A clinical founder heading into the first real conversations about a sale will hear the same thing from every adviser who looks at the practice honestly. Buyers will discount heavily for how much of the clinical credibility walks out with the owner. That conversation usually focuses on client relationships and reputation. It rarely gets to the regulatory dimension. In many clinical practices, the registered clinical lead also holds the regulatory standing on which the practice’s licence to operate depends.
AI is now splitting clinical founders into two different directions. One group is using it to capture the practice’s decision logic, the protocols, triage pathways, and clinical rules that currently live in one person’s trained judgement. The other is deploying AI to mirror the founding clinician’s individual style. The first builds something a buyer can assess. The second adds another reason to reduce the price.
If you have an exit somewhere in the next five years, which direction you are moving in matters considerably.
What is the founder-dependency discount in a clinical business?
The founder-dependency discount is the reduction a buyer applies when business value is tied to one person’s continued presence. In a typical owner-managed business, that means client relationships and institutional knowledge. In a clinical business, a third element comes into play. The founder may hold the regulatory standing and the referral network trust that give the practice its licence to operate. That makes the discount harder to price away.
In practice, this plays out in concrete ways. A physiotherapy clinic where the founder is the only registered manager faces a regulatory review process if that person leaves. A GP practice where the clinical lead holds the NHS contract may lose that contract if the lead changes. A mental health practice where the consultant relationships all run through the founder has a referral base that cannot simply be reassigned. Each of these is a reason for a buyer to reduce the purchase price, or to structure the deal with an earn-out that keeps the founder tied in for years after the sale.
Clinical businesses can be sold, and regularly are. The value a buyer will pay at arm’s length is constrained by this dependency in ways that take real work to address.
Why does it matter more here than in other sectors?
In a non-clinical professional services business, the founder-dependency discount reflects one risk. If the founder leaves, clients might follow. That risk can be reduced through a good handover, a strong second team, and an overlap period. In a clinical business, some of the dependency cannot be resolved through management quality alone. Parts of it sit with the regulator, and that takes time and a formal process to move.
The Care Quality Commission requires regulated services to have a fit and proper registered manager. That individual must satisfy a fitness assessment. If the registered manager is also the clinical founder, a change of that person is not a paper exercise. The buyer faces the regulatory process of getting a new person approved, plus the operational uncertainty of the practice running during that transition. Those are risks that cannot be mitigated through goodwill or good management alone.
Referral relationships in healthcare carry a similar constraint. Clinical referrals exist because of trust in a specific clinician, built over years. That trust is not transferable in a transaction. A new owner, even a highly capable one, starts each referral relationship from the beginning. A buyer’s corporate finance adviser will see both of these risks clearly and price accordingly.
Where does AI make the dependency better, and where does it make it worse?
There is a fork in how clinical practices are using AI, and the direction chosen determines whether the technology reduces the founder-dependency discount or deepens it. AI that captures the practice’s clinical protocols, decision pathways, and triage rules in documented, auditable form builds something a buyer can assess independently of the founder. AI configured to reflect one clinician’s individual working style does the opposite.
The documentation path works like this. The practice uses AI to turn its clinical decision-making into structured, written form, covering which presentations trigger which protocols, how triage decisions are made, and what the criteria are for escalation. These are things the founding clinician has always known, but they have lived in experience and judgement rather than in any document a successor could follow. When AI is used to extract and codify them, that knowledge becomes portable.
The imitation path creates a different outcome. AI trained or configured to reflect how one clinician works may produce good outcomes for patients. But it creates an asset that is only valuable while the person who shaped it is still present to update and validate it. A buyer looking at that system sees a black box that belongs to the person who is leaving.
When should you start the documentation work?
The standard answer is before the data room opens. A more useful answer is three to five years before you think you might sell. Buyers discount for risks they cannot see resolved within the deal timeline. If the documentation work starts after a letter of intent arrives, the price is already set. The buyer saw the gaps before you began filling them.
The documentation work has a natural sequence. The first stage is identifying which clinical decisions are currently undocumented. That means triage criteria, referral thresholds, escalation rules, and clinical pathways for common presentations. These are usually held by the founding clinician as experience rather than as written process. The second stage is capturing them in a form that a trained successor could follow.
The ICO’s guidance on AI and data protection is clear that any system making or assisting clinical decisions must have documented evidence of how it operates, what data it uses, and how clinical oversight is maintained. That documentation requirement is both a compliance obligation and, in an exit scenario, a due diligence asset. Practices that have met it proactively are in a different position than those that meet it under deadline pressure. A buyer’s diligence team will ask for consent records, audit trails, evidence of clinical oversight, and whether the AI systems the practice relies on have been formally documented or have simply accumulated over time. Undocumented AI is a liability in diligence, not an asset.
What should you think about before adopting any clinical AI tool?
The question to ask is whether each tool you bring into the practice generates transferable documentation or personalises something that belongs to one person. Any AI system the practice adopts should produce, not consume, the documentation that makes clinical knowledge portable. Every protocol captured, every decision rule written down, every consent process evidenced is material that a buyer can review and rely on.
Two things help with this. The first is treating clinical AI as a governance project from the beginning, alongside any productivity gains. The documentation standard, the consent framework, and the clinical oversight evidence all need to exist and be maintained regardless of whether a sale is on the horizon. The second is actively separating the practice’s clinical knowledge from the clinician who holds it. A practice that has done this work genuinely reduces the discount, because the buyer can see the evidence in the data room.
This is also where the founder-dependency problem and the AI question meet most usefully. Many clinical founders deploy AI to improve their own workflow, which is a reasonable starting point. The question is whether the AI being built is capturing something that will outlast the founder’s presence, or personalising something that depends on it. The distinction is rarely obvious at the point of adoption. If you want to book a conversation about where your clinical AI sits on that fork, that option is always there.
