Training Tuesday · AI Pro Writers Studio

Training Tuesday

Stop Guessing at Your Comps — Reverse-Engineer Them Instead

Topic: Using Book Reporter (CONAN) and Series Reporter (SONJA) to forensically analyze comp titles before you write a word of your own story

The problem this solves

Most of us pick comp titles the same way: we read a book we loved, we feel something, and we say "I want to write something like that." Then we sit down to plan our own story and realize "like that" was never actually specific. Was it the pacing? The trope execution? The voice? The way the series built across four books instead of one?

Vibes aren't a blueprint. If you can't name exactly what a comp title is doing — structurally, at the sentence level, across its beats — you can't reliably reproduce what made it work, and you definitely can't improve on it.

That's what Book Reporter and Series Reporter are for. They're the forensic layer of the Professional Writing System — the tools you run before Phase 1 (Story Development Room) to turn "I want something like this" into an actual architecture you can build from.

The two tools, and how they're different

Book Reporter (CONAN) analyzes one book. He reads it the way a mechanic pulls an engine — not to admire it, but to understand exactly what's load-bearing and what's cosmetic. He produces a Forensic Morphology Report: 15 layers covering structure, character architecture, prose voice, world-building, tropes, emotional design, and hooks, plus:

A Beat Rhythm Map — a chapter-by-chapter tension chart that shows you exactly where the book accelerates, where it breathes, and where it stalls (dead zones, flagged by name)

Genre Beat Benchmarks — where this book's inciting incident, midpoint, dark night, and climax land as percentages of total length, compared against genre norms

A Prose Quality Score Card — eight dimensions scored 1–10, so "the writing is good" becomes "voice distinctiveness is an 8, but foreshadowing/payoff execution is a 4, and here's why"

A Reconstruction Brief — the condensed DNA of the book (genre, structure, archetypes, thematic argument, world type, tropes) formatted to drop directly into Story Development Room, plus 2–3 original seed concepts that share the book's structural DNA without copying it

CONAN works on one book per session. Full text gives you the deepest report; title-and-author alone still gets you a research-based analysis, clearly flagged as such.

Series Reporter (SONJA) is the sibling skill — same forensic philosophy, wider lens. Where CONAN goes deep on one book, SONJA reads across installments to answer questions CONAN structurally can't: Does this series compound, or does each book just reset? Do the promises made in Book 1 pay off by Book 4? Does the author's voice hold steady across six books, or drift?

SONJA's report has two tiers:

Tier 1 — Per-Book Profiles: a condensed six-point read on each installment (premise, beat placement, character arc snapshot, promises made, promises paid)

Tier 2 — Cross-Series Synthesis: eight sections including the Promise/Payoff Ledger (every setup tracked to its payoff or flagged as broken/overdue), the Character Evolution Audit (is the arc credible and compounding, or repetitive and reset every book?), Series Pacing Diagnosis, and Voice & Prose Consistency across the run

SONJA doesn't read raw manuscripts — she works from Book Reporter reports (preferred) or from title-and-knowledge for a published comp series. If you only have one book analyzed, she'll tell you synthesis needs at least two installments and point you back to CONAN.

The important thing: these tools are sequential, not competing. CONAN goes book by book. SONJA takes CONAN's output and reads the whole shelf.

Book Reporter (per book) ──► Series Reporter ──► Series Bible Creator seed

│

├──► Story Development Room (new installment)

└──► Market Scout

The workflow: comping a series before you outline your own

Here's the actual sequence, step by step.

1. Pick 2–3 comp titles — book and series comps if relevant. If you're planning a trilogy, don't just comp the vibe of one standalone. Pick at least one comp that's also a series, so you can study how someone else built a multi-book arc, not just a single-book one.

2. Run each installment through Book Reporter separately. Give CONAN full text where you can (excerpts or the full manuscript get you the deepest layer — the Beat Rhythm Map and Prose Quality Score both require it). For published books you don't have text access to, title + author still works — CONAN flags it as research-based and pulls in web research on category placement, Goodreads shelf tags, and comps.

3. Feed the Book Reporter reports to Series Reporter. This is the step people skip, and it's the one that actually earns you series-level insight. Hand SONJA the CONAN reports for each installment (or title + author if you're doing research-based comps for a whole published series). She'll build the per-book profiles, then run the Promise/Payoff Ledger and Character Evolution Audit across the set.

4. Read the Promise/Payoff Ledger like a checklist for your own series. This is the single most useful artifact in the whole workflow. SONJA tracks every promise the comp series makes — narrative, emotional, character, thematic — and marks each one Paid (earned or exceeded), Paid but underdeveloped, Open and overdue, or flat-out Broken. That's a direct list of traps to avoid in your own outline: if the comp series' Book 3 broke a promise made in Book 1, you now know exactly the kind of setup discipline your own series needs that theirs didn't have.

5. Take the Reconstruction Brief and Series Bible seed into Story Development Room. CONAN's Reconstruction Brief and SONJA's Series Bible Creator Initialization Handoff are both formatted for direct handoff — the Brief drops into Phase 1 (Story Development Room) as your architectural starting point; the seed drops into Series Bible Creator Mode 2 if you're planning a series from the outset. You are never copying the comp's plot — you're borrowing its proven architecture (pacing rhythm, promise discipline, arc-compounding pattern) and building an original story on top of it.

What this catches that "just reading the comps" doesn't

A few things came up repeatedly in testing this workflow that are easy to miss reading casually:

Dead zones you didn't notice because you were enjoying the book. The Beat Rhythm Map flags chapters where narrative momentum actually drops — often invisible on a first read because voice or character charm is carrying you through, but structurally still a stall.

Beats landing "off" from genre norm without you clocking it. A midpoint at 41% instead of 48–52% might be the exact reason a book felt like it dragged in the middle third — even if you couldn't have named why at the time.

Promises that were never actually paid. Readers forgive individual books for a lot. They remember, at the series level, when a Book 1 hook goes unanswered by Book 4. SONJA's ledger surfaces this explicitly instead of leaving it as a vague "something felt unfinished" impression.

Voice drift across a long series. If your comp author's prose voice shifted noticeably by book five, that's either a deliberate craft choice or a warning sign — and now you know which, going in.

Try this before your next project

If you're about to start a new series, don't skip straight to Story Development Room with a vibe and a Pinterest board. Pick your top 2 comp titles — ideally ones that are also series — run them through Book Reporter, then Series Reporter, and actually read the Promise/Payoff Ledger before you outline a single beat of your own.

You'll walk into Phase 1 with a Reconstruction Brief, a Series Bible seed, and a specific list of structural traps to avoid — instead of a feeling you're hoping translates into a finished manuscript.

This is a Training Tuesday post for the Professional Writing System community.

Questions or results from running this on your own comps? Drop them in the thread below.

Always testing...The first image is Grok, the second is Nano Banana and the third is GPT2 all from the text above using the image gallery in WCP

6 comments

Training Tuesday