An Intro to Entrepreneurship Syllabus That Assumes AI Exists

Most introductory entrepreneurship syllabi were written when building a product was hard. That assumption is doing quiet damage, because almost every structural choice in the standard course depends on it.

The classic shape is: weeks one to five on concepts, weeks six to ten on planning, weeks eleven to fifteen on building and pitching. It worked because the build in weeks eleven to fifteen was a genuine constraint. Students learned by hitting it.

A student can now have a working application before the first class. The constraint is gone, and with it the part of the course that was doing the teaching.

Here is a fifteen-week outline that puts the constraint back somewhere it still exists: in front of real customers. I taught this material at BYU for twelve years and have run it since across more than fifty bootcamp cohorts, so the notes at each stage are about what actually goes wrong, not what should happen.

The four graded deliverables

Before the week list, the spine. Everything else is support for these.

Deliverable	Week	Weight
Assumption memo	3	15%
Customer discovery round one	6	20%
Customer discovery round two	11	25%
Evidence review and recommendation	15	30%

Participation and in-class exercises make up the remaining 10%.

Note what is absent: there is no business plan, and no pitch deck as a graded artifact. Students may present in week fifteen, but the grade attaches to the evidence behind the recommendation, not the delivery.

Weeks 1 to 3: finding a problem worth solving

Week 1. Why most startups fail, with specifics. Not a motivational lecture. Walk through three or four companies that died of a problem nobody had. Students arrive believing failure comes from execution. It mostly comes from building something unwanted, and the earlier they internalise that, the more the rest of the course makes sense.

Week 2. Problem selection. Students bring three problems they have personally experienced or observed closely. Rule: no problem they have only read about. The quality of the whole semester is set here, and a student who picks a problem they have no access to will struggle in week six when they need to find interviewees.

Week 3. Assumptions and falsifiability. The first deliverable. Students write what they believe about the customer, the problem, and the willingness to pay, in a form that could be proven wrong. Then they mark which single belief, if false, kills the idea.

Instructor note: this is the week that determines whether the course works. Collect the memo on a fixed date and do not accept revisions afterward, because it is the only artifact in the course a student cannot reconstruct after the fact. Everything downstream is graded against it.

Weeks 4 to 6: first contact

Week 4. How to run a customer interview. Teach the mechanics: open questions, past behaviour rather than future intention, silence, and never pitching. The single most useful rule to give them is to ask what the person does today and what it costs them, rather than whether they would like a better way.

Week 5. In-class interview practice. Students interview each other about a real problem, badly, in front of the room. It is uncomfortable and it is the highest-value hour in the course. They discover that they pitch instead of listen, and they cannot unlearn that once they have seen it on themselves.

Week 6. Round one due: five interviews. Graded gently. The purpose of this round is to get the fear out of the way, and the grading should say so.

Instructor note: expect roughly half the class to have interviewed friends. Name it without punishing it, then price it properly in round two.

Weeks 7 to 10: the business model

Week 7. Business model canvas. The nine blocks, with the emphasis on the fact that a canvas is a set of hypotheses rather than a plan. Students fill it in forty minutes, which teaches them how little they know.

Week 8. Customer segments and the trap of “everyone.” Narrowing practice. A student who cannot name the single person who feels the problem most acutely does not yet have a customer.

Week 9. Revenue, pricing, and what people actually pay for. Keep this practical. Pricing conversations in interviews, not spreadsheets.

Week 10. Channels and distribution. The most neglected block on the canvas and a common cause of death. How will this person find out the thing exists?

Weeks 11 to 13: evidence

Week 11. Round two due: ten interviews, strangers only. Graded hard, against the week three memo. Full marks require reporting what contradicted them.

Week 12. Pivot or persevere. Each student presents what changed. The class pressure-tests whether the change was driven by evidence or by preference.

Week 13. Returning to a previous interviewee. An underused exercise. Students go back to one person they spoke to in round one and show them how their thinking has changed. Second conversations are where the real information is, and almost nobody does this unprompted.

Weeks 14 to 15: the recommendation

Week 14. Writing the evidence review. Not a pitch. A memo: here is what I believed, here is what I did, here is what I learned, here is what I recommend and why. Killing the idea with good evidence earns full marks, and you must say this early and repeatedly or no student will believe you.

Week 15. Presentations and recommendations. Students present the delta. Invite someone from outside, ideally an investor or operator, to ask the uncomfortable question.

What to cut if you have ten weeks

Drop week 9 and week 13, and merge weeks 7 and 8. Keep both customer discovery rounds. If you only have room for one round, the course becomes a lecture series with homework attached.

The AI question, handled once

Students will use AI. The useful posture is not prohibition, which is unenforceable, but placement. AI is excellent at helping a student prepare interview questions, summarise transcripts they actually gathered, and spot patterns across conversations that happened. It cannot generate the evidence itself, because a synthetic customer agrees with you, and agreement is the one thing this entire course is designed to guard against.

Make that argument in week one, in those terms, and most of the enforcement problem goes away. For the part that does not, require provenance rather than detection: interview dates, raw notes, and a contactable name for a sample you will spot-check. I wrote up the grading side of that in more detail in a customer discovery rubric that survives AI.

---

Try the ToolSuite free with your class Customer discovery and business model canvas tools built for coursework. Free to start, with academic pricing for programs. Check out our software free →

If you are running a course on this shape, the customer discovery and business model canvas tools in Startup Ignition ToolSuite were built for it. Students log interviews with dates and sources, the canvas keeps version history so the assumption changes are visible between class sessions, and the scored outputs give you something gradeable that is not a presentation. Academic pricing is available, and it sits alongside whatever LMS you already use.