Your AI Doesn't Give Bad Demo Advice — You Gave It Nothing [53]
AI can effectively coach demo performance, but only when leaders provide rich context about their ICP, personas, and standards. Generic AI advice fails because it lacks situational awareness; the solution is building a shared, context-loaded assistant that scores demo transcripts against company-specific criteria rather than generic best practices.
Summary
The episode challenges the assumption that AI cannot coach demos effectively. It opens with a real example where someone asked AI for demo tips and received generic advice (focus on decision makers, don't let end users derail you, hit economic buyer motives first). While not wrong, this advice was useless because the prompt contained zero context about audience, deal stage, or sales motion. The hosts argue this generic guidance can actually be harmful—for example, telling an SE to focus on the economic buyer when facing five admins who want a specific workflow results in a disconnected conversation.
The core argument is that the problem isn't AI's capability but rather leadership's responsibility to feed it information. Two key structural elements must be in place: first, a centralized shared assistant loaded with company-specific knowledge (ICP, personas, demo standards, methodology); second, a system for SEs to upload demo transcripts and receive feedback scored against those internal criteria rather than generic books.
Nate and Ava discuss implementation, revealing that one director built this setup in a day and a half using existing enablement materials. The coaching should operate in two separate layers: structural decisions (who's in the room, pacing) and demo craft (opening, clicking, narration). The transcript analysis is highlighted as particularly powerful—by the third session, SEs can see concrete improvements like stopping unnecessary click narration. The hosts address concerns about standardization by proposing a split: team-owned base layer (non-negotiable standards) and SE-owned personal layer (how the assistant communicates with that individual). They acknowledge different team members need different coaching approaches. Finally, they clarify this augments rather than replaces human manager feedback—humans should provide judgment about context (like knowing the CFO's boss was replaced), while AI handles pattern detection in demos.
About this episode
Nate saw a one-line prompt produce textbook-generic demo advice, and a comment section full of people saying "it needs more context" without ever building any. He and Ava work out how SE leaders turn that context into a shared enablement asset — and what it does for coaching between 1:1s.
Key Insights
- The hosts argue that generic AI advice about demos is actively harmful when it lacks context—an SE told to focus on the economic buyer when facing end users creates a disconnected room, not a successful demo.
- Nate and Ava claim that companies have already written the necessary context (in enablement decks and standards) but haven't made it machine-readable, so the real leadership job is translating existing knowledge into a shared assistant's prompt rather than creating new content.
- The hosts contend that demo transcript scoring generates behavioral change faster than traditional enablement because it provides immediate feedback and demonstrates concrete improvement over multiple sessions, unlike deck-based training that asks for attention without returning value.
Topics
Transcript
Hey there and welcome to Leading Pre-Sales, the show for solution engineering leaders who want to build teams that drive revenue and not just demos. My name is Tim and I'm the co-founder of SE Rockstars and together with Jan, we've coached over 350 solution engineers and their leaders across several dozens of companies. Every conversation you hear on this show is based on real coaching situations, real challenges, real problems that as leaders like you are dealing with right now. None of this is made up. We use AI to bring these stories to life through our two hosts, Nate and Ava, but the insights come straight from the trenches. Each episode gives you one actionable takeaway you can…
Full transcript available for MurmurCast members
Sign Up to AccessMore from Leading PreSales | The Solution Engineering Leadership Show
Core vs. Context: When Your SEs Should Let a Build Deal Go [92]
The episode teaches solution engineering leaders to distinguish between core and context business functions when customers consider building custom solutions versus buying software. SEs should let lightweight, context-focused builds go while fighting harder for core capabilities, and leaders must publicly validate this approach to change team culture.
The Iceberg Under Every Build Deal [91]
Solution engineering leaders need a standardized five-year cost model to evaluate Build vs. Buy decisions in the AI era, grounded in hidden maintenance costs and opportunity expenses rather than vendor pitches. The key is creating an 'iceberg checklist' of often-overlooked costs and having customers co-create the financial analysis rather than presenting a pre-built case.
Leaderboards, Not Cameras [89]
The episode examines why AI adoption dashboards and leaderboards often fail to drive behavior change, and proposes an alternative approach: publicly sharing AI workflows that actually shipped on real deals, ranked by peer reuse rather than usage metrics.
Your SEs Just Inherited the Customer's Bill [88]
Solution engineers must own usage estimation and sizing in consumption-based pricing models, even though they don't set prices. When SEs fail to accurately estimate how often workflows trigger, customers face billing surprises that damage relationships and renewals.
Rebrand or Real Role? The FDE Test [86]
The episode examines whether Forward Deployed Engineers (FDEs) represent a genuinely different role from Solution Engineers or merely a rebranding strategy. Through analysis of real job postings and market trends, the hosts present a three-question test to distinguish authentic FDE positions from SE roles with inflated titles.