outside-the-box way that I can see, is it time to get radical and put this technology we have falling out of our pockets to use for more than short video clips and cancellations?

Finale Word on Indeed.com

Since I had already used Indeed.com passively for years with my search feeds appearing in Feedly (before I canceled that, but save for another post) and sometimes showing up in my inbox via email, and since I had actually gone through their CV creation process at some point so I could have a somewhat-structured CV to output and send to someone about three years ago, this is where I began to look for the state of affairs in the gig and remote contract (and perhaps other) markets. Here is what I see from two weeks of daily use:

  • Indeed does not use their own platform. They post many jobs, but they do not make use of the notification and application status features of their own tools, and when they rejected me (I sent applications to about 6 or 8 Indeed jobs as a test, I have no idea if I was a match, to be honest) they sent me a form letter email from a completely separate email domain, outside of their platform gateway. In other words, they do not eat their own dogfood. This is a problem.
  • I encountered many bugs in those two weeks, some intermittent, many not. These are usability in general, odd navigation, weird search results, and so on. All sorts of little things that add up to questions, though that first bullet is one of the answers, perhaps.
  • Third parties (the people who are looking for me, in a disappointingly small minority of cases) may or may not be buying in to the Indeed.com workflow. Several reqs include instructions to NOT apply via Indeed, but to apply via yet another portal or specific candidate gathering database. I do not have direct knowledge, but I assume that where I have had issues on the job seeker side, there must be some really annoying issues on the candidate searcher side. At the baseline, if I need to re-submit my CV to another candidate database, why am I starting over, when a structured document could be submitted directly (or even better, a link to my self-hosted structured document).

On the surface, Indeed.com seems pretty cool, as it did way back when. After taking it for a ride for many days now, it has gaps, it has holes, and it leaves opportunities for people like us to fix them. No, not to fix their platform, but to get it right in such a way as people can spend their time matching CV and Req, not figuring out why tools like Indeed are not working.

Artwork Attribution

I grabbed logos from SimplyHired.com and Indeed.com, and constructed my cover image for this post with Canva.com, making use of some of their art elements. I am not a graphic illustrator by any stretch, as you know.