Biotech companies usually begin with a scientific idea. Someone finds a compound that does something interesting, and a company forms around the question of what it can do. I have helped build companies from that starting point, and it is a reasonable one.
I’ve also come to believe the order deserves a second look. The first question a company asks shapes every question after it. “What can this molecule do?” sends a team in one direction. “What does this patient need?” sends it somewhere else, and the two paths rarely converge by accident.
Building backwards means starting at the end. You describe the medicine a patient would want to take, in the life they actually live, and then work back through each decision a company makes to get there. I’ve watched those decisions from several seats: as a pharmacist, a researcher, a founder, an investor, and now as a CEO. What follows is how I’d sequence them.
Write the Label Before the Lab Work
Drug developers already have a tool for this. A target product profile, or TPP, summarizes the medicine you’re aiming for: who it’s for, what it should achieve, how it’s dosed, how safe and tolerable it needs to be, and how stable it must stay on a shelf. The NIH’s SEED program describes it as a planning tool that begins development with a clear end goal, and advises that “a TPP should be integrated at the inception of a drug development initiative.”
The trouble is how teams fill it in. A TPP written outward from the molecule records the potency someone measured and the dosing early studies happened to support. A backwards-built company writes it inward from the patient. What does a good week look like for someone with this disease? What would they trade for convenience, and what would they refuse? Which side effect would make them quietly stop?
Minimum and preferred attributes are listed for each characteristic, and I read the minimums closely. If the floor was set by whatever the chemistry delivered, the molecule has already made a decision that belonged to the patient.
Let Daily Life Choose the Format
Formulation, route of administration, and dosing frequency can sound like engineering details. For the person taking the drug, they are the product.
HIV prevention offers one of the clearest recent cases. Daily oral pills for pre-exposure prophylaxis had worked in clinical studies for years, yet a pill every day remained a barrier for many people. In June 2025, the FDA approved lenacapavir, sold as Yeztugo, as a twice-yearly injection for HIV prevention. In the PURPOSE 1 trial among cisgender women in South Africa and Uganda, the lenacapavir group recorded zero HIV infections. The two daily oral comparison arms recorded 39 and 16.
The pharmacology was impressive, and the product decision behind it was just as deliberate: design around what daily life demanded, and a medicine gets taken. A medicine that gets taken is the only kind that works.
That logic can point either way. Some patients would choose a pill over an injection; others would trade a daily dose for a clinic visit twice a year. A backwards-built company asks early, while the answer can still change the program.
Dose belongs in the same conversation. I’ve written before about the habit of carrying the highest tolerable dose forward because it produced the most striking early result. Starting from the patient reframes the question as which dose someone can live with for years while still getting the benefit.
Recruit the Trial You’ll Wish You Had Run
Clinical development is where a patient-first company proves it meant what it said, and most of the proof comes down to who enrolls and what gets measured.
Enrollment is the easier one to get wrong. If the people in a study are healthier, younger, or more uniform than those who will eventually receive the drug, the evidence travels imperfectly from the trial to the clinic. Measurement runs a close second. Biomarkers tell scientists and regulators what they need to know; patients care about whether they can work, sleep, eat, and care for their families. A good program tracks both.
Then there is the burden of participation itself. Every extra clinic visit, blood draw, or day of travel shrinks the pool of people who can realistically take part, and that pool tends to shrink unevenly. A protocol designed for the sponsor’s convenience can win approval and still disappoint the patients it was meant to reach.
Build the Road to the Pharmacy Early
Commercialization is usually treated as a chapter that opens after approval. A backwards-built company drafts it alongside the clinical plan.
The questions are practical. How will patients learn they have the disease? Which doctors will treat them, and do those doctors recognize the condition? Who pays, and what does the medicine need to cost for that payer to agree? How will it be stored, shipped, and dispensed?
A therapy can clear every scientific and regulatory hurdle and still stall here, at great cost to the company, its investors, and the patients who never receive it. Asking these questions during early development gives a team years, instead of months, to build the diagnostic partnerships, clinical relationships, and health-economic evidence that access requires.
Questions I Ask Founders
As an investor, I’ve turned this sequence into a short set of diligence questions. Perfect answers are rarely given at an early stage, and I don’t expect them. What I’m listening for is whether a founder has thought about them at all.
Describe the patient who will take this medicine. What does their day look like?
Why this route of administration and this schedule? Did patients help decide?
Who will be in your late-stage trial, and how closely do they resemble the people who will use the drug afterward?
How will patients get diagnosed, and who will pay?
The founders who answer well are seldom the ones with the flashiest early data. They are the ones who can describe a patient’s week in detail and point to the step of care where people usually fall away.
Who Sits at the Table
Organizations reflect the people who build them. A company that wants patients to shape its decisions needs people around the table who carry that knowledge: clinicians who treat the disease every day, patient advocates, pharmacists who see where treatment breaks down, advisors who have taken a product through launch. I often encourage scientist-founders to build an advisory board early. A good one is the fastest way to learn what you don’t know, and the gaps it fills often sit on the patient side.
My own path supplied some of this by accident. Pharmacy practice showed me how a carefully engineered medicine behaves once it meets a crowded, complicated life. Research taught me rigor. Company building taught me that a good idea needs an organization behind it to travel anywhere. At EIT Pharma, which is developing therapies for serious viral diseases, I try to keep all of those perspectives in the room. They argue with each other, and the argument is useful.
The Destination Comes First
None of this asks anyone to give up scientific ambition. New mechanisms are how patients get options they never had. Building backwards is about order: the patient’s needs set the destination, and the science charts the route.
Companies that work this way tend to have fewer late surprises. Their product profiles hold up because they were tested against real lives from the start, and their launches are less likely to stall on questions nobody thought to ask. The founders I most want to back already work like this. They start every program by picturing the person who will open the package, and they build everything else to reach that moment.
Leen Kawas, Ph.D., is CEO of EIT Pharma and co-founder and managing general partner of Propel Bio Partners, a Los Angeles–based life sciences venture firm.