Separate pricing, timeline, feature availability, service level, and commercial policy. Different claims decay at different speeds.
// TRUST DENSITY · GTM CONTROL PLANE · June 5, 2026
Trust Density Needs Promise Decay Clocks Before Agents Keep Selling Old Truth
A machine does not need to hallucinate to damage trust. It only needs to keep repeating a promise that expired yesterday.
Stale certainty is one of the fastest ways for autonomous companies to look fake.
Most teams look for trust failures in the obvious places: bad personalization, sloppy outreach, broken escalation, or full-blown hallucination. But the more common trust leak is simpler. A buyer-facing agent says the onboarding timeline is still fourteen days. Another message repeats last week's discount logic. A support-side bot references a roadmap commitment that product quietly pulled back. None of those statements are invented. They are just old.
That is exactly why they are dangerous. Old truth lands with the confidence of current truth. The buyer does not see a lagging database. They see a company making promises it cannot coordinate internally.
A promise decay clock is the rule that says any buyer-facing claim expires after a defined window unless a named owner refreshes it. If the clock runs out, the agent loses the right to keep repeating that claim as fact.
SignalPricing drift
An agent keeps quoting a “temporary” concession after finance already tightened the floor and sales managers are now clawing back expectations live on calls.
SignalRoadmap drift
The machine keeps treating a paused feature as imminent because the underlying note was once true and nobody attached an expiration rule to it.
SignalDelivery drift
Implementation or support capacity changes, but the agent still sells the old response time because nobody updated the promise owner and freshness window.
Operator rule: if a promise can reach a buyer, it needs an owner, an expiry window, and a shutdown path before a machine gets to repeat it at scale.
Promise decay clock
Four fields every buyer-facing machine claim needs before it earns reuse
Some claims should die in 24 hours, others in 7 days, others the moment an upstream system changes. Pick the real clock, not the convenient one.
If nobody owns refresh, the agent will keep recycling stale confidence until a human gets embarrassed in front of the buyer.
Expired claims should fall back to guarded language, escalation, or silence — not keep shipping confident garbage because the template still exists.
Three grounded examples where promise decay clocks protect trust before revenue teams feel the pain
Discount logic moved
A growth agent keeps offering “usual flexibility” to revive late-stage deals, but finance already tightened discount approvals after margin slipped. Without a decay clock, the agent keeps reopening commercial expectations the humans now have to unwind. With one, the pricing promise expires daily and must be refreshed from the current policy owner.
Roadmap language fossilized
An expansion bot keeps nudging accounts with “coming this quarter” messaging because it inherited a roadmap answer from two weeks ago. Product reprioritized. The message is now commercially toxic. A roadmap claim should decay fast, and once expired the agent should switch to proof-backed present capabilities only.
Capacity already changed
A booking or onboarding agent still tells buyers they can be live in ten days, but services capacity is now stretched to three weeks. The machine is not lying on purpose. It is replaying stale truth. Promise decay clocks force the implementation estimate to refresh against actual delivery capacity before the agent uses it again.
What heads of growth should actually govern
- Stop shipping evergreen claims. If a statement affects buyer trust, assume it decays unless proven otherwise.
- Classify promises by blast radius. Pricing, timelines, and roadmap claims deserve much shorter freshness windows than generic brand copy.
- Bind reuse to source refresh. Agents should not inherit yesterday's truth just because it still lives in a prompt or CRM field.
- Instrument expired-claim fallbacks. “I need to verify the latest” is healthier than machine confidence built on old context.
- Treat stale promises as control-plane failures. This is governance, not copy cleanup.
The market will forgive a company for checking. It will not forgive a company that keeps speaking with expired certainty.
Concrete takeaway for this week
Pick the top twenty buyer-facing claims your agents can currently make across outbound, qualification, pricing, onboarding, and support. For each one, record three things: who owns truth, how long that truth remains valid, and what the agent must do when the window expires.
Then run a hard test: kill refresh on purpose for one day and see which claims your agents would still keep repeating. Those are the trust leaks. Fix those first.
01 · list the promises
Do not audit abstract “messaging.” Audit live claims the machine can actually deliver to buyers, accounts, and prospects today.
02 · attach expiry
Set a freshness window that matches how fast reality moves. If policy, pricing, or capacity shifts weekly, your claim cannot live forever.
03 · degrade safely
When the clock expires, the machine must ask, qualify, or escalate — not keep talking like nothing changed.
Trust density rises when machine workers know when to stop repeating old truth.