Ask how many demos an SE should run weekly and you will hear a capacity answer: as many as the calendar survives. The better question is which demos deserve an SE at all. Run the arithmetic honestly and the answer restructures the pre-sales week: humans for judgment demos, agents for the repeatable rest.
The capacity math every pre-sales leader already knows
Count an SE's real week: meetings, prep, follow-up documentation, internal syncs, POC work. The hours left for demos are fewer than anyone admits, and each demo costs prep and recovery time around the calendar slot itself. Now put that against demand: every AE wanting SE coverage, every mid-funnel deal requesting a technical session, plus the long tail of smaller deals and trials that never get coverage and quietly convert worse. The demand curve and the capacity line have never intersected at any company, ever. Rationing was the industry's answer, and rationing is a strategy of pre-decided losses.
The segmentation the math forces
Sort demo requests into three piles. Judgment demos: complex architectures, strategic accounts, bake-offs, anything where reading the room changes the outcome; these are worth every SE hour and deserve more prep than they currently get. Repeatable demos: standard walkthroughs of documented capability for qualified buyers; these follow patterns and an SE adds politeness, not information. Uncovered demand: everyone who never got a session, self-served badly, or churned in the queue. The first pile is the SE's actual job. The second and third are agent work: a live demo agent runs them at unlimited parallel volume, qualifies as it goes, and escalates the ones that reveal judgment-level complexity. The right SE demo number is not a bigger one; it is a purer one.
What changes when the math is applied
SE calendars shift toward the deals where technical selling wins or loses money. Coverage extends to the entire funnel, including the tail that was never worth a human slot. Every demo, human or agent, generates a record: questions asked, objections raised, features that landed, which turns pre-sales from anecdote into data. And the SE role rises in seniority as its repetitive floor disappears, the same pattern playing out across onboarding and support. (Disclosure: we build Skippr; Skippr AI demo runs the repeatable pile and hands SEs the judgment pile with full context, which is the division this whole essay argues for.)
Questions buyers actually ask
What is the right SE-to-AE ratio once agents cover repeatable demos?
It becomes a judgment-workload question rather than a coverage question; teams typically rediscover the ratio from their own escalation data within a quarter.
Do AEs accept agent-run demos for their deals?
They accept booked meetings with briefed buyers; the resistance fades the first time a calendar arrives with context attached.
See it rather than read about it
The difference between a recording and a live agent is hard to argue and easy to watch. Fifteen minutes is enough to judge whether it fits your motion.