The AI-Native Search Firm Con: Why Your Algorithmic Shortlist Is Just LinkedIn With a Ranking Model

Let us say the quiet part loudly

Almost every "AI-native" executive search platform on the market today is doing the same three things: scraping LinkedIn, running a language model over the profiles, and re-ranking the results against a job description. That is not a market map. That is a search engine with a coat of paint. And the candidates it surfaces are, by definition, the candidates who have already told the internet who they are.

The dirty secret of executive recruitment is that the best candidate for a Brisbane CFO or COO role almost never has an up-to-date LinkedIn profile. They have not posted in four years. Their headline still says the job they left in 2022. Half of them have their profile set to private. The most placeable executive in the SEQ market right now is invisible to every tool that starts with "we index the professional web".

The 5 percent problem nobody wants to talk about

Roughly 5 percent of any given executive market is publicly visible, actively engaged on LinkedIn, and open about what they want next. The other 95 percent are the ones who actually move the needle when they change roles. They are the operators quietly building. The GMs who have not needed a new job in a decade. The founders who just sold and are working out what is next. The Chief People Officers who took nine months off after the last acquisition and are only telling three people they are looking.

An algorithm cannot find those people. Not because the algorithm is bad, but because there is nothing to index. Their information does not live on the internet. It lives in someone else's head, from a conversation seven months ago, over a coffee that had nothing to do with a specific brief.

So when an AI-native firm sends you a "market map of 240 qualified candidates" 72 hours after the brief, what they have actually sent you is a beautifully formatted list of the visible 5 percent. The same 5 percent every other AI-native firm in the market just sent to your competitor. Ranked by a slightly different model, but drawn from an identical pond.

What the demo does not show you

Sit through the sales demo of any of these platforms and you will see the same theatre. A slick interface. A confidence score next to each name. A little chart claiming to compare candidates on "leadership signal" or "growth stage fit" or some other invented metric that sounds rigorous and is, on inspection, a language model guessing from a bio.

What the demo does not show you is the referee check. The private conversation about why the last CFO actually left. The board dynamics that make one candidate a nine and the same candidate on paper a two in a different company. The context that this particular CEO cannot work with anyone who thinks in quarters. The fact that the highest scoring candidate on the platform has been quietly asked to leave two of their last three roles for reasons that were never made public.

None of that is in the model. None of it can be in the model. Because none of it is written down anywhere the model can reach.

The uncomfortable question for boards

Here is what should keep any Chair awake. If your search partner's core differentiator is a platform, and that platform is trained on the same public data every other platform is trained on, what exactly are you paying a premium for? The candidates are the same. The rankings are broadly the same. The pitch deck is definitely the same. The only variable left is the human sitting behind the tool, and the whole point of the tool was to reduce reliance on that human.

A search firm that sells you the algorithm has, by their own logic, told you the algorithm is doing the work. Fine. Then why is the fee still built like it is 2015?

Where AI is genuinely useful, and where it is theatre

None of this is anti-AI. We use AI every day. It writes cleaner search summaries than a human intern. It parses transcripts faster than any researcher. It is genuinely helpful for compensation benchmarking, competitor org mapping, and turning a two hour reference call into a defensible written record.

What AI cannot do, and what the "AI-native" pitch quietly relies on you not noticing, is originate a relationship. It cannot pick up the phone to a GM who has not returned a recruiter's call since 2019 and get a conversation because they trust the person calling. It cannot read the room in a private dinner. It cannot tell a client, with the weight of Jason's relationships built through 17 years recruiting behind it, that the shortlist they think they want is not the shortlist they actually need.

Those are still the parts of executive search that decide whether the appointment works. And they are still, stubbornly, human.

The honest test

Next time a search firm pitches you their AI capability, ask two questions.

• "Of your last five successful placements, how many came from your platform, and how many came from a phone call to someone you already knew?" The honest answer is almost always the second. If they cannot answer, that is the answer.
• "Show me a candidate on your platform who does not have a LinkedIn profile." Watch what happens. This is the entire game in one question.

The firms with a serious answer to both are worth talking to. The rest are selling you a very expensive Boolean search dressed up as the future of the industry.