Intelligent Recruitment – Contradiction in Terms?

It took a while, as it usually does these days, but I finally got it. For years I’ve been a LinkedIn user. Not particularly active, but it touches my vanity when people want to connect with me, even though their reasons for wanting to make my acquaintance usually remain obscure, since the act of connecting is almost always the last I hear of them. Except that they lurk forever in that very eclectic group of people LinkedIn refers to as my contacts.

Lately I’ve been receiving emails from the site telling me about ten jobs I might be interested in. Which is rather odd, since at no time since I signed up with LinkedIn have I ever stated or even implied that I was looking for a job. But as the former owner of a recruitment business I’m interested to see how effective the software is at figuring out what jobs might be appropriate for me.

The answer is: not very effective. How I might qualify for the position of Senior Building Surveyor with a property company is an interesting question, for example. Perhaps the software has figured out that I live in a house, which it thinks is qualification enough.

Another opportunity it helpfully pointed me towards is to be the manager of a testing service business for a computer retailer. Years of wrestling with laptops that go into cardiac failure after a couple of years clearly make me a good pick for that one.

A third job in the current batch got a bit warmer. Would I like to be an executive assistant to the founder of a hedge fund? Sounds interesting – a chance to learn the tricks of the trade and run off a couple of years later to start my own hedge fund. The only problem would be that to disguise myself for any length of time as a smart woman in her mid-twenties would be a serious challenge to my acting ability, not to mention the problem of paring down my voluminous. Mrs Doubtfire perhaps, but definitely not Miss Moneypenny.

The other jobs were at the BBC as a “service owner”; with a digital advertising agency as a “male lifestyle media owner”; as a “connected care feature owner” with Jaguar; and as a “Programmatic Sales Manager – Media Owner” with a digital publisher.

And that’s when I finally twigged what was going on. I’d been working on the assumption that LinkedIn, as one of the world’s leading social media companies, had some pretty smart software people on their team. People who use all kinds of sophisticated algorithms to harness their big data on my behalf. Like Facebook and Amazon, for example. Not necessarily very nice people, but certainly smart.

That may be the case in other areas of their business, but not with the engine they use to match people to jobs. The clues I hadn’t picked up on until my eureka moment were the words “owner” and “founder” – words that feature on my LinkedIn profile because I have indeed been an owner and founder of businesses.

So basically these guys are using nothing cleverer than word searches to come up with these ridiculous suggestions of jobs that might suit me. Doubtless on an industrial scale, consuming enough electricity power for each search to power a small town for a year. And wasting enough time on the part of users to last several lifetimes.

The technology employed – as far as I can see – is no different than what my recruitment business used twenty years ago. Databases and word searches. Except that we went a step further and used a system that allowed relevant terms to be specifically coded, so that the search results yielded the skills that we were actually looking for rather than an endless list of random and irrelevant occurrences of search terms.

Thus if I said in my CV that I was a “Jaguar owner” (which I’m not and never have been by the way),  the software we used all those years ago would have ignored that, or presented the information way down the search. Not so LinkedIn it seems.

Perhaps I’m wrong. That’s entirely possible, since I’m a mere earthling far from the cutting edge of the technology business. But I would have thought that by now companies like LinkedIn would have figured out by now how to guess people’s career aspirations a little more accurately.

In my case, for example, it should be pretty clear that I’ve been around the block a few times, and that I’d rather climb up Everest without oxygen than contemplate working in the middle ranks of some corporation, even if that company was foolish enough to want to employ me. My profile should be screaming out: PAIN IN THE ASS – UNEMPLOYABLE. The definition of a freelance consultant, perhaps.

I suppose you could argue that LinkedIn is merely a conduit. That it makes no judgement about the suitability of a person for a job. Presumably it leaves that to all the headhunters and networkers who pay premium fees for easy access to the millions on its database.

In which case it’s missing a trick. Why, for example, could it not provide a private space for people to talk at more length about themselves? About their aspirations, their career preferences, their favourite companies, their role models. And if it was concerned about the data security implications of doing that, it could sell access to a profiling engine to headhunting companies that can’t afford such software, so that they could retain the information in their own space.

I shouldn’t really be too hard on LinkedIn. Sadly, recruitment technology doesn’t seem to have advanced much over the past 20 years, except possibly in improving process. The real silver bullet lies not in doing things more efficiently and quickly, but in finding the right people, meaning people who can add value, not just now but in the future. And for that the traditional solutions seem still to apply – database and internet searching, psychometric tests with varying degrees of clunkiness, and of course interviews – none of which can be relied upon to supply the critical ingredient: judgement.

Artificial intelligence barely seems to get a look in. Business games have been around for quite a while, and are starting to be used in assessment centres. HR companies will tell you that their assessment techniques focus on attitude rather than skills, and can predict the success (or otherwise) of a candidate in a particular role. But the trouble is that they rarely get the opportunity to use their science on the other side of the equation: the employer. Predicting whether a person will succeed in a vanilla environment is one thing, but getting a handle on a company with thousands of employees and hundreds of subcultures that change each time an influential employee leaves is quite another. And I’m sorry, but the mission statements and beautifully crafted company values won’t change that – they reflect “like to be”, rather than “actually is”.

Most recruiters take the employer’s word for what it’s like to work for them. Some talk to former employees, who may or may not be biased depending on their reason for leaving. Others use their common sense by picking up indicators of culture from visits. But all of this information is based on the present and the past, not the future. A potential employee might be perfect for today, but what about tomorrow?

Businesses and their cultures are dynamic, not static. Some companies recognise this, and hire consultancies to define the attributes of their ideal employees based on future plans and aspirations, but how many do this on a regular basis? Not many, I suggest. And what about the people whose attitudes and skills match the present and not the future? They either adapt or leave. And if they leave, it can be a colossal waste of investment. I’ve seen that happen time and time again. One warning sign is when a company asks its employees to “apply for their own jobs”. I remember Digital doing that in the late 90s, much to the consternation of its staff. Within a year or two it was gone – gobbled up by Compaq.

All of which suggests that while the gurus warn us that artificial intelligence threatens to make huge numbers of workers redundant over the coming decades, people who hire other people will not lose their jobs any time soon, no matter how dumb they are, because no amount of intelligent software can compensate for dumb recruiters, or for dumb executives who don’t think further than their noses when figuring out who they want to hire. And yes, there are some smart recruiters and employers out there, but I haven’t met one yet that would accept the recommendation of an expert system over their own experience, intellect and gut feeling – in other words, judgement.

So perhaps LinkedIn are quite smart in keeping things simple and leaving expensive mistakes to the employers and employees that use them.

Anyway, my next career move will probably be as a greeter for Asda. I don’t think I’ll need LinkedIn for that.

(Illustrations are by the illustrious Hunt Emerson for a book I wrote on recruitment – among other things – back in 1995)