Pattern: Peer-gap chart

What it is

A single visual that benchmarks the target against its best-in-class peers on a single quantitative KPI — and shows the target losing. The gap is the punchline.

Why it works

• It's the most credible claim in the deck because it's arithmetic, not opinion.
• It's immediately actionable — close this gap and value accrues.
• It pre-empts the "context is different" defence by pitting the target against direct peers in the same category.

When to use it

• ✅ The target is a public company with ≥3 comparable listed peers
• ✅ There's a measurable KPI where the target lags (operating margin, ROIC, same-store sales, EBITDA margin, operating ratio, revenue growth, cost per unit, FCF conversion, NPS, employee productivity, etc.)
• ❌ Don't force it when the peer set is weak (the comeback is "these aren't real comps")

Recipe

1. Pick one KPI. Not three. One. The one where the gap is widest.
2. Choose 4–6 peers. Real operating comps, not stretch comps. Keep the set defensible.
3. Use a bar chart. Sorted descending (or ascending if "lower is better").
4. Colour the target differently — typically a warning colour (red/orange) against the peers' neutral grey.
5. Label the gap in the title, not the chart legend. Title carries the punch. E.g.: "CP has the worst operating ratio of all Class I railroads."
6. Annotate the "winner" — the peer whose performance you're implicitly challenging the target to match.
7. One extra line: what closing the gap would deliver. "Matching CN's 63% operating ratio would add $X per share."

Headline language that works

• "Company X has the worst / highest / lowest [metric] in its peer group."
• "Every peer we studied outperforms on [metric] by [X]pp / [X]%."
• "To match the peer median, the target must improve [metric] by [X] points."

Common mistakes

1. Too many KPIs → the reader doesn't know what to focus on. One chart, one metric.
2. Wrong peer set → the target's PR will list "more appropriate" peers. Pre-empt this by using peers the target itself cites in its own 10-K or investor day.
3. Outdated data → any comp older than the last fiscal year is vulnerable to "that's not current".
4. Hiding the y-axis baseline → feels manipulative; use full range unless the distortion is material to your argument.

Exemplars

See examples/by_pattern.json → peer_gap. Canonical examples:

• Pershing Square · Canadian Pacific (Feb 2012) — the archetype. "CP has the worst operating ratio of any Class I railroad while closest peer CN has the best." 5 peers, single chart, the gap carries the entire deck.
• Starboard · Darden Restaurants (Sep 2014) — margin gap to casual-dining peers, opened at 300bps, closed campaign at parity under new board.
• Pershing Square · McDonald's (Nov 2005) — unusual inverted use: McDonald's trades at a discount to peers despite owning vastly more real estate.
• Starboard · Autodesk (Aug 2024) — operating-margin gap vs. Adobe and Intuit.

The slide itself

See slides/peer-gap-chart-recipe.md for exact layout, typography and colour decisions.