The Synthesis Playbook — AI-Assisted Recipes for Turning Workshop Outputs Into Decisions
The Facilitation Series — Book 6

The Synthesis Playbook

Six proven recipes — and the four moves behind them — for design sprints, retros, discovery, and strategic offsites. Prompts included.

auto_stories 12 chapters + recipe index schedule ~3 hr read person Intermediate

Coming in June — get notified

About this book

The workshop ends Friday afternoon. You were good in the room — teams aligned, decisions made, energy high. Then everyone leaves, and you're staring at a wall of stickies, a pile of transcripts, and half-legible whiteboard photos, facing a lonely weekend of turning it all into something the client will respect.

This is the part of facilitation nobody trained you for, and the part that eats your weekends. The fear is real and it cuts both ways: miss the signal, or flatten the dissent that made the session worth running. AI could help — but generic "50 prompts" advice produces tidy, confident, wrong answers. This book exists to prevent that synthesis slop.

The promise is plain: finish the workshop Friday, hand the client a defensible synthesis by Sunday — one proven recipe per workshop type, with the prompts that actually work. The governing principle: human-authored, machine-assisted. AI clusters; you decide.

The four moves

Every recipe in the book is built from the same four primitives. Learn them once in Part 1, then compose them into any workshop type. The split between what AI owns and what you own never changes.

workspaces

Cluster

Group raw outputs into candidate themes. AI-led, human-checked — the one move the model is genuinely good at.

psychology

Interpret

Turn groupings into insight statements — the "so what." Human-led, AI-drafted. Where slop sneaks in if you let it.

format_list_numbered

Prioritize

Rank, sequence, force the choice. Human-owned; the model surfaces options but never makes the call.

edit_note

Narrate

Shape meaning into the client-facing story. Human-authored, AI assists the prose. Your name is on the deliverable.

What you'll learn

  • check_circle The four moves of synthesis — cluster, interpret, prioritize, narrate — as reusable primitives
  • check_circle One complete, tested recipe per workshop type — inputs, outputs, workflow, prompts, pitfalls, worked example
  • check_circle Exactly where AI helps (clustering, first-pass theming) and where to keep it out (the politics)
  • check_circle Copy-ready prompts that preserve nuance instead of flattening it
  • check_circle How to build your own reusable prompt library — and a gentle on-ramp to agents
  • check_circle What to show the client and what to keep in the working layer

Chapters

Part 1: Synthesis fundamentals
  1. 01

    Synthesis is a craft, not a summary

    Deciding what a room meant, not compressing what it said. Summary vs. synthesis vs. slop.

  2. 02

    The four moves

    Cluster, interpret, prioritize, narrate — defined as reusable primitives, with the AI-vs-human split for each.

  3. 03

    Your synthesis stack, and where AI helps vs. hurts

    Transcript tool, LLM, clustering, output format — plus the AI-off zones you'll defend in every recipe.

Part 2: The recipes
  1. 04

    Discovery & qualitative synthesis

    Interview transcripts → insight statements → opportunity areas. The anchor recipe. [Cluster] + [Interpret].

  2. 05

    Strategic offsite & stakeholder politics

    The "AI hurts" chapter — conflicting leadership input into a defensible recommendation. [Interpret] + [Narrate].

  3. 06

    Design sprint synthesis

    "Decide" outputs + user-test notes → a Monday-morning ship / iterate / kill call. [Prioritize] + [Narrate].

  4. 07

    Ideation & roadmap synthesis

    200 ideas → a ranked shortlist with rationale → sequenced bets. [Cluster] + [Prioritize].

  5. 08

    Retro & post-mortem synthesis

    Themes and actions without flattening dissent or assigning blame. [Cluster] + [Interpret], dissent-preserving.

  6. 09

    Training & workshop debrief synthesis

    Feedback forms + field notes → what to change next time + a skills-gap signal. [Cluster] + [Narrate].

Part 3: Building your own synthesis system
  1. 10

    Your reusable prompt library

    Turn the book's prompts into a saved, versioned library tuned to your clients and tools.

  2. 11

    A gentle on-ramp to agents

    From copy-paste prompts to a repeatable, semi-automated workflow — and where Workshopr fits.

  3. 12

    Client-facing artifacts — what to show, what to hide

    The deliverable vs. the working layer, showing your authorship, and the AI-disclosure question.

menu_book

Bonus: The recipe index + the prompt appendix

A one-page cheat sheet — workshop type → recipe → moves — plus every prompt collected and copy-ready.

The companion tool

terminal

Workshop Synthesizer — the recipes, as slash commands

Six recipes from the book turned into a Claude Code plugin. The mechanical work — clustering, formatting, transitions — gets handled; the tool deliberately pauses on the interpretive steps that need you. Same principle as the book: human-authored, machine-assisted. Explore the Workshop Synthesizer →

Who this is for

person_pin

Solo consultants

You run client workshops and own the weekend that follows. You need a defensible synthesis by Monday, alone.

groups

Internal facilitators

You synthesize for your own org. If You're Internal sidebars cover where the politics differ.

smart_toy

The AI-curious

You're not a skeptic, but you've been burned by generic prompt advice. You want recipes that fit the workshop you just ran.

Previous book arrow_back Did It Work?