Skip to content This library is free, offered as dana. If it helps your work, donate here →
penguin tree ai
← back to libraryDonate here

This is one of 3,622 free prompts in the penguin tree ai library. Sign up once and the whole library opens, with new prompts by email.

open the library →

Shadow P&L builder

Reverse-engineers a competitor's income statement from public disclosures, press releases, job postings, sell-side research, and industry benchmarks into a fully estimated shadow P&L — with revenue build, cost structure…

download .md

Purpose

Construct a defensible estimated income statement for a competitor — public segment, private company, or a business unit that isn't separately reported — by triangulating public data points into a structured revenue build and cost stack, with an explicit confidence rating on every assumption. Turns scattered disclosures into a board-ready view of a rival's margin profile and cost leverage.

When to Use

Invoke this skill when the user asks to:

  • Estimate a competitor's revenue, gross margin, or operating margin from public information
  • Build a cost model for a private or non-reporting competitor
  • Reverse-engineer unit economics or contribution margin for a rival product line
  • Size the financials of a specific business segment a public company doesn't break out
  • Prepare competitive cost benchmarking ahead of a pricing, M&A, or strategy review
  • Pressure-test a competitor's ability to sustain a price cut or fund a new initiative

Inputs Required

Before starting, collect:

  • Competitor: Company or segment name and a one-line description of the business being modeled
  • Available sources: What's on hand — 10-Ks/10-Qs, earnings call transcripts, press releases, sell-side notes, job postings, headcount estimates, pricing pages, industry reports
  • What's already known: Any confirmed figures (reported revenue, headcount, ARPU, store count, GMV, etc.)
  • Business model type: SaaS / marketplace / hardware / retail / services / manufacturing / hybrid
  • Fiscal period to model: The year or quarter the shadow P&L should represent
  • Purpose of the model: Pricing decision / M&A screen / margin benchmarking / board briefing — this sets the required precision
  • Peer set: 2-4 comparable public companies to source cost-structure benchmarks from

If the peer set or business model type is missing, infer the closest public comparables and note the assumption. If no confirmed revenue anchor exists, state that clearly and treat the entire revenue line as low-confidence.

Instructions

Step 1: Establish the Revenue Anchor

Find the single most reliable top-line data point and build outward from it.

  • Rank available signals by reliability: reported segment revenue > management-guided figures > sell-side consensus > unit-count × price (e.g., customers × ARPU, stores × per-store sales, GMV × take rate) > headcount-based backsolve (revenue/employee benchmark from peers)
  • Build the revenue two ways where data allows (e.g., top-down TAM/share and bottom-up unit economics) and reconcile the gap
  • State the anchor explicitly: name the source, the figure, and the derivation method
  • Assign a confidence rating: High (disclosed/guided), Medium (triangulated from 2+ signals), Low (single-signal estimate)

Step 2: Decompose Revenue by Stream

Break the top line into its components so cost logic can attach cleanly.

  • Split by product line, geography, or recurring vs. one-time as the disclosures support
  • For SaaS: separate subscription, services, and usage revenue; note net revenue retention if inferable
  • For marketplaces: separate GMV from net revenue (take rate) — never model cost against GMV
  • For hardware/retail: separate product from attach/services revenue
  • Flag any stream that is pure estimate versus disclosed

Step 3: Build the Cost Stack Line by Line

Estimate each cost line using peer benchmarks anchored to the competitor's specifics. Work down the P&L:

  • Cost of Revenue / COGS: Start from the peer-set gross margin range, then adjust for known structural differences (scale, cloud vs. on-prem, in-house vs. outsourced manufacturing, hosting intensity). Output an estimated gross margin band, not a point estimate.
  • Sales & Marketing: Triangulate from job postings (sales headcount), reported CAC or payback where available, ad-spend disclosures, and peer S&M-as-%-of-revenue. Flag whether the company is in land-grab (elevated S&M) or efficiency mode.
  • Research & Development: Anchor to engineering headcount (LinkedIn/job postings × fully-loaded cost per engineer) cross-checked against peer R&D-as-%-of-revenue.
  • General & Administrative: Use peer G&A ratios, adjusted for public-company overhead if applicable and for M&A/legal intensity.
  • Stock-Based Compensation: Estimate separately for tech/pre-IPO competitors — it materially distorts GAAP margins and should be called out as a reconciling item.

For each line: state the benchmark used, the adjustment logic, the estimated value or range, and a confidence rating.

Step 4: Assemble the Margin Waterfall

Roll the estimates into a structured P&L:

  • Revenue → Gross Profit → Contribution/Operating Expenses → EBITDA → Operating Income (EBIT)
  • Show both a GAAP view and an adjusted/cash view (ex-SBC, ex one-time items) where relevant
  • Present each line as a value and as a % of revenue
  • Where a line is a range, carry the range through to the margin lines

Step 5: Run a Sensitivity Range

Convert the point model into low / base / high scenarios.

  • Identify the 2-3 assumptions that most move operating margin (usually revenue anchor, gross margin, and S&M intensity)
  • Produce a low / base / high operating margin outcome by flexing those drivers within their confidence bands
  • State which assumption the whole model is most fragile to

Step 6: Interpretation for the Stated Purpose

Translate numbers into a decision-relevant read:

  • What the margin profile implies about the competitor's pricing headroom, ability to fund growth, or M&A affordability
  • Where the estimate is most likely wrong and what public data point would tighten it fastest
  • One or two "if this is right, then…" strategic implications tied to the model's purpose

Output Format

  • Structure: Narrative-light, model-forward — a revenue build section, a cost stack section, a formatted shadow P&L table, a sensitivity table, and an interpretation section
  • Length: 900–1,300 words plus tables
  • Format: Two tables (the shadow P&L with value and %-of-revenue columns; the low/base/high sensitivity table) surrounded by structured supporting prose
  • Sections: Revenue Anchor & Build → Revenue by Stream → Cost Stack (line by line) → Shadow P&L Table → Sensitivity Range → Interpretation & Fragility
  • Tone: Strategic finance analyst — precise, assumption-explicit, appropriately hedged. Every estimate is labeled as an estimate; no false precision.

Quality Checklist

Before delivering output, verify:

  • [ ] The revenue anchor names its source and derivation method, and revenue is built at least two ways where data allows
  • [ ] Every P&L line (COGS, S&M, R&D, G&A, SBC) carries a benchmark reference and a High/Medium/Low confidence rating
  • [ ] The shadow P&L table shows both absolute value and % of revenue for each line
  • [ ] GAAP and ex-SBC views are distinguished for any tech or pre-IPO competitor
  • [ ] The sensitivity table shows low/base/high operating margin driven by named assumptions
  • [ ] The single most fragile assumption is explicitly called out
  • [ ] No figure is presented as certain when it was derived from a single unverified signal

Notes

  • Marketplaces are the most common modeling error: attaching operating costs to GMV instead of net revenue overstates the cost base by 5–20x. Always confirm whether a disclosed figure is gross volume or net revenue.
  • Headcount-based backsolves (engineers × loaded cost, sales reps × quota) are often the most reliable estimator for private competitors because hiring data is public even when financials aren't.
  • Resist the urge to output a clean point estimate — the value of a shadow P&L is the labeled uncertainty. A ranged answer with confidence flags is more decision-useful than a false-precision number.
  • Revisit the model each time the competitor reports or raises capital; new disclosures usually re-anchor at least one line and can collapse the sensitivity range meaningfully.

How to use this skill

This file is a Claude Code skill in Anthropic's standard format — markdown with YAML frontmatter at the top.

Installing in Claude Code:

  • ~/.claude/skills/<folder-name>/SKILL.md — auto-loads in every session anywhere on your machine (recommended for skills you'll use across multiple projects)
  • <your-project>/.claude/skills/<folder-name>/SKILL.md — auto-loads only when working in that project (recommended for project-specific skills)

Restart Claude Code or open a new session to register.

Using in Claude.ai, ChatGPT, Gemini, or other AI tools:

  1. Drag this file into a new chat, or copy and paste its contents
  2. Send your task or request as your next message — the AI will apply the skill where it fits

For Claude Code skill details: https://docs.anthropic.com/en/docs/claude-code/skills