Talk track

1. The point of this scenario isn't any single API — it's that six Creative Platform capabilities (Music, Voice Design, v3 TTS with audio tags, SFX, Dubbing, Image/Video Flows) land on one page, triggered by one event. Say that framing out loud before walking the breakdown, or it reads as six unrelated demos instead of one platform.
2. Use the Breakdown tab to explain which API produced which piece, then switch to Email or Landing Page to show the same underlying assets assembled as a finished deliverable with none of the module labels — that pairing is the point: an SE explains the machinery, a stakeholder just sees the birthday page.
3. Be upfront about the Voice Design workaround rather than letting it pass as a full demo of the capability: the Creative Platform MCP tools used to generate these assets only expose TTS against an existing voice, not voice creation itself, so the greeting uses a stock ElevenLabs voice (Alice) as a stand-in. The real integration path is creating the bespoke voice from the ElevenLabs dashboard's Voice Design feature directly, then swapping the voice_id these assets use. Naming this yourself, before a technical buyer finds it, is the stronger move.
4. Every asset here is pre-generated into storage (D4) — say plainly that this is a reliability choice for a live demo, not a platform limitation, and that the value story is content velocity and personalisation at scale, a genuinely different buying conversation from every Agents scenario's contact-deflection story.

Discovery questions

• How many personalised lifecycle touchpoints do you run today — birthday, anniversary, milestone — and how much of that is actually produced content versus a templated email?
• Of these six capabilities — voice, music, TTS, SFX, dubbing, generated visuals — which do you already use today, and which would be entirely new to your stack?
• How many languages or markets would a campaign like this realistically need to reach, and how is that localisation handled right now?
• What's your current cost and turnaround for producing one polished, multi-asset personalised campaign, compared to what you just watched happen on one page?

Objection handling

If a tool call fails or times out

1. An asset shows 'Not yet generated' unexpectedly on stage: don't try to trigger generation live — per D4 nothing here depends on live latency, so a pending asset means the wrong vertical/case is loaded or a reset is needed, not that something is broken mid-demo. Reset the scenario rather than waiting.
2. The live 'type a name, generate on the spot' TTS demo in the Breakdown tab fails or times out: acknowledge it plainly, fall back to the pre-generated spoken greeting already on the page, and note that this is the one moment in the scenario that does depend on live generation — everything else on the page does not.
3. Asked to prove this isn't just a static mockup: don't argue the point verbally — show the admin panel triggering a fresh send and the activity panel logging it, which is the concrete evidence the flow is real infrastructure, not a canned screenshot.