Politics last week, AI today. Yes, I can hear your collective groans already. All you need is Consult, Robert Walters, and something wildly inappropriate and you’d be shouting “BINGO!”. Unfortunately, it’s an election year and it’s 2026, so both topics are painfully unavoidable. And whether you like it or not, AI is proving to be the most transformative “thing” (not just technology) since the interweb and/or 1988s “Above the Law“. Denying this (or criticising Seagal for being fat/a sex offender) will put you on the wrong side of history – and if you’re famous enough, will turn you into an AI-generated meme in 10 years’ time. From social media, actual media, journalism, and the world of work, AI is attacking us from all sides. Even reading about Spark’s financial performance yesterday, you only have to skim a few lines before it’s mentioned. In this instance, it’s about taking jobs. That’s kinda standard.
AI raises some big questions for the world of work. Firstly, what happens when businesses require fewer people? Do we all just work a bit less (this is good) or do fewer of us work (this is bad)? No one has yet explained to me how society can flourish with fewer taxpayers doing tedious tasks. Where will the administrators, pen-pushers, graphic designers, and junior coders go? For recruiters, we’re in the thick of it. Fewer workers means fewer vacancies of course, but will this be compounded by fewer vacancies requiring less human recruitment work? A double whammy of decreasing relevance. Yikes.
Let’s see where things are sitting right now shall we?
Firstly, the reality. Someone text me yesterday asking how much I reckon the Auckland agency headcount has dropped in the last 2-3 years, stating someone had suggested 35-40%. I think that high, but not drastically so. The reality is, if we count non-billers, it’s at least a third down. We have of course seen this before, and each time, like a bout with Rocky Balboa, it’s bounced back and then some. Bruised, slurring, but always bound to make a comeback in the next movie. It would be easier, and I would sleep better, if we could continue this trend. However, I feel this time, we’re experiencing something slightly different. Recessions, elections, and wars for oil, land, and religion have happened before. Believe it or not, but I’m old enough to have recruited through all of the above. Each time, managers have “waited for things to pick up before hiring“, and when things do “pick up“, us recruiters get phone calls. Now, for the first time in history, we’re not the first solution on our clients’ minds. Most GMs and business owners now think “I know what will replace 25 years niche experience for the cost of a coffee!! ChatGPT!“. Now of course they’re wrong, but perception always trumps reality, at least at first. To keep pace with this, and mirroring society as a whole, we’re seeing recruitment firms downsize, and increasingly look to AI to do the heavy lifting. Deliver more with less is the mantra.
This is the mildly interesting bit. Some aren’t just using AI to prospect, reference check, and source. Instead, they’re spotting a wave and deciding to ride it. Not content with using AI to make money the old fashioned way, they’re turning AI into products. Twice in a week I’ve spoken to recruiters who have built sourcing and screening tools with the help of AI. They always ask me my opinion on the commercial viability of such a thing, and I say the same thing. It’s very slim and minimal. Had they created this product 3 years ago, they would have had several million dollars thrust into their sky rocket right now and they’d be drinking spicy margs from the bum cleft of Di Caprio’s cast-offs. However I don’t think regular folk monetising AI has any legs. Not long term. And I think this for a couple of reasons. Firstly, without AI, they couldn’t have created an AI product. And this means that I can also create it. And so could you. And you (but not you). And if idiots like me can get AI to program itself, imagine what those who are truly skilled can do. This democratisation of technology sure makes it hard to make money out of technology. Secondly, we don’t actually own what we’re building. Essentially, we are making reasonably useful AI agents which use a blend of OpenAI, Google, Anthropic, and xAI LLMs. It’s these f*ckers that are eventually going to be calling the shots (as if they aren’t already). SEEK have a huge AI team and I’m sure they’re putting AI to all sorts of clever stuff. However, SEEKs value won’t be based on the AI technology that they have. Having AI technology is the increasing norm, and can be built increasingly well by an increasing number of people. Their value is the years they’ve spent getting us lot addicted to suckling their teat. Their products come and go – but what keeps us coming back is those f*cking addictive ads which somehow, against all logic, still f*cking work. It’s so annoying.
So where do humans fit into this? Well, there’s still value in the human side of recruitment. Negotiating offers, assessing if someone is a bit weird, taking a client for a beer. We need to double down on this whilst embracing AI for the other stuff. What isn’t going to work is for us to try and monetise what is essentially a well-prompted AI agent, dressed up as a product. Soon (and I mean in a couple of months) every ATS will have agents that write ads, PDs, source, screen, and rank candidates. The leg work in these products is building the infrastructure that AI hangs off of, the UX, the human thinking behind it, and how strong and long the contracts you can sell to clients are. Same with job boards. AI is the easy bit. The track record of ROI is the gold dust. And us agencies have to accept the fact that “growth” is no longer about headcount, or even revenue. Profit is king. And you’re not going to be a profitable recruitment firm in 2 years if you’re not using AI in some form. I’m not sure I like it either.
^SW
