Evaluating the Credibility of Promises, Threats, and Signals

Welcome back. In the last lesson, you built a rule for recurring relationships: cooperate constructively, respond to meaningful breaches with proportionate boundaries, and make repair possible when behavior genuinely changes. That rule depends on a difficult practical judgment: when someone says “I’ll do better,” “I won’t accept that,” or “we are fully committed,” should you treat it as information—or merely as talk?

This lesson gives you a practical way to assess promises, threats, and signals. The goal is not to become suspicious of everyone. It is to make your trust, concessions, and boundaries proportionate to the evidence available.

---

When words should change your choice

A statement is strategically important only when it changes what you reasonably expect another person to do or what you learn about their situation.

Consider three common forms:

• A promise is a claim about a future action: “I will send the work by Thursday.”
• A threat is a claim about a conditional future action: “If the payment is late, we will pause service.”
• A signal is an observable action meant to reveal something that is otherwise hard to see: “We have assigned two experienced people to support this launch.”

The central question is not, “Does this sound confident?” It is:

When the decisive moment arrives, will this person still have a reason and ability to do what they say?

That future moment matters because incentives can change. A supplier may sincerely promise rapid delivery while trying to win a contract. But after receiving full payment, expedited work may become costly and less urgent for them. Conversely, a teammate’s promise to complete a visible task can be reliable even without a formal contract if their reputation, workflow, and incentives all favor follow-through.

Sometimes a person makes a statement credible by deliberately limiting their own freedom to reverse course. For example, a manager may publicly commit to a deadline, allocate named staff, and publish a delivery plan. Backing out is still possible, but it now carries visible reputational and operational costs.

A commitment does not have to be dramatic or irreversible. In everyday settings, credibility may come from:

• a calendar invitation with accountable owners;
• payment in stages rather than entirely upfront;
• a public, specific commitment that colleagues can observe;
• a refundable deposit or other concrete stake;
• a written policy that constrains what someone is authorized to do;
• a record of reliable behavior in similar circumstances.

None of these is absolute proof. They are pieces of evidence that alter the likelihood that a claim will be honored.

---

A five-part credibility checklist

Use this checklist before giving up an option, accepting a promise, or changing your behavior in response to a threat. Think of it as a disciplined pause between hearing a claim and acting on it.

Check	What to examine	Useful evidence
1. Specificity	What exactly is being promised, threatened, or signaled? Under what condition and by when?	Named action, deadline, owner, and clear condition
2. Incentives	At the future moment of action, does following through still serve the speaker’s interests?	Ongoing relationship, reputation, mutual benefit, avoided cost
3. Control	Does the speaker actually have the authority, resources, and capacity to do it?	Decision rights, budget, time, access, relevant skills
4. Exposure	What does the speaker lose if they break the claim? Have they made reversal costly or difficult?	Deposit, contract, public commitment, operational setup, policy
5. Evidence	Can you verify the claim now or test it safely? What does their record show?	Past conduct, independent confirmation, observable preparation, small trial

A practical mnemonic is SICER: Specificity, Incentives, Control, Exposure, Record and evidence.

1. Specificity: can a neutral observer tell what happened?

“I’ll try harder” may be emotionally meaningful, but it is not a strategically strong promise. It leaves too much open: what counts as trying, how long, and what happens if circumstances become inconvenient?

Compare:

• “We will make the handoff process better.”
• “Before each Thursday release, the owning engineer will add the rollback procedure to the shared checklist by 2 p.m.; the release owner will confirm it.”

The second statement is easier to assess because it creates a visible test. Specificity does not itself make a claim truthful, but vagueness makes accountability nearly impossible.

2. Incentives: what will be best for them later?

Imagine a regular collaborator promises to give you feedback by Friday. Their incentive may be strong if your work blocks their own deliverable, if the relationship is ongoing, or if the commitment is visible to a shared manager. It may be weak if they gain nothing from the feedback and are overloaded with more urgent priorities.

This is the most important check:

Do not assess a promise by what the person wants now, while making the promise. Assess it by what they will want later, when keeping it becomes inconvenient.

The same logic applies to threats. If someone says, “If you decline this offer, I will never work with you again,” ask whether carrying out that threat would still make sense for them after you decline. If they need your expertise, have no replacement, and would lose heavily by cutting contact, the threat may be a bluff. If a policy, budget rule, or clear alternative makes withdrawal their best remaining option, it may be credible.

3. Control: can they actually deliver?

A sincere promise can still be unreliable. Someone may want to approve your proposal but lack the authority. A teammate may want to cover an urgent shift but lack the time. A sales representative may promise an exception that only a separate operations team can grant.

This is why “they seemed genuine” is incomplete evidence. Genuineness addresses intention; credibility requires both intention and capability.

4. Exposure: what makes backing out costly?

Credible commitments often work because the speaker has placed something at stake. The PBS reading’s entrepreneur is more convincing to investors when they risk their own money: someone who knows the venture is poor has less reason to take that risk.

In ordinary life, the stakes are usually smaller. A person may:

• reserve a meeting room and invite all participants;
• transfer a deposit;
• assign staff publicly;
• share a plan whose progress others can see;
• accept contractual or reputational consequences for non-delivery.

The key is not simply that an action is costly. A cost is informative when it would be less attractive for someone whose claim were false.

For example, an extended warranty can signal confidence in product quality because a firm selling a poor-quality product expects more expensive claims. But a flashy marketing campaign is not necessarily a quality signal: both good and poor firms may consider advertising worthwhile.

5. Evidence: what can you see, verify, or test?

A good record of similar behavior is evidence, particularly in recurring relationships. So are preparations that are hard to fake: a completed draft, an assigned owner, a paid deposit, or access already provisioned.

Still, avoid treating history as destiny. A previously reliable colleague may face a new constraint; an unreliable counterpart may be changing under a redesigned process. The strongest response is often neither blind trust nor permanent distrust, but a small, observable test.

---

Promises, threats, and signals require different tests

The same checklist applies to all three, but the decisive question changes slightly.

Promises: “Will they follow through?”

A promise should influence you when the action is specific, feasible, and aligned with the other person’s future incentives—or when they have made breaking it meaningfully costly.

Suppose another team says:

“We will provide tested rollback documentation before the next release.”

A reasonable response is not to label this either “trusted” or “untrusted.” Instead, assess it:

• Specificity: Is “before the next release” a real deadline, and is the documentation standard defined?
• Incentives: Does reliable documentation reduce their own incident risk, or only shift work away from them?
• Control: Is there a named engineer with time and access to write it?
• Exposure: Is the commitment visible in the release plan? Is there a review gate that cannot be bypassed casually?
• Evidence: Have they completed comparable handoffs? Can you inspect a draft early?

If evidence is mixed, do not necessarily reject the promise. Reduce your exposure: ask for a draft, retain a review gate, or divide the work into smaller milestones.

This continues the previous lesson’s repair logic. An apology after an incomplete handoff becomes more credible when it includes ownership, correction of the immediate problem, and an observable future commitment—not merely reassurance.

Threats: “Would they really carry this out?”

A threat is credible if the threatened action is one the speaker can and will take if the condition occurs. The relevant test is not whether the threat is unpleasant or forcefully delivered. It is whether executing it remains a sensible choice for the speaker.

For example:

“Without a security review, this service cannot be released.”

This may be credible if the speaker controls a real release gate, the policy is established, and releasing without review would expose them to significant responsibility. It is less credible if they lack authority, routinely waive the rule, or would be harmed more by delay than you would.

A credible threat can still be unreasonable, manipulative, or inappropriate. Credibility tells you what may happen; it does not tell you what you ought to accept. You may respond by proposing an alternative, documenting the issue, escalating through an appropriate process, or declining the arrangement.

Threats involving coercion, harassment, abuse, blackmail, serious safety risks, or major power imbalances are not ordinary bargaining tools. In such situations, prioritize safety, preserve relevant evidence where safe, and seek appropriate organizational, legal, or personal support rather than trying to “play the game” better.

Signals: “What does this action reveal?”

A signal is often more persuasive than a statement because it is observable and costly, difficult, or unattractive to fake.

Imagine a prospective contractor says they are committed to a long-term relationship. Their words are weak evidence by themselves. Their behavior may be more informative if they invest time in understanding your workflow, provide a carefully scoped pilot, accept staged payment linked to outcomes, and assign a senior person to the account.

Yet no single action proves quality or goodwill. A contractor may invest heavily because winning your contract is valuable even if their eventual delivery will be poor. Ask: Would a low-quality or uncommitted counterpart have a similar reason to take this action? If yes, the signal is weaker than it first appears.

---

Turn an assessment into a proportionate response

Credibility is not a binary label. Your response should reflect both the strength of the evidence and the cost of being wrong.

Assessment	Appropriate response
Weak or unclear	Do not make an irreversible concession. Ask for detail, seek independent confirmation, or use a small reversible test.
Mixed	Proceed in stages. Share limited information, set a checkpoint, document expectations, or use reciprocal commitments.
Strong	Let the claim meaningfully shape your plan, while still keeping normal safeguards appropriate to the stakes.

Consider a realistic cross-team situation. A partner manager says:

“We will provide twenty-four-hour support during launch. If the integration is not approved by Friday, we will reassign those engineers. We have already named the on-call rotation in the shared plan.”

There are three strategic claims here:

1. Promise: support will be available during launch.
2. Threat: support will be withdrawn if approval does not occur by Friday.
3. Signal: named engineers and a shared rotation demonstrate some preparation.

A measured response might be:

• Verify whether the on-call rotation is real and whether the manager can actually reassign those engineers.
• Ask what operational constraint creates the Friday deadline: a budget cycle, another client commitment, or merely negotiation pressure.
• Clarify what “approval” requires and which issues could block it.
• Offer a staged plan: approve the low-risk integration component now, complete the outstanding checks by an agreed date, and retain support for the defined launch scope.

This response does not assume bad faith. It also does not surrender to urgency merely because it was announced. You let the evidence determine how much the statement influences you.

A useful two-minute habit for a low-stakes interaction is to write down four lines before responding:

1. The claim: What exactly was said or done?
2. The future moment: When would the person have to follow through?
3. The evidence: Which parts of SICER are genuinely present?
4. My response: What is the smallest sensible commitment, request for verification, or boundary?

That short pause prevents two predictable errors: dismissing a reliable commitment because you dislike the speaker, and accepting an unreliable claim because it sounds confident or reassuring.

---

Key takeaways

Promises, threats, and signals should influence your choices only to the extent that they give you credible information about future behavior or hidden intentions.

• A promise is credible when follow-through remains worthwhile and feasible.
• A threat is credible when the speaker can and would carry it out if the condition occurs.
• A signal is informative when it is difficult or unattractive for someone without the claimed quality or intention to imitate.
• Use SICER: specificity, incentives, control, exposure, and record/evidence.
• Calibrate your response. When evidence is uncertain, use verification, staged commitments, and reversible tests rather than all-or-nothing trust.

In the next module, you will move from credibility to negotiation: assessing each side’s fallback if no agreement is reached.