M&A due diligence has a systematic analytical failure that the deal-making process tends to amplify rather than correct. At Rung 1, the three teams who disagree at the board table — finance on valuation, strategy on fit, risk on customer concentration — are all partially observing the same upstream pressure: competitive deal dynamics and strategic urgency that simultaneously inflate every metric the board is looking at. At Rung 2, earnout structures are offered to targets with revenue uncertainty — which means observed integration outcomes for earnout deals already started from a higher-uncertainty position than cash-at-close deals. At Rung 3, management retention clauses are resisted by strong management teams, meaning the teams most likely to deliver integration are the ones the clause was designed to retain. The intervention query isolates each of these structural confounds from the observational comparison.

QuestionStandard ApproachCausal Approach
Which board concern is structural?Review valuation, fit, and concentration independentlyDiagnostic model traces all three back to Deal Competitive Pressure and Strategic Urgency confounders
Earnout effect on integrationCompare earnout vs cash-at-close deal outcomesIntervention query severs Revenue Uncertainty confounder; isolates incentive mechanism from deal selection
Did the dropped clause cost us?Compare integration to deal model targetAbduction anchors Management Quality and execution environment; counterfactual restores the clause
The deals that get approved despite risk flags are not evidence that the risk flags were wrong. They are evidence about the board’s risk tolerance — which is also a predictor of which credits they approved.
3 Questions, 3 Rungs
  1. Would insisting on the management retention clause have changed integration performance on a deal that is now 35% below target? — Rung 3 (Counterfactual). Answering it requires abduction to anchor Management Quality and the actual adverse execution environment before restoring the contractual mechanism; the confounder is Management Quality, which must be fixed at its observed value before the intervention is applied.
  2. What does requiring an earnout actually cause to management retention and integration speed, separate from the fact that earnouts are only offered when revenue is uncertain? — Rung 2 (Intervention). A do() query severs the Revenue Uncertainty and Target Management Motivation confounders, isolating the causal effect of the earnout structure from the conditions that make it more likely to be offered.
  3. Which of the three competing board concerns — valuation multiple, strategic fit, or customer concentration — is the structural driver versus an artifact of deal competitive dynamics? — Rung 1 (Association). The graph encodes which dependencies exist between Deal Competitive Pressure, Strategic Urgency, and integration outcomes; entering each team’s evidence updates only the genuinely connected nodes.

Reading the screenshots: a black check mark on a node means it has been set as observed evidence — a fact entered into the model, acting as a filter. A red check mark means it has been set as a do intervention — a decision applied to the model, severing the influence of its parents.

Reading the spec tables: each Run the Analysis block lists the exact steps to reproduce each screenshot in Bayes Server. The Obs / Do column uses three italic control tokens: clear — reset the model to a blank no-evidence state; abduction step — enter the factual observations that anchor the U nodes to this specific case; use abduction result — apply a do() intervention with the U nodes held from the abduction step.

Rung 3 — Counterfactual

Would insisting on the management retention clause have changed integration performance?

“We acquired a target 18 months ago. Integration has underperformed the deal model by 35%. During negotiation we dropped a management retention clause when the seller’s team resisted it. Would keeping it have changed the outcome?”

Management Quality is the confounder: strong managers resist retention clauses because they have leverage, and they also deliver better integration regardless of contractual structure. Observing that the clause was dropped tells you the management team was capable enough to resist — which is a positive signal for underlying integration capability. Abduction first anchors the unobserved background risk to this deal’s actual execution context (the below-target outcome tells us the execution environment was adverse). Then restoring the clause improves Personnel Continuity through the contractual pathway.

Answer

The model shows that the retention clause would have improved Personnel Continuity and through it Integration Performance — but not enough to bring the deal to target, because the Management Quality direct path to Integration Performance independently explains a significant portion of the 35% underperformance. The clause was not the deal’s primary failure point. The counterfactual answer: insisting on the clause would have moved the needle, but the board’s fundamental miscalculation was about management quality rather than contract structure.

MADueDiligenceCausal.bayes
ImageObs / DoNodeSetResult
ma-cf-0Management Quality30% High / 48% Moderate / 22% Low
Retention Clause54.3% Included / 45.7% Dropped
Key Personnel Continuity37.9% Full / 35.2% Partial / 26.8% Departed
Integration Performance22.7% Above / 35.7% On Target / 41.6% Below
Deal Value Realisation25.0% Above / 34.5% At / 40.6% Below Model
ma-cf-1obsRetention ClauseDroppedObservable from signed agreement
obsIntegration PerformanceBelow TargetU nodes update — deal background anchored
Management Quality23.7% High / 58.1% Moderate / 18.2% Low (inferred)
Key Personnel Continuity8.4% Full / 32.6% Partial / 59.0% Departed
ma-cf-2doRetention ClauseIncludedSevers Management Quality → Clause back-door
Management Quality13.4% High / 51.5% Moderate / 35.2% Low (updated)
Key Personnel Continuity41.0% Full / 38.7% Partial / 20.2% Departed (improved)
Integration Performance100% Below Target — still clamped from abduction step
Deal Value Realisation4.0% Above / 22.0% At / 74.0% Below Model
Prior state — Counterfactual model, no evidence set
Prior — no evidence set

Before any evidence is entered, the model reflects the population prior for M&A integration outcomes across deals of this type and size.

Rung 2 — Intervention

What does an earnout structure actually cause to management retention and integration speed?

“If we require an earnout tied to post-close EBITDA rather than clean cash-at-close, what does that actually cause to management retention and integration speed — separate from the fact that we only offer earnouts to targets where we have revenue uncertainty?”

Revenue Uncertainty Level and Target Management Motivation are the confounders. Earnouts are offered specifically when acquirers have high revenue uncertainty — so observed earnout deals already started from a more uncertain revenue position. Highly motivated management teams resist earnouts because they want certainty, and they also independently deliver better integration. Observing an earnout deal tells you both that revenue was uncertain and that management was not strong enough to resist — both of which negatively predict integration outcomes. The intervention query holds both confounders at their priors, isolating just the incentive mechanism.

Answer

The pure causal effect of earnout structures on management retention is negative at high revenue uncertainty — earnout disputes over target definitions become the primary reason senior management exits early. However, the intervention query at moderate uncertainty shows a positive effect on integration speed through incentive alignment. The board is not deciding whether earnouts are good or bad in general — it is deciding whether to impose an earnout on this specific target under its specific uncertainty and management motivation profile. The model separates those conditions.

MADueDiligenceCausal.bayes
ImageObs / DoNodeSetResult
ma-int-0Revenue Uncertainty Level30% High / 48% Moderate / 22% Low
Target Management Motivation35% High / 45% Moderate / 20% Low
Deal Structure62.9% Earnout / 37.1% Cash-at-Close
Management Retention58.9% Full / 29.1% Partial / 12.0% Departed
Integration Speed39.2% Fast / 35.9% On Track / 24.9% Delayed
ma-int-1doDeal StructureEarnoutSevers both confounder back-doors
Revenue Uncertainty Level30% High / 48% Moderate / 22% Low — stays at prior
Target Management Motivation35% High / 45% Moderate / 20% Low — stays at prior
Management Retention53.6% Full / 31.6% Partial / 14.8% Departed — true causal effect
Integration Speed37.2% Fast / 36.1% On Track / 26.7% Delayed
ma-int-2obsDeal StructureEarnoutBack-door open — confounded estimate
Revenue Uncertainty Level24.0% High / 48.4% Moderate / 27.5% Low — updates (selection)
Target Management Motivation30.2% High / 46.0% Moderate / 23.8% Low — infers lower
Management Retention55.6% Full / 30.7% Partial / 13.7% Departed
Integration Speed37.1% Fast / 36.2% On Track / 26.7% Delayed
Prior state — Intervention model, Management Retention
Prior — no evidence set

At the prior, the population-average 12-month management retention rate reflects the full mix of deal structures across the acquirer’s historical portfolio.

Rung 1 — Association with Causal Structure

What does the diligence data actually support as primary value driver and primary risk?

“The board is reviewing a £400M target. Finance says the multiple is too high. Strategy says the strategic fit justifies it. Risk says the customer concentration is the real issue. What does the diligence data actually support?”

Deal Competitive Pressure and Strategic Urgency are the confounders. Contested auctions simultaneously inflate the multiple, compress diligence time, and suppress risk focus — so all three teams’ concerns are partially downstream of the same deal dynamics. Strategic urgency inflates fit perceptions through motivated reasoning and simultaneously reduces appetite to act on concentration risk. At Rung 1, the graph encodes which dependencies exist: confirming a high valuation multiple updates Deal Competitive Pressure, which also shifts customer concentration risk through the same shared confounder.

Answer

When only the deal approval decision is entered as evidence, all three concerns update — no single issue dominates. Adding competitive auction evidence shifts valuation multiple and concentration simultaneously through the shared confounder, showing that finance and risk are both partially right about the same underlying problem. The diagnostic result: the board is not choosing between three independent concerns but managing one compound risk from deal dynamics that manifests across all three dimensions at once.

MADueDiligenceCausal.bayes
ImageObs / DoNodeSetResult
ma-diag-0Competitive Auction40% Yes / 60% No
Strategic Urgency25% High / 50% Moderate / 25% Low
Deal Approval76.1% Approved / 23.9% Rejected
ma-diag-1obsDeal ApprovalApproved
Deal Competitive Pressure32.8% High / 40.7% Moderate / 26.5% Low
Strategic Urgency23.5% High / 49.4% Moderate / 27.1% Low
Valuation Multiple34.5% High / 41.0% Fair / 24.5% Low
Customer Concentration30.8% High / 37.0% Moderate / 32.2% Low
ma-diag-2obsCompetitive AuctionYes
Deal Competitive Pressure65.0% High — amplifies from 32.8%
Valuation Multiple49.6% High — co-moves via DCP
Customer Concentration45.9% High — co-moves via DCP
ma-diag-3obsStrategic UrgencyHigh
Strategic Fit Score62% Strong — inflated by urgency
Risk Focus10% Adequate / 45% Reduced / 45% Minimal — suppressed
Prior state — Diagnostic model, Deal Assessment
Prior — no evidence set

Before any evidence is entered, the three board concerns — valuation, strategic fit, and customer concentration — are independent and weighted roughly by their historical base rates in deals of this sector and size.

MADueDiligenceCausal.bayes — All Three Rungs
25-node causal model covering the full M&A due diligence lifecycle. Rung 1: enter obs(Deal Approval = Approved) then add obs(Competitive Auction = Yes) and obs(Strategic Urgency = High) — read how Deal Competitive Pressure, Valuation Multiple, Customer Concentration, and Risk Focus update through shared confounders. Rung 2: compare do(Deal Structure = Earnout) vs obs(Earnout) — Revenue Uncertainty Level and Target Management Motivation stay at prior under do(), revealing the selection bias in observed earnout performance. Rung 3: abduct with obs(Retention Clause = Dropped) + obs(Integration Performance = Below Target), then do(Clause = Included) — Key Personnel Continuity improves through the contractual mechanism; Management Quality’s direct path to Integration Performance shows the clause alone cannot close the full gap.

All models require Bayes Server (free edition available). See Download Models for the full library.

Next Step

If your deal committee is weighing three board concerns that all intensified during the competitive process, the diligence data cannot tell you whether those concerns are independent risks or the same upstream problem wearing three different faces.

The models are free. What I provide is the judgment to build the right structure for your specific situation, encode your experts’ knowledge into it, and turn the output into decisions your board can act on. The discipline stays with your team.

info@rung3.ai

This case study is a composite drawn from published M&A research, integration practice literature, and corporate finance case studies. Specific figures are representative. No individual organization or engagement is described. The Bayes Server models are working files: download, set evidence, and run inference.