Platt Perspective on Business and Technology

Dissent, disagreement, compromise and consensus 6 – the jobs and careers context 5

This is my 6th installment to a series on negotiating in a professional context, starting with the more individually focused side of that as found in jobs and careers, and going from there to consider the workplace and its business-supportive negotiations (see Guide to Effective Job Search and Career Development – 3, postings 484 and following for Parts 1-5.)

I wrote Part 2 and Part 3 of this series with a focus on the earliest stages of a job search and particularly for those who have not carried out that type of exercise for a long time, who see need for more fundamental change in what type of work they would do and even in their overall career path, or both. And I offered Part 4 and Part 5 with a goal of bringing the next step of an overall job search campaign process into this narrative. And in that, I focused on background and preparation issues that enter into consideration when getting ready to reach out to, and apply for a specific position with a top choice for you, hiring company.

To be more specific there, I wrote in Part 5 “that you have a fairly solid idea as to what you would see as an ideal next job for you, and with a business that would be a best fit for you. And I assume that you have updated and refined the job search marketing tools that you would need for this: with your resume and other written documents, and with your interviewing and other skills as they could most effectively be used for these jobs in presenting your case. And that brings me to this posting, which I begin with a challenge of sorts.” And I then went on to question this set of assumptions, and with a goal of helping you to refine how you would fulfill them from that.

My goal here, at least to start is to re-challenge that starting set of assumptions but from a second direction, in order to address changes in how businesses select, consider and hire job candidates. More specifically, and in contrast to a number of series that I have been offering here that touch upon the issues of automated algorithm-driven information processing and filtering, and artificial intelligence-driven agents: I have been writing in this series in terms of human readers and human decision makers only. I will move beyond that presumption here.

I began posting to my Guide to Effective Job Search and Career Development in this blog in October 2009. And one of the first issues that I addressed in that, and certainly with regard to reaching out to a possible employer, and to submitting a resume and application to them in an attempt to secure a new job, was how businesses have actively sought to automate their first round candidate screening. Why? Their underlying rationale for pursuing that type of course has not changed and for years now, as more and more, and by now essentially all businesses that allow candidates to send in resumes online, have been inundated with floods of generic submissions that can best be considered spam resumes and spam applications. I briefly made note in Part 5 of a job candidate’s need to prove their genuineness, and that they are not simply a “to whom this may concern” generic submission as would be sent to any possible receiving email address or web site form, as set up by any business that happens to have a job opening.

How does this challenge fit into this series, with its focus on more effective communications and negotiating skills and their use? My answer is very simple. Anyone who would at least initially approach a prospective employer, looking to hire new people into their overall team, has to be able to craft a message that can simultaneously meet the screening needs of a mindless automated filtering system, and still work in creating interest in a real person, when and if their resume and cover letter make it through that first automated gatekeeper round.

I began writing about this set of challenges in terms of what might be called a version 1.0 context, where all of the pre-human viewing filtering is carried out by what amounts to simple key word searches. All resumes and cover letters received, under this system are date and time stamped and coded to identify job candidate source, so anything coming from that applicant can be included in a same single record. And all of this documentation and from all such sources, once tagged this way, is uploaded into a large data base. Then this pool of accumulated document files is data mined, and filtered for relevance to the particular job description that a business is hiring for, using SQL queries that would seek out submitted documents that include at least a threshold level of specific sought after key words in them, and with the sought-for list of those key words prioritized and weighted for their value according to how important their relevant inclusion would be for a hiring manager who set up this job candidate search.

This is a technical solution for what is framed as a strictly technical problem, so it should not be too surprising that it works best, to the extent that it does work, in more technical fields and for more routine positions, where for example a hiring manager might need a new hire who knows some specific computer programming language, or some particular content editing tool, or who has worked in some specific industry, or perhaps (ideally) with some specific business there. It breaks down a great deal when new hires would be needed for positions that are less easily captured, and even just in rough cartoon form by some specific set of background indicating key words or phrases. And that brings me to a version 2.0 to this still rapidly evolving saga, that is only just beginning to be developed and deployed: the use of single task oriented artificial intelligence-driven agents for carrying out this initial screening, designed to go beyond simply key word filtering searches in selecting a fraction of all submissions that might merit further, more directly human consideration.

This, I add is where neural network-based hardware and self-learning AI software systems, grounded in a database expert system foundation of how human experts have selected out viable and best candidates, will really take off. And that is where the challenge that I write of here will become both more interesting, and a lot more complex, and certainly for job applicants.

Automated key word filtering as a first round mechanism for culling out the generic spam, has primarily meant that a real candidate, specifically applying to some particular position with a specific hiring company, has to be careful to use the same type of wording that they use in their posted job description and when discussing this area of their business on their web site, and certainly for the key skills and experience words used there. At the risk of misusing a term or two here, the emergence of candidate culling and filtering version 2.0 will mean a real candidate having to navigate their way through the filtering and selection processes and priorities of two types of at least loosely algorithmically prescribed thought processes: one relatively rigidly algorithmically defined and the other much less so, but with its own biases, preferences and preconceptions (as set by overall company hiring policy and practice, and by the personal biases and assumptions of the hiring manager.)

How would I propose navigating this more complex and nuanced path going forward? I can only offer a partial and even tentative response to that question here, with the still embryonic stage of version 2.0 candidate pre-filtering and culling in mind, that we are coming to face. Focus on improving how you would respond to a more advanced version 1.0 system but with additional care made to avoid colloquialisms, or vagueness of expression. This means being as clear as possible and it means avoiding ambiguity of possible interpretation in what you do offer in your submitted documents, that you will be judged and selected-in, or filtered out by. Beyond that, I at least categorically make note of a possible and even likely source of trade-offs that will have to be addressed. From the version 2.0 automated screening perspective that your submission will have to successfully pass through first, effectively mirroring the phrasing of the job description and related business-sourced content that you are responding to, will probably offer positive value to you. But too much mirroring of this type, might very well trap you when your resume, cover letter and any related documents are passed on to a next step human review – and certainly if an excess there does not red flag your submission as the product of a robo-applicant, or similarly fraudulent applicant before any human can get to see it. And even if you’re perhaps too-close-a-match application documents do not get caught up in that type of problem, you will still need to show that you can and do think through the issues that you are presenting yourself, and that you are not just parroting back what you see the hiring business as wanting. So you have to mirror back what the company wants, in their type of wording, but selectively and with a view towards your best understanding of their hiring priorities and needs. This, among other things means that a switch from a version 1.0 here and now to a version 2.0 as it more fully arises, will increase the pressure on you to more thoroughly research and understand the businesses that you apply for work with. I offer this as a briefly stated preliminary and even anticipatory note for what I am certain is to come and soon.

With that offered, think through what version 2.0 will be like for this, and certainly as far as the current state of artificial intelligence per se is concerned. And consider the implications of businesses pursuing this type of shift in the basic job candidate selection process followed by them, and the limitations that are sure to arise in that, and throughout any more immediately foreseeable future. In that, consider the sometimes extreme lack of apparent judgment and understanding that still comes through from even the best, most advanced automated agents coming from sources such as Google, Microsoft and other leaders in the development of consumer-facing AI agents, with their Siri, Alexa and so on. Now consider the implications and complications of having one of those agents carrying out a candidate screening and filtering, and first round job candidate culling function for an employer that happens to get way too many applications to be able to deal with them without resorting to automation. Anyone who has ever struggled to convey what would seem to be a simple request or command to those AI agents, knows how limiting and frustrating that can become. And the version 2.0 candidate selection AI agents that I write of here are certain to go through a protracted, similarly “awkward” development period too.

I am going to continue this discussion in a next series installment where I will assume that your submission has made the initial automated cut and that you are now going to face human gatekeepers, perhaps starting with a member of Human Resources, but with that process leading towards your meeting with a hiring manager. And I will begin to more fully and directly consider negotiating and the skills that go into that, there.

Meanwhile, you can find this and related material at Page 3 to my Guide to Effective Job Search and Career Development, and also see its Page 1 and Page 2. And you can also find this series at Social Networking and Business 2 and also see its Page 1 for related material. And I particularly recommend your at least briefly reviewing a specific job search best practices series that I developed here on the basis of both my own job search experience and from working with others going through that: Finding Your Best Practices Plan B When Your Job Search isn’t Working, as can be found at Page 1 of my above-noted Guide as its postings 56-72.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.

%d bloggers like this: