01
Market Assumption Debt.
The founder believes they know which segment will buy, at what volume, and on what timeline. The segment has not been validated through actual purchasing behaviour. Only through demo enthusiasm or discovery conversations with warm contacts.
Heard: We had seventeen discovery calls. Everyone said the problem was real. Then we went to market and nobody moved.
02
Positioning Assumption Debt.
The product is right. The pitch is wrong. In regulated markets this debt often manifests as positioning the product as autonomous when buyers require human oversight, triggering compliance rejection before the product is ever evaluated on its merits.
Heard: Buyers kept comparing us to ChatGPT. We are nothing like ChatGPT. We never fixed the message.
03
Champion Assumption Debt.
The person who loves the product cannot buy it. The founder has invested months of relationship capital in a contact who has enthusiasm but no purchasing authority. The deal goes quiet after security review.
Heard: My champion went quiet after the security review. I still do not know what happened in that room.
04
Pricing Assumption Debt.
The founder assumes buyers will pay the price needed to make unit economics work. Demo enthusiasm is not purchasing intent. The real test is what happens when the invoice arrives in procurement.
Heard: We dropped the price on the first ask. We never tested whether they would have paid the original number.
05
Hiring Assumption Debt.
The founder assumes a senior hire will solve a problem that is actually a market or product problem. The hire lands in a broken system and cannot perform. The constraint is not headcount. It is evidence.
Heard: We hired a VP Sales before we had a repeatable motion. He had no playbook to work from. We burned six months and eighty thousand pounds.
06
Fundraising Timing Assumption Debt.
The founder assumes they know when to raise, at what valuation, and on what narrative. Raising at the wrong moment on the wrong metrics locks in dilution that cannot be undone.
Heard: We raised too early on hype. Six months later the metrics we thought would hold up did not survive investor scrutiny.
07
Inference Cost Assumption Debt.
Demo margins are not production margins. Founders assume inference costs at demo scale reflect production economics. They do not. Every enterprise contract signed on demo pricing before production costs are understood compounds this debt.
Heard: Our demo margins looked amazing. Production margins collapsed at scale. We had already signed the contracts.
08
Product Assumption Debt.
The founder assumes the product solves the problem buyers actually have rather than the problem founders assumed they had. For AI products this often surfaces as building past the accuracy threshold buyers actually need.
Heard: We built for the wrong customer. By the time we figured it out we had eighteen months of technical decisions to unwind.
09
Channel Assumption Debt.
The founder assumes their go to market channel will work at the scale they need before testing it with real buyers at real price points. Early customers from the founder network are not evidence that the channel scales.
Heard: Our first ten customers all came through warm intros. When we went cold the CAC blew up and the motion stopped working.
10
Enterprise Deployment Assumption Debt.
The founder assumes their product integrates into buyer workflows and clears compliance requirements more easily than it does. Pilots succeed. Production deployment hits IT, governance, and compliance walls that were never visible in early conversations.
Heard: The pilot worked perfectly for six months. Then their IT security team got involved and the deployment stalled completely.
11
Platform Dependency Assumption Debt.
The founder assumes the capability gap their product fills will remain open long enough to build a defensible business. Foundation model providers update their core capabilities and render the product layer obsolete.
Heard: OpenAI shipped a native version of what we built. We had no data moat. We had to pivot in three months.
12
Regulatory Assumption Debt.
The founder assumes their AI architecture, data governance, and compliance posture will survive investor due diligence and enterprise procurement. In regulated markets this debt surfaces faster than anywhere else and at the worst possible moment.
Heard: We assumed GDPR was enough. The enterprise buyer asked for an EU AI Act conformity assessment and we had nothing to show them.
What is the difference between an assumption and assumption debt?
+
How do I know which of my assumptions are load-bearing?
+
Are all twelve categories relevant to every founder?
+
What is the most expensive category of assumption debt for AI founders in regulated markets?
+
How is assumption debt different from technical debt?
+