Self-Efficacy on a Program Everyone Has Quietly Given Up On

The programme landed on my desk the way stalled systems usually do: still scheduled, still funded on paper, and already abandoned in the room. The steering pack was current. The RAID log was mostly green. The operations lead who used to challenge dates had stopped opening the document. Nobody had cancelled anything. Belief had simply left.

That inheritance is more common than the industry likes to say — a live programme with a calendar, a vendor contract, and a quiet consensus that the next date is theatre. The question is not how to "stay positive." It is how to keep a usable sense of agency after the collective story has collapsed.

The Quiet Abandonment — Belief Dies Without a Memo

Quiet abandonment rarely arrives as a kill decision. It shows up as skipped reads, empty decisions, and status meetings that still fire because the invite series outlived the sponsorship.

Portfolio writers have a blunt name for the pattern: initiatives that fail to deliver yet keep shuffling along, consuming attention without strategic impact. Survey work puts numbers on the everyday version — stalled work the boss still calls "active," channels gone silent, status untouched for weeks — and finds large shares of office workers carrying that zombie-project weight into a new year.

The persistence reasons are ordinary. Nobody owns the kill decision. Sunk cost starts to feel like a moral claim. Stopping looks like personal failure. The programme does not need a villain. It needs a decision owner who never arrives.

Watch the practical signals. Missed milestones with no decisive action. Updates that cannot name the value still being created. Weakened executive sponsorship. Scope that keeps widening because narrowing would force a choice. Energy leaving the team while the calendar stays full.

On a logistics CRM replacement I stayed with too long, the RAID stayed green for six weeks after the integration lead stopped booking the vendor call. Nothing was cancelled. The people who could have made it real had simply stopped looking.

That gap — declared commitment versus actual belief — is where self-efficacy either shrinks to a workable unit or evaporates into heroics.

The programme still has a status pack. Belief left earlier.

Why Morale Theatre Makes It Worse

When belief leaves, organisations reach for cheerfulness. Status slides fill with wins. Leaders ask for "energy in the room." Someone insists the roadmap is still right if people would only believe harder.

That is not optimism. Optimism can still name the risk. The forced version dismisses the evidence and replaces it with false reassurance — what workplace writers call toxic positivity.

Research on the workplace form is still catching up to the phrase. The early findings are not gentle. Forced positivity tracks with shame, emotional exhaustion, and intentions to withdraw — and the path runs through those middle states, not around them.

Optimism is not the antidote to demoralisation. Proof is.

Heroics are the twin failure. Working nights to "save" a programme sponsors have already mentally cancelled teaches the room the wrong causal story: that effort was the missing ingredient. It was not. Ownership was. Overnight heroics also spend the one resource self-efficacy needs next — a credible past performance you can still point to without sounding desperate.

Heroics do not reverse abandonment. They rehearse it with better lighting.

There is a quieter version of the same mistake: polishing the steering pack until the pack becomes the product. Beautiful status is still status. It does not restore agency.

Proof-Sized Work — The Unit Belief Can Still Afford

Self-efficacy, in the sense worth keeping for delivery work, is not belief that the whole programme will succeed. It is belief that you can execute a specific course of action in a prospective situation.

If that word is new in this register, the one-sentence version worth keeping is: confidence in your capacity to finish this slice, not faith in the roadmap's mythology.

Workplace summaries of the research line up on the source that matters most. Past performance — mastery — outweighs pep talks. Verbal persuasion is weaker and depends on who is speaking. Meta-analytic work has long tied efficacy beliefs to work-related performance with a meaningful average correlation. None of that says "believe in the programme." It says arrange a success you can verify.

That is Proof-Sized Work: one falsifiable slice, one owner, one visible consequence a sceptic can check without joining a belief system.

You do not need the whole roadmap believed. You need one slice you can still finish and show.

One falsifiable slice

Shrink until the claim can fail in public. Not "stabilise integrations." Prefer "reconcile receiving stock for Store Cluster A against the overnight feed for five trading days, and publish the mismatch count."

On a retail ERP cutover whose sponsors had stopped reading packs, that was the slice that still made sense. Receiving-only. One cluster. A before/after mismatch number. Nobody had to pretend the full go-live date was honest for the number to matter.

The warehouse supervisor did not ask whether steering still believed. She asked whether Cluster A's receiving numbers would stop lying by Friday. That is the constituency proof is built for.

The failure mode on this beat is shrinking into busywork. A tidy spreadsheet nobody will use is not proof. If the slice does not change a decision or a control for someone who still has a stake, cut it again — or stop.

One owner

Name the person who can finish the slice without a steering decision that will never come. If the slice needs three committees, it is still programme-sized. Cut again.

Shared ownership here is usually a soft cancel. Two names on the ticket, and neither will risk being the one who finishes while the room looks away.

One visible consequence

Close with a check the sceptics can run themselves: "If the mismatch count drops below X by Friday, Cluster A can trust receiving for the pilot; if it does not, we still have the same feed risk — and we have evidence, not a mood."

Proof-Sized Work is not a secret plan to rescue the whole programme by stealth. It restores a usable unit of agency while the cancel-or-continue decision is still missing. Sometimes the proof also makes that decision possible. Treat that as a feature.

The Fair Objection — If Nobody Believes, Why Continue?

The strongest objection is honest: continuing fiction is waste. Proof-sized deliveries can become a new form of status theatre if they postpone an overdue kill.

Concede that. If the only audience for the slice is your own anxiety, stop. Archive the work. Ask the kill question out loud: if we were not already running this initiative, would we approve it now?

Where the objection stops holding is when someone still needs a true answer — operations still receives stock, finance still closes, a regulator still expects a control — even though the steering table has checked out emotionally. In that window, agency is not loyalty to a dead story. It is one artefact the remaining constituency can use.

Self-efficacy on a quietly abandoned programme is not the same as faith in the programme. Confusing the two is how people burn out defending a fiction.

What You Can Try This Week

Pick one slice you can finish without a hero sponsor. Write the falsifiable claim in one sentence with a number or artefact. Name the owner in the ticket, not in a RACI nobody opens. Share the consequence check with the one person who still has a stake — warehouse, finance, support — before you update the steering pack.

If you cannot find a stake-holder outside the belief theatre, you may already have your answer about continuing. That answer is also a form of agency.

Belief in the programme is optional. Proof of your next slice is not.