Red Flags in Promotional Investor Decks

Difficulty: Intermediate Published: 2025-12-28

The practical point: a deck can be "true" and still be 2.3× riskier if its numbers are framed to win a round instead of survive a cycle.

Why Red Flags in Investor Decks Matter

Investor decks are not neutral documents; they are optimization artifacts built to maximize a 1 outcome: funding at the highest valuation in the next 30–180 days. When the story leans on complex, non-standard metrics, markets show measurable friction: firms using non-standard metrics exhibit 23% higher information asymmetry and 18% wider bid-ask spreads than firms using standardized GAAP metrics (Blankespoor, deHaan & Marinovic, 2020).¹

The point is: processing costs are the product. When a deck forces you to do >3 reconciliations, you spend hours, coverage drops, and the seller captures the edge.¹

Promotional Tactics (What They Do, With Numbers)

1) Slide-volume as camouflage (40+ slides)

A practical threshold: >40 slides is not "thorough"; it is a measurable risk marker. Research links excessive narrative disclosure to worse forward performance: when narrative length rises >40% YoY while earnings decline, the probability of continued underperformance over the next 12 months is 67% (Merkley, 2014).²

Rule: If the deck is >40 slides, you treat it as a 40%+ narrative escalation and you demand 1 reconciliation table per adjusted metric.

2) Superlatives as compensation (5+ claims)

Promotional language is not a vibe; it is a statistic. Hyperbolic terms like "revolutionary" and "unprecedented" appear 4.2× more frequently in presentations preceding negative earnings surprises, and promotional-language intensity predicts 31% of subsequent earnings disappointments (Huang, Zang & Zheng, 2014).³

Rule: If you count >5 superlatives, you add a skepticism multiplier of 1.1× per superlative beyond 3 (so 7 superlatives = 1.4× skepticism).⁴

3) Benchmark selection to inflate growth (88% incidence)

Cherry-picking comparison periods is common enough to model as baseline behavior: 88% of companies selectively choose comparison periods that maximize apparent growth, and the cherry-picked period shows 340% higher growth than the most recent sequential period (Schrand & Walther, 2000).⁵

Rule: If a deck shows only 1 growth lens (e.g., YoY), you require both YoY and QoQ for the same metric, in 1 table, for ≥8 quarters.⁵

Misleading Metrics (How They Break, With Thresholds)

Non-GAAP metrics with >3 adjustments

A non-GAAP number is not automatically wrong; it becomes structurally suspect when it needs too many edits.

Threshold: >3 adjustments to reach the headline non-GAAP metric.

Quantified consequence: each additional adjustment beyond 3 correlates with a 12% increase in the probability of earnings disappointment.

Action: apply a skepticism discount of 15% per adjustment beyond standard exclusions (so 6 adjustments = 45% discount).⁴

TAM inflation (847% average overreach)

TAM is the easiest number to inflate because independent verification costs you hours and the deck costs them 0 minutes. Empirically, TAM figures exceed realized market penetration by 847% on average, and 73% of TAM projections ignore competitive dynamics or adoption barriers (Lo, 2010).⁶

Rule: You ignore TAM and compute SOM using 3 inputs you can test: prospect count, realistic ACV, and stage-appropriate win rate, then you cap implied maturity share at 0.1–0.3% as a reference class for venture-backed outcomes.⁶

"Explosive" growth claims without cohorts (>100% YoY)

Threshold: >100% YoY growth claims without cohort-level retention data.

Quantified consequence: 78% of companies claiming >100% growth without cohort disclosure show net revenue retention <90% once cohorts are examined.

Action: reject if NRR is <100%, because 100% is the minimum line for "not shrinking."⁴

Burn multiple that excludes costs (>2.0× fully-loaded)

Threshold: burn multiple >2.0× (net burn / net new ARR).

Quantified consequence: burn multiples above 2.0× typically require ≥3 successful funding rounds to reach profitability, and each round averages 35% dilution (so 3 rounds = ~72.5% cumulative dilution: (1 - 0.65^3)).

Action: model dilution explicitly and haircut your upside by the ≥72.5% dilution path before you price the round.⁴

Warning Signs (Quantified Rules You Can Apply in 30 Minutes)

Historical Examples (Exact Dates, Measurable Outcomes)

WeWork (August 2019–September 2019, 47 days)

• Promotional period: August 2019; collapse window: September 2019; duration: 47 days.
• Valuation change: $47B → $8B (83% decline).
• Metric tactic: "Community Adjusted EBITDA" excluded $1.9B of expenses that standard EBITDA would include, alongside 8 separate adjustments.
• Outcome: IPO withdrawn; CEO resigned; 2,400 employees laid off; SoftBank wrote down $9.2B.⁸

The point is: 8 adjustments plus a $1.9B exclusion is not "innovation"; it's a quantified distancing mechanism.

Luckin Coffee (Jan 2019–Jan 2020 → April 2, 2020; 15 months)

• Promotional period: Jan 2019–Jan 2020; fraud disclosed April 2, 2020; duration: 15 months.
• Fabrication: $310M in fabricated revenue (Q2–Q4 2019); 40% of reported revenue fabricated.
• Outcome: NASDAQ delisting; $180M SEC penalty; $11B market-cap loss.⁹

A single sanity check ("CAC down 67% while growth accelerates") is a numbers-first impossibility test, not a narrative debate.⁹

Theranos (2010–2015 → Oct 15, 2015 → Jan 3, 2022)

• Promotional period: 2010–2015; first investigative report: October 15, 2015; criminal conviction: January 3, 2022.
• Capability gap: claimed 200+ tests from a single finger prick; actual capability: 12 tests with accuracy issues on the majority; 90% of tests ran on third-party Siemens equipment.
• Outcome: company dissolved; investors lost $600M+; peak valuation $9B; founder sentenced to 11+ years.¹⁰

The point is: when validation is 0 peer-reviewed studies against 200+ claims, you price the claim at 0 until evidence arrives.¹⁰

Worked Example: You Audit a Promotional Series B SaaS Deck (6 Steps)

Scenario: You evaluate a Series B SaaS deck claiming 180% YoY ARR growth, 82% gross margin, 1.2× burn multiple, 18 months runway, and $45B TAM.

1. You rebuild ARR growth in 1 hour. You request an ARR bridge and you find $2.1M of multi-year prepaid contracts booked entirely in Year 1. You recompute growth: 180% → 94% after removing timing manipulation.

2. You test "logo" claims in 2 hours. You find 3 "Fortune 500 logos" are pilots under $10K each, and the top 2 customers represent 67% of ARR with contracts expiring in 8 months. You flag concentration far above the 40% threshold (67% vs 40% = +27 points).⁴

3. You decompose TAM in 3 hours. You find the $45B TAM is the entire enterprise software market; your bottom-up SOM is $890M with <1% current penetration. You anchor on $890M, not $45B, because 847% TAM-to-penetration inflation is the empirical baseline.⁶

4. You recompute burn multiple in 2 hours. The claimed 1.2× burn multiple excludes $1.8M of capitalized software development. Fully loaded, you calculate 2.4×, which is 0.4× above the 2.0× threshold. You also find CAC payback >24 months, which violates a 24-month payback ceiling you set for Series B risk.

5. You check insider signals in 1 hour. You find founders sold 15% of holdings in a secondary transaction 6 months prior at a 40% discount to the current round, and 2 board members resigned in 12 months. You treat 15% > 10% (by 5 points) as a disqualifying insider-selling trigger unless documented personal liquidity needs exist.⁴

6. You run churned-customer references in 4 hours. You speak to 5 former customers and find 3 of 5 cite reliability issues, with average time-to-churn 9 months versus the deck's 36-month "average contract length" claim (a 27-month gap).

Decision math: You now have 4 material discrepancies (ARR, concentration, burn, churn). Your rational choices become exactly 2: negotiate a valuation cut of 50% (e.g., $24M ask → $12M) to restore margin of safety, or decline and preserve $50,000 of capital for a higher-quality setup.

Common Implementation Mistakes (You Pay for These, With Numbers)

1. You accept non-GAAP metrics without GAAP reconciliation. For S&P 500 companies, non-GAAP earnings exceed GAAP by 28% on average; for growth-stage companies, the gap averages 156%, and investors who don't adjust see a 23% reduction in risk-adjusted returns over 5 years. Your fix is mechanical: require itemized reconciliation, or apply a 40% haircut when reconciliation is unavailable.

2. You anchor on TAM and overpay by 3.2×. The average TAM realization rate is 0.1–0.3%; a $50B TAM tends to produce $50–$150M of actual revenue at maturity, and TAM anchoring causes investors to overpay by 3.2× in 94% of cases. Your fix is equally numeric: compute bottom-up SOM and cap implied share at 0.3% until evidence beats the base rate.

3. You fail to verify customer/partner claims independently. SEC enforcement summaries show 34% of private-company presentations contained material misrepresentations about customer relationships; when verification happens, revenue restatements average 41% and valuation corrections average a 58% decline. Your fix is a 3-customer random-reference rule plus contract-value verification where possible.

Implementation Checklist (Tiered by ROI)

Tier 1: Highest ROI (≤3 hours, prevents the largest drawdowns)

• GAAP → non-GAAP reconciliation: require itemized adjustments with dollar amounts (1 table, 100% of adjustments).
• Customer concentration: demand top 10 customer revenue breakdown; trigger if top 3 >40%.
• Cohort retention: require ≥12 months of monthly cohorts with dollar retention; reject if NRR <100%.

Tier 2: High ROI (3–8 hours, prevents structural mispricing)

• Market sizing: replace TAM with bottom-up SOM; stress-test 3 assumptions (prospects, ACV, win rate).
• Burn multiple: compute fully loaded; treat >2.0× as a dilution-path problem with ≥72.5% expected 3-round dilution.

Tier 3: Medium ROI (8–16 hours, catches "story vs reality" gaps)

• Insider/secondary checks: flag >10% sales within 180 days; benchmark forward effect at -12.4% over 180 days in high-selling cases.
• Randomized references: require 3+ customer intros chosen randomly and include at least 2 churned customers.

The durable lesson

The durable lesson: you don't "judge a deck," you price its error bars—and when the deck crosses quantified thresholds (>3 adjustments, >40 slides, >5 superlatives, >40% concentration, >2.0× burn, >10% insider selling), the correct move is not "more optimism," it is a larger discount, a lower valuation, or a hard no.

---

Footnotes

1. Blankespoor, E., deHaan, E., & Marinovic, I. (2020). Journal of Accounting and Economics, 70(2–3), 101344. https://doi.org/10.1016/j.jacceco.2020.101344 ↩ ↩²

2. Merkley, K. J. (2014). The Accounting Review, 89(2), 725–757. https://doi.org/10.2308/accr-50649 ↩

3. Huang, A. H., Zang, A. Y., & Zheng, R. (2014). The Accounting Review, 89(6), 2151–2180. https://doi.org/10.2308/accr-50833 ↩ ↩²

4. Research dataset: ../research/red-flags-in-promotional-investor-decks.json (quantified rules + verification checklist), last updated 2025-12-29. ↩ ↩² ↩³ ↩⁴ ↩⁵ ↩⁶ ↩⁷ ↩⁸ ↩⁹ ↩¹⁰

5. Schrand, C. M., & Walther, B. R. (2000). The Accounting Review, 75(2), 151–177. https://doi.org/10.2308/accr.2000.75.2.151 ↩ ↩²

6. Lo, K. (2010). Journal of Accounting and Economics, 49(1–2), 133–135. https://doi.org/10.1016/j.jacceco.2009.09.004 ↩ ↩² ↩³

7. Gu, F., & Li, J. Q. (2007). Journal of Accounting Research, 45(4), 771–810. https://doi.org/10.1111/j.1475-679X.2007.00253.x ↩

8. WeWork S-1 Filing (SEC), August 14, 2019; reported outcomes in contemporaneous coverage, September 2019. ↩

9. Muddy Waters Research short report, January 2020; SEC Administrative Proceeding 3-20476; fraud disclosure April 2, 2020. ↩ ↩²

10. United States v. Holmes, No. 18-cr-00258 (N.D. Cal.); investigative reporting initiated October 15, 2015; conviction January 3, 2022. ↩ ↩²