What artificial intelligence really means for executive search in pharmaceuticals and biotechnology

There is a rather delicious irony in the executive-search industry worrying about artificial intelligence.

For years, search firms have explained their fees by invoking proprietary networks, deep market knowledge, painstaking research and the mysterious art of identifying people whom lesser mortals could never possibly find.

Then along came LinkedIn and, before it became brochureware, made half the industry’s theatre faintly embarrassing.

Will AI finish the job?

This is particularly relevant in pharmaceuticals and biotechnology, where executive search has long enjoyed unusually favourable conditions. Our industries are technically complex, the talent pools are international, the stakes are high and boards are cautious. A failed CEO, Chief Medical Officer or commercial leader can destroy years of work and hundreds of millions in capital.

The result has been a market willing to pay for reassurance.

Unfortunately, reassurance and value are not always the same thing.

What’s already becoming cheap

Much of what executive search firms actually do is already becoming tractable at pace.

Company mapping. Candidate identification. CV comparison. Organisational research. Competitor analysis. Therapeutic-area mapping. Interview preparation. Written assessments. Market intelligence.

Once, these things took teams of researchers and consultants several weeks.

Now a reasonably competent person with good AI tools can do a startling amount of what looks like credible work before lunch.

This does not mean executive search is dead.

It means that quite a lot of what has been sold as executive search was never terribly valuable in the first place.

That distinction may prove brutal.

The cult of the CV

Life sciences makes the problem especially interesting because it is an industry that worships expertise.

We like people who have “seen this movie before”.

Preferably at Roche.

Or Genentech.

Or AstraZeneca.

Or perhaps during the brief period when everybody wanted their chief executive to have been at Moderna.

Or wherever someone’s binary bet came good last time.

The CV is expected to tell a reassuring story: right company, right therapeutic area, right financing environment, right board, right exit.

There is logic in this.

There is also a danger.

When “knowledge” stops being scarce

AI is beginning to make “knowledge” easier to access.

Not wisdom. Not judgement. Not courage.

“Knowledge”.

A pharmaceutical executive no longer has the same advantage merely because he has spent 25 years accumulating information that somebody else can now interrogate in seconds.

A biotech CEO does not need to carry the competitive landscape in his head if a machine can update it continuously.

A CMO does not become useless because AI can synthesise clinical evidence, but the value of being a walking filing cabinet is plainly reduced.

The same is true of your search partner.

“Knowing” who runs oncology at Novartis is not a business model…actually knowing them still is.

Nor, increasingly, is producing a list of people who once did something similar somewhere else.

And this is where things become more interesting.

Familiarity and judgement are the scarce resources

If AI makes information abundant, then judgement becomes more valuable and familiarity brings context.

That sounds suspiciously like management-consulting prose, so let us put it more plainly.

The difficult part of hiring a chief executive is not finding ten people with the right CV.

The difficult part is knowing which one will be the most likely to succeed in YOUR business.

That has always been true, but the industry has been able to disguise the distinction because finding the ten people used to be expensive.

Soon it will not be.

The executive-search firm of the future therefore faces an awkward question:

What exactly are you being paid for?

Not your database. Everybody has access to databases.

Not as much of “your” research as before. AI can help make research more tractable.

Not your “global reach”. So can the internet.

Not your beautifully formatted candidate report. AI can write one in about 40 seconds, including the prose.

The answer has to be judgement and access.

And judgement is troublesome because it cannot be standardised nearly as easily as the industry would like. It means:

• Telling a chairman that the obvious candidate is wrong.

• Backing somebody who has never held the title before.

• Recognising that a glittering big-pharma career may have been achieved with 400 people and a billion-dollar budget behind it.

• Asking whether the charismatic biotech founder is actually capable of running a public company.

• Distinguishing achievement from proximity to achievement or market effects.

• And sometimes, telling the client not to hire anyone at all.

The question search firms would rather not ask

That last point may be the most disruptive.

The retained-search model contains a rather obvious economic incentive. A search firm only makes money when they complete the search, the first two phases almost cover their cost of delivery. It is not normally paid to conclude that the job should not exist.

Yet AI may force precisely that conversation.

• Does a biotech company really need a full-time Chief Strategy Officer?

• Does a 70-person therapeutics business need the same functional architecture as a 7,000-person pharmaceutical company?

• Does the organisation require another permanent executive, or a fractional operator for 18 months?

• Could one unusually good leader, properly supported by AI, do work previously requiring a small department?

These are not theoretical questions. They are beginning to become organisational-design questions.

And executive search firms that refuse to ask them will gradually become more like contingent agency recruitment suppliers to companies that have already made the important decisions elsewhere.

That would be a considerable demotion.

What this means for candidates

The candidate implications are equally uncomfortable.

Biotech hiring has become addicted to literal experience. If the company works in immunology, find an immunology executive. If it wants to launch in rare disease, find somebody who has launched rare disease. If it needs a CEO, find somebody who is already a CEO.

Preferably three times.

This is understandable when technical knowledge is difficult to acquire. But what happens when technical context becomes much easier to acquire?

Then the genuinely scarce things start to look different:

• Can this person make decisions with incomplete evidence?

• Can they kill a programme they personally championed?

• Can they tell investors something they do not want to hear?

• Can they spot when consensus is wrong?

• Can they recruit people better than themselves?

• Can they operate without the machinery of big pharma behind them?

• Can they use AI intelligently without either worshipping it or pretending it does not exist?

• Can they persuade a regulator, a board, a scientific founder and a frightened management team in the same week?

These are rather harder things to scrape from LinkedIn. They are also rather harder for AI to determine.

Bad search dies. Good search gets scarcer.

Which is why I suspect the effect of AI on executive search will be paradoxical. Bad search becomes almost worthless. Good search becomes more valuable. The middle gets crushed.

The firms whose principal skill is producing plausible lists of plausible people will have a problem. The firms that genuinely understand companies, markets, personalities, incentives, boards and human behaviour may find themselves in greater demand.

Because once everybody can find everybody, finding people is no longer the point. Choosing them is.

The end of information asymmetry

There is another implication that the industry may enjoy even less.

Clients will become much better informed. A board that once depended on its search adviser for market knowledge can increasingly challenge the adviser’s assumptions:

• Why this candidate? Why not that one?

• What evidence supports your assessment?

• Which comparable appointments failed?

• What did the unsuccessful candidates do next?

• How much of this recommendation is analysis, and how much is familiarity?

Quite right too.

For too long, parts of executive search have depended upon information asymmetry dressed up as expertise. AI will be extremely effective at stripping away information asymmetry.

What remains will be expertise. Or nothing.

For pharmaceuticals and biotechnology, that could be healthy. Leadership appointments matter too much to be reduced to CV matching with better stationery.

What business are you actually in?

The great irony is that AI may force executive search to become what it has always claimed to be. Not a sourcing business. Not a database business. Not recruitment with nicer lunches.

An advisory business built around judgement.

For some firms, that will be an exciting opportunity. For others, it may be the first time they discover what business they were actually in.