The Molecule Was the Easy Part: The GLP-1 Talent Bottleneck

The industry spent decades trying to get a peptide into a pill. The people who finally managed it, and the ones now being asked to make it at scale, are the real story behind the GLP-1 boom.

‍

For the better part of two decades, getting a peptide into a pill was the kind of problem ambitious scientists were quietly steered away from. A small number of formulation chemists kept at it regardless, working on something much of the field had filed under theoretically possible and practically hopeless. The payoff, for most of their careers, looked unlikely to ever arrive.

What they were up against was the gut. The human digestive system is, in effect, built to dismantle exactly the kind of large, fragile molecule a peptide is. Stomach acid and enzymes go to work the moment it arrives, and very little survives the journey across the gut wall intact. For decades, that left one reliable way to get a peptide into the bloodstream, which was to bypass the gut altogether with a needle. The science was never short of ideas for getting around it. It was short of ideas that held up at a usable dose.

The breakthrough, when it came, was as much persistence as inspiration. Novo Nordisk's oral semaglutide tablet uses an absorption enhancer called salcaprozate sodium, which briefly lifts the local pH around the tablet and helps carry the peptide across the gut wall before it degrades. It reached the US market as a weight-loss pill in January 2026, the first of its kind. Eli Lilly and Company arrived from a different direction with orforglipron, a small molecule that hits the same receptor without being a peptide at all, approved a few months later in April. Two routes to the same destination, both of them the work of people who'd backed a problem most of their peers had given up on.

Solving the molecule simply moved the difficulty somewhere else. The GLP-1 medicines behind all of this have created demand on a scale the industry has rarely dealt with. The global market has been valued at around $53 billion, with forecasts pushing it past $157 billion by 2030. Proving a peptide could go into a pill was the decades-long puzzle. Making enough of it, quickly and to standard, turned out to be the harder one.

Part of that bottleneck is physical. A GLP-1 peptide isn't made like an ordinary tablet. It's built through solid-phase peptide synthesis, a slow, multi-step chemical process, then purified at scale before it ever reaches a fill-finish line. These are technically complex facilities that take years to build and qualify, which is why production can't react quickly when demand jumps. Novo's answer was to spend $16.5 billion buying the CDMO Catalent in 2024, largely to pull fill-finish capacity in-house and control one of the tightest points in the chain.

The harder constraint is people. Around 60% of the world's sterile fill-finish capacity for GLP-1s currently sits in three locations, in North Carolina and Indiana in the US, and Kalundborg in Denmark. A company can announce a new facility in a press release. It can't announce the senior process development lead who knows how to scale a peptide synthesis from kilograms to tonnes without bleeding yield, or the peptide chemist who has run that purification at commercial volume rather than studied it. Those people exist in small numbers, they're largely already employed, and every company expanding at once is recruiting from the same shallow pool. When the industry refers to a capability backlog, this is mostly what it means.

That part rarely makes the trade press, and it's the part that decides who wins. The scarce profiles aren't junior. They're the people who've already done the hard version once, who've scaled a process under real pressure and taken a peptide from clinical supply to commercial volume. You don't build that in a year, and you can't hire your way out of the shortage any faster than your competitors are trying to do exactly the same thing. The companies moving fastest right now tend to be the ones who worked out early that capacity on paper means little without the people to run it, and who started building those teams before the rest of the market caught on.

‍

Oral delivery doesn't relieve any of this. It adds to it. A tablet that has to protect a peptide and release it intact at the right point in the gut is its own formulation problem, needing its own specialists. As delivery turns into a discipline that spans the drug and the device, with the data around adherence not far behind, the people who can move across those worlds only get rarer.

The direction of travel is set. More peptides will follow the same path out of the syringe and into a pill, and the science that looked impossible ten years ago is on its way to becoming ordinary. The part that stays difficult is the one that was always going to be difficult once the chemistry was cracked, which is finding the people who can make these things reliably and at scale. Having built CDMOs from the inside, that shift is the thing we watch most closely. The next decade in this market will belong to the firms that grasp how far the real bottleneck has moved, from the lab to the org chart, and staff for it accordingly.