Artist Insights — Crawley Feedback & Editorial Rules

Consolidated from: 3/16 in-person meeting, 3/20 in-person meeting, 3/20 written feedback, multiple build iterations.

Non-Negotiable Editorial Rules

1. Show, Don't Tell

• Charts first, commentary second. Every data section needs a visualization.

• The primary purpose is getting people's head around where the artist is — not just insights and takeaways.

• People should be able to interpret the data themselves from the charts. Commentary supports the visual, it doesn't replace it.

2. Don't Be Too Self-Assured

• Present findings, don't make proclamations. "The data shows X" not "X is the story."

• Don't editorialize beyond what the data directly supports.

• If making an interpretive claim, hedge it. "This may suggest..." not "This means..."

• Exception: if something is genuinely striking (like a 6,795 index), call it out.

3. Every Deck Is Independent

• NEVER reference other artists' decks. No "two audiences again," no "similar to [other artist]."

• Treat each build as the first one ever made.

• The audience may not have seen any other deck.

4. Don't Force Department Insights

• This is an overview, not an operations document.

• Do NOT tag actions to departments (no "→ Commercial Partnerships" callouts).

• No ACTION / QUESTION / FLAG toggle system.

• If there's an overwhelming insight, note it once. Otherwise: present the data.

5. No Names

• No author names or analyst names in any output.

• Credit line: "Data: Strategic Analytics" only.

• Meg's and Crawley's names removed from published HTML. PPTX/PDF still need update.

V2 Feedback (3/16 Meeting)

• Linear scale on genre charts (no log)

• Tone down confidence in "5 Things" section

• Remove psych source line

• Kill "Thrifted Y2K" (too specific/editorial)

• More charts, less editorial throughout

• Each deck independent (no cross-references)

V5 Feedback (3/20 Meeting — Applied Literally)

Section Restructure

• Rename "Social & Content" → "Social Performance"

• Break socials into three sub-sections, each with individual commentary:

- Followers & Growth → commentary

- Engagement → commentary

- Posting Frequency → commentary

• Lead with follower counts

Data Presentation

• Remove ALL P-notation. Say "92nd percentile — excellent" or "92nd percentile, Very Strong" in plain English.

• Simplify benchmark display

• Top 15-20% is fine as description

• Time windows: one year, six months, three months — people get those

Specific Fixes

• Restore missing age distribution commentary

• Fix gender commentary (remove "significantly female")

• Update gender colors from pink/blue to chartreuse/teal (pink/blue is outdated)

Commentary Style

Crawley's exact framing example: "92nd percentile, excellent. However, more recent windows show the growth rate is slowing down slightly rather than accelerating."

Scorecard

• Likes it as proof of concept, not needed for every artist yet

• Wants to see what builds the score (not a black box)

• Build order: follower growth (done) → streaming growth → TikTok growth → release performance → catalog strength

Color System (V6 — Locked)

Crawley's core confusion: chartreuse was doing too many jobs. Fixed with shade-based system:

Color	Hex	Purpose
Chartreuse	#CDF851	Brand/structural ONLY
data-us	#B8E64A	US/Instagram data identity
eval-green	#4ADE80	Positive evaluation
eval-red	#EF4444	Negative evaluation
Flame	#FF4A23	TikTok identity EXCLUSIVELY

Rule: Every color has ONE meaning. Decorative vs. evaluative must be visually distinct.

Instagram Follower Growth Benchmark Framework

Crawley built a percentile framework for IG follower growth. Applied to horsegiirL:

• P87 composite (Very Strong, Stable momentum) in the 200K-500K tier

• Tier matters: growth in 200K-500K tier is different from growth in 5M+ tier

A&R Evaluation Framework (Next Phase — Jordan/Crawley Direction)

Goal: Build KPI evaluation framework:

1. For each metric (engagement, follower growth, streams), use AI + all artist data to determine what "good" looks like

2. Segment by artist tier

3. Combine into overall artist score

4. Eventually apply to unsigned artists as A&R eval tool

Crawley's key insight on volume vs. growth:

An artist doing 1M streams consistently with no growth is more valuable than one growing from 118 to 186. Volume matters, not just growth. Need a way to benchmark and contextualize volume — "even if you're not growing, it's okay because your sheer volume is impressive."

Outputs (Always Produce All Three + CSV)

1. Interactive HTML — chart-heavy, dark-mode, sticky nav, Chart.js. Primary deliverable.

2. PPTX deck — 11 slides, brutalist corporate dark mode. For meetings.

3. PDF — converted from PPTX. For send-arounds.

4. Psychographics CSV — generated from web research. For Tableau.

Artist Build Status

Artist	HTML	PPTX	PDF	Psych CSV	Notes
CA7RIEL & Paco Amoroso	✅	✅	✅	✅	Complete. Names need removal from PPTX/PDF.
horsegiirL	V6	Needs rebuild to match V6	Needs rebuild	✅	V5 (literal) + V6 (editorial). Photos placeholder.
Sleepy Hallow	—	—	—	—	Next in queue.