Model Best, Expected, and Downside Negotiation Scenarios

Scenario modeling negotiation planning should prepare a team to act when conditions change—not merely produce three forecasts. In practical Negotiation Scenario Planning, every best, expected, and downside case needs a traceable chain: evidence, assumptions, modeled outcome, observable triggers, response options, and an accountable decision-maker.

This structure turns Offer Response Scenarios into operational choices. It helps procurement, finance, legal, operations, and business sponsors agree in advance on what the negotiation lead may offer, when the team must escalate, and which evidence would justify changing the mandate.

Quick answer

Teams should connect scenarios through one chain: evidence → assumptions → modeled outcome → observable trigger → pre-approved response → accountable decision-maker. Build best, expected, and downside cases around coherent conditions, not arbitrary percentage changes. For each case, document evidence quality, commercial outcomes, trigger thresholds, response options, approval authority, and the date or event that requires reassessment.

What the three scenarios mean

A scenario is a coherent set of conditions and assumptions used to examine a plausible outcome. It is not simply an optimistic, central, or pessimistic number. The HM Treasury Green Book describes scenario analysis as examining plausible futures and evaluating a proposal under those different conditions. It also cautions against assigning probabilities that imply more precision than weak evidence supports.

For an enterprise negotiation, define the three cases as follows:

• Best case: A favorable but supportable combination of conditions. It is not the team's aspiration or opening anchor.
• Expected case: The planning case most strongly supported by current evidence. “Expected” does not necessarily mean statistically most likely.
• Downside case: A plausible adverse combination that could materially change price, service, timing, risk allocation, or leverage.

The three cases should model the complete package. In a procurement negotiation, price may interact with volume, payment timing, implementation milestones, service levels, liability, term, indexation, and termination rights. A price-only model can obscure trades that create or protect value.

The EATRA three-scenario model

Use the EATRA model to make each scenario reusable and reviewable:

1. Evidence: What verified information supports the scenario?
2. Assumptions: What must be true for the scenario to hold?
3. Terms: What commercial package results?
4. Response trigger: What observable threshold changes the team's action?
5. Authority: Who may activate the response, and who must approve an exception?

This model supports broader Negotiation planning by connecting analysis to a mandate. It also prevents a common failure: writing a downside narrative without deciding what anyone will do if that downside begins to materialize.

Reusable EATRA template

Each row should have an owner. “Procurement” or “the business” is not specific enough; name a role or individual accountable for validation and approval.

Separate facts, assumptions, estimates, and recommendations

A scenario workbook becomes unreliable when these evidence classes are blended together. Label them explicitly.

Verified facts

Facts are confirmed by reliable evidence, such as signed contract terms, actual consumption history, validated bids, audited supplier statements, confirmed production capacity, or a published index.

Record the source, measurement date, owner, and next refresh date. A six-month-old capacity statement may remain a historical fact while no longer supporting a current planning conclusion.

Assumptions

Assumptions are propositions required to complete the model but not yet established as facts. Examples include:

• a supplier will preserve capacity through a specified date;
• forecast demand will stay within a defined range;
• a second source will qualify by a target milestone;
• an index approximates the supplier's actual input exposure;
• the counterparty will accept a proposed risk-sharing mechanism.

The GAO Cost Estimating and Assessment Guide emphasizes documented, traceable assumptions and sensitivity analysis. It warns that subjective plus-or-minus percentages are not a valid substitute for examining the variables that drive an estimate.

For every material assumption, document an owner, confidence assessment, contrary evidence, and validation action.

Estimates

Estimates are calculated values rather than verified future facts. Examples include a should-cost range, demand forecast, transition cost, concession value, switching timeline, or supplier-margin estimate.

Show the method, input dates, range, and sensitivity. Avoid displaying a point estimate with false precision. If several inputs may move together, test a coherent scenario rather than changing one variable while holding implausibly related variables constant.

Recommendations

Recommendations are management judgments: the opening package, concession sequence, trigger threshold, reservation position, or decision to pause. Every material recommendation should state who approved it and what new evidence would cause reconsideration.

This separation is especially important when AI assists with synthesis. A generated summary should never silently convert an assumption into a fact or an estimate into an approved position.

A six-step Negotiation Scenario Planning process

1. Define the decision before gathering data

Write the decision the scenarios must support. For example: “What package can the negotiation lead offer if the supplier substantiates a material-cost increase but cannot demonstrate a capacity constraint?”

Without a decision statement, teams often gather large amounts of market data that do not change their action.

2. Model the package and the BATNA

Define the issues that create total value or exposure: price, volume, scope, timing, payment, service, risk allocation, implementation, and term. Then test the BATNA—the operational course available if no agreement is reached.

A BATNA should be costed, time-bounded, funded, and assigned to owners. “We can switch suppliers” is not a credible BATNA if qualification requires an uncertain nine-month program that nobody has approved.

Use the BATNA to inform the reservation position, but do not treat the two as identical. The reservation position is the least favorable package the team is authorized to accept; the BATNA is what the organization can actually do without agreement.

3. Identify high-impact assumptions

List the assumptions that most affect the package or BATNA. Typical inputs include supplier capacity, market participation, input costs, demand, foreign exchange, implementation timing, switching costs, and internal approval ceilings.

Run sensitivity analysis to identify which assumptions matter most. Then combine relevant assumptions into best, expected, and downside cases. For more detail on structuring the model itself, see negotiation scenario modeling.

4. Write executable triggers

A trigger must change an action, not merely describe the environment. Use this syntax:

If a named indicator crosses a defined threshold by a stated date, then the negotiation lead may take a specified action, subject to named evidence checks and approval.

Weak trigger: “If costs rise, revisit pricing.”

Stronger trigger: “If the agreed input index rises above the approved threshold before final offer, validate the observation period and the supplier's actual exposure, then consider indexation, a volume adjustment, or shorter offer validity. Any package exceeding the reservation position requires finance and commercial-sponsor approval.”

The threshold does not have to be numeric. A qualified bidder withdrawing, a missed certification milestone, or a material contractual exception can also be observable triggers.

5. Design Offer Response Scenarios

For each likely supplier move, prepare more than a spoken answer. Define:

• the evidence the team will request;
• the questions used to test the claim;
• the issues available to trade;
• the concessions prohibited without approval;
• the response package within the lead's mandate;
• the point at which the team pauses or escalates.

Offer Response Scenarios should remain conditional. A concession given in exchange for a verified commitment is different from an unconditional movement that merely rewards the counterparty's opening demand.

6. Approve, monitor, and refresh

The model should be approved before material negotiations begin and refreshed after events such as a revised bid, supplier disclosure, scope change, missed milestone, or significant market movement.

The U.S. Federal Acquisition Regulation's acquisition-planning requirements call for attention to milestones, constraints, alternatives, and cost assumptions. The UK Sourcing Playbook similarly supports pre-agreed positions, defined roles, early risk identification, and iterative validation of assumptions.

Maintain an audit trail showing what changed, why the scenario changed, and who authorized the revised response.

Hypothetical procurement example: packaging renewal

The following example is hypothetical. Its figures and events illustrate the method and are not market evidence or benchmarks.

A manufacturer is renewing a regional packaging agreement. The supplier requests a price increase and cites higher material costs and limited production capacity. Procurement has one incumbent, one qualified alternative, and a third supplier still undergoing quality approval.

Best scenario

• Evidence: The qualified alternative confirms bid validity and available capacity; the relevant published input index declines.
• Assumptions: Competition remains active and the alternative can meet the implementation date.
• Modeled package: Current price is preserved, service reporting improves, and an optional volume tier is added.
• Trigger: The alternative reconfirms capacity in writing and passes the scheduled operational review.
• Response: Ask the incumbent to improve price and service terms in exchange for a defined volume commitment.
• Authority: Procurement may trade within the approved volume and term limits; operations must approve any allocation change.

Expected scenario

• Evidence: The supplier demonstrates partial cost exposure, but the claimed increase is not fully aligned with the selected index.
• Assumptions: Demand remains within the internal forecast band and both qualified suppliers remain available.
• Modeled package: A limited adjustment is paired with index transparency, a cap, and reciprocal downward movement.
• Trigger: Validated evidence demonstrates exposure during the contract's defined measurement window.
• Response: Use a contingent package rather than accepting a permanent unqualified increase.
• Authority: The negotiation lead may agree within the approved range; finance validates the model.

Downside scenario

• Evidence: The alternative misses a qualification milestone and the incumbent provides documented notice of a capacity constraint.
• Assumptions: Switching will take longer than planned, and a service interruption would impose material operational cost.
• Modeled package: Shorter term, reduced optional scope, allocation protections, and a time-bounded adjustment rather than a broad long-term commitment.
• Trigger: The qualification delay is confirmed by the accountable quality owner and capacity evidence is validated.
• Response: Repackage the deal, preserve termination or reopener options, and escalate any package beyond the reservation position.
• Authority: Business sponsor, operations, finance, and legal review the revised risk allocation.

This example also illustrates the ZOPA. In the best and expected scenarios, overlap may exist across several issues even if price positions initially differ. In the downside case, the ZOPA may narrow or disappear. The model should reveal that possibility, not force a deal where no acceptable overlap exists.

Multi-party governance: assign authority before the meeting

In an enterprise negotiation, stakeholder disagreement can be as consequential as counterparty resistance. Create a compact decision-rights table:

Accountable human review remains mandatory for deciding whether the downside case is plausible, setting the reservation position, determining whether a BATNA is operationally credible, interpreting conflicting supplier data, approving out-of-mandate concessions, and accepting residual financial, operational, legal, or reputational risk.

Teams should also define what happens when an approver is unavailable. An escalation path with no decision deadline can become a hidden concession to the counterparty.

How AI can support the workflow without owning the decision

AI can help organize evidence, compare scenario versions, identify missing fields, generate questions against weak assumptions, and draft Offer Response Scenarios for review. In a concrete Negotiations.AI procurement workflow, a team might use Negotiations.AI to assemble source-linked assumptions from a supplier proposal, market files, and internal demand data, then route the resulting scenario brief to finance, operations, legal, and the business owner for validation.

The human team must still verify sources, correct classifications, approve thresholds, and own every external commitment. AI output should be treated as decision support—not automatic authority to change an offer or accept a package.

Useful review prompts include:

• Which statements are verified facts, and which are assumptions or estimates?
• Which source and date support each material input?
• What evidence contradicts the expected case?
• Which two assumptions have the greatest effect on the reservation position?
• Which triggers lack an owner, threshold, or response?
• Which proposed responses exceed the negotiation lead's mandate?

For a broader preparation workflow, see how to build a negotiation brief from fragmented evidence.

Scenario quality scorecard

Score each item as ready, needs revision, or missing:

• The best case is favorable but credible, not merely aspirational.
• The expected case is supported by current evidence.
• The downside case is plausible and materially different.
• Facts, assumptions, estimates, and recommendations are labeled separately.
• Each material input has a source, date, owner, and refresh point.
• The model covers the commercial package, not only price.
• BATNA timing, cost, funding, and ownership are documented.
• ZOPA implications are treated as hypotheses to test, not known facts.
• Every trigger is observable and changes the team's action.
• Every response has evidence requirements and an approval owner.
• Reservation positions and prohibited concessions are explicit.
• Cross-functional reviewers have approved the mandate.
• Material changes create a new version and audit trail.

If a scenario is missing a trigger, response, or decision-maker, it is analysis—not an operating plan.

Common mistakes to avoid

Applying arbitrary percentages

“Best is 5% lower and downside is 10% higher” does not explain what changed or what the team should do. Model the drivers and their interactions instead.

Treating probabilities as facts

Do not assign a 60% probability because the expected case feels plausible. Use probabilities only when a defensible empirical basis and suitable method exist. Otherwise, compare plausible scenarios without creating false precision.

Using lagging triggers

A quarterly index or annual filing may confirm a problem too late. Where possible, pair lagging data with leading indicators such as bid withdrawals, lead-time changes, order-book disclosures, or missed qualification milestones.

Confusing a target with a best case

A target is a recommendation. A best case is a modeled outcome supported by favorable conditions. Keeping those concepts separate improves discipline.

Ignoring correlated variables

Capacity, lead time, logistics costs, and allocation terms may move together. One-variable sensitivity analysis can mislead when conditions are connected.

Treating the model as a substitute for judgment

A scenario cannot decide legal permissibility, organizational risk appetite, or whether a strategic relationship outweighs a short-term economic gain.

Limits and when this approach does not fit

Three scenarios cannot represent every possible future, and best and downside cases are not automatically statistical confidence bounds. Market indexes may not reflect a supplier's actual purchasing basket, hedge position, or timing. Public filings may be consolidated and reveal little about contract-level economics. A stated alternative may also prove slower or costlier than modeled.

This framework may be disproportionate for low-value, reversible transactions with minimal operational risk. It also needs adaptation where tender rules restrict negotiation, emergency conditions leave no meaningful response time, or legal and regulatory requirements prescribe decision authority. In those situations, follow the applicable process and obtain appropriate specialist review.

Use the model to expose uncertainty and prepare choices—not to disguise uncertainty behind a spreadsheet.

FAQ

How many scenarios should a negotiation plan include?

Three is a practical minimum: best, expected, and downside. Add a separate scenario only when it represents a materially different condition, response, or approval path. More scenarios are not automatically better.

How are Offer Response Scenarios different from forecasts?

A forecast describes what may happen. Offer Response Scenarios specify what the team will ask, offer, withhold, validate, or escalate when a defined supplier move or external trigger occurs.

Should every trigger have a numeric threshold?

No. A trigger must be observable and decision-relevant. It can be a numeric threshold, a missed milestone, withdrawal of a bidder, receipt of verified documentation, or an unresolved contract exception.

How often should Negotiation Scenario Planning be refreshed?

Refresh it after any material bid, disclosure, scope change, market movement, approval change, or BATNA milestone. Also set a scheduled review date so stale assumptions do not survive by default.

Can AI approve a scenario or change the reservation position?

No. AI may support classification, synthesis, comparison, and drafting, but accountable people must validate evidence, approve the mandate, authorize exceptions, and accept residual risk.

Further reading

• HM Treasury: The Green Book 2026
• U.S. GAO: Cost Estimating and Assessment Guide
• U.S. Federal Acquisition Regulation: Part 7—Acquisition Planning
• UK Cabinet Office: The Sourcing Playbook
• UK Cabinet Office: Should Cost Modelling Development Guidance

Disclaimer: This article provides general negotiation-planning information and is not legal, financial, or procurement-regulatory advice.