Bad news early: how a business teaches people when to speak up

A job sits at “a bit behind” for six weeks, then turns into a serious problem in a single conversation, arriving with a recovery plan that has obviously been in preparation for some time. Nobody lied. Every update was defensible on the day it was given. The real question is not who held back what, but what the business had already taught its people about the cost of raising a problem early.

Most owners and managers believe their door is open, and the evidence they offer is that they have said so. Their staff are consulting a different source: what happened to the last person who brought an unwelcome fact into a room where everyone had already agreed things were on track. That memory, not the stated policy, decides when the next piece of bad news arrives.

Every business makes decisions on a picture assembled from what people choose to report. This article explains why that picture is fragile, how well-meant management advice can quietly close the channel, what late news costs in lost options, and the practical design choices that make early warning normal rather than brave.

Why upward information is fragile

Financial information travels upward inside a system built to protect it: accounting standards, reconciliations, external review and personal consequences for misstatement. Operational information, such as whether a job will finish on time, whether a supplier is struggling or whether a new process is working, has almost none of that. There is no audit, no agreed definition, no independent check, and its flow depends on what the person reporting expects to happen next.

Operational information is also time-sensitive in a way financial information is not. A misstated figure is worth correcting whenever it is found. A supply problem is worth knowing about while there are still options to respond, and less with every week that passes. How quickly bad news travels is therefore a business variable, not a matter of style.

Advice that quietly closes the channel

Some popular management advice, sensible in its original setting, becomes harmful when it turns into workplace culture.

“Focus on the goal, not the problem.” As a personal technique for staying steady after a setback, this has value. As a business norm, it is risky, because managing a business means holding the goal and the threats to it at the same time. Risk lists, contingency allowances and staged commitments all exist because goals are uncertain. A team told to focus on the goal learns which half of the picture to stop reporting.

“Look for the positive before responding.” Useful for composure. Applied before reporting, it becomes a filter: what reaches the decision-maker is an interpretation with the uncomfortable details removed. Establish the facts, pass them on, then choose the stance.

“Don’t bring me problems, bring me solutions.” This deserves care. Expecting the owner of a known, well-understood issue to arrive with analysis and options is reasonable and raises the quality of discussion. Requiring a worked solution before anyone may raise a concern is a tax, and it falls hardest on early, weak signals that cross team boundaries, which nobody yet owns and nobody can yet solve. Those are often the most valuable signals of all.

Organisational researchers such as Edgar Schein have long distinguished between the values an organisation states and the assumptions that actually operate beneath them. The stated value may be openness. The operating assumption is inferred from the last time someone raised a problem, and it decides what you will be told next month.

Latency and fidelity

Two properties of upward information are worth managing directly:

• Latency: the time between the first moment anyone in the business knew about a problem and the moment the person who could act on it knew.
• Fidelity: how much of the original signal survives the journey, especially its uncertainty. A problem stripped of its uncertainty is usually a problem stripped of its urgency.

Neither is measured in most businesses. Both can be reconstructed from records the business already holds, such as emails, job notes and meeting minutes.

Speaking up has a price

Raising a problem costs the person raising it something: exposure, the risk of being made responsible for a problem they merely noticed, or the loss of standing as someone who copes. The benefit is spread across the business and arrives later. Nobody needs to be dishonest for that arithmetic to produce silence. They need only be sensible and slightly patient, waiting one more week in the hope that the problem resolves itself.

The result is that the main failure is delay, not fabrication, which is why exhortations to be transparent rarely work. And the filter selects badly. It suppresses the early, uncertain signals that have the most value, and lets through the late, well-evidenced ones that can no longer change anything.

Not all of this comes from fear. The psychologist Irving Janis described how cohesive groups suppress dissent through a shared preference for agreement, often called groupthink. A leadership team that has publicly committed to a plan creates that pressure simply by having committed.

What silence costs

Consider a supplier problem noticed by a supervisor in week two of a job. Raised then, several responses exist: find another supplier, resequence the work, or talk to the customer about dates while it is still a conversation rather than a breach. Raised in week six, perhaps one response remains, and it is the most expensive. Nothing about the problem changed between those dates. The number of options attached to it did. That loss of options is the true cost of latency, and it belongs not to the person who waited but to whoever made waiting sensible.

There is a second cost. When owners suspect the picture they receive is unreliable, they compensate by checking everything themselves. It works well enough to become permanent, and the business ends up dependent on one person’s direct observation, with all the limits that brings.

Four design choices

You cannot exhort people into early disclosure. A statement that the door is always open costs nothing to make and carries little information. What changes behaviour is changing the price of speaking up:

1. Separate the report from the fault. If the business cannot distinguish the person who raised a problem from the person who caused it, reporting becomes a confession.
2. Do not make the reporter the default owner. Otherwise every warning is a self-assigned extra job, and people will price it accordingly.
3. Guarantee a response within a stated time. An escalation that disappears without reply teaches more than one met with irritation.
4. Judge raising and handling separately. Recognise the person who raised an issue early, even if the issue itself was not handled well.

Layers smooth the signal

Each layer of management applies a small, reasonable smoothing. A delay becomes “timing pressure”. A range becomes its middle. No step is dishonest; each is a summary for a busier audience. But the effect adds up. In a business with several layers, the signal reaching the top can differ in kind from what the front line saw.

Two remedies help. First, keep at least one route for information that does not pass through every layer, clearly understood to carry information, not decisions. Second, ask about exceptions rather than summaries: “What is the one thing that might go wrong this month?” rather than “How is it going?” Summaries are where smoothing lives.

A simple self-check

None of this needs a staff survey. Each question can be answered from the past few months of your own records:

If the business cannot reconstruct when a problem was first known internally, it does not have an information channel it can manage. It has a sequence of announcements.

A worked example

This is an illustration. The owner of a 25-person fabrication and installation business has long told staff, “Don’t bring me problems, bring me solutions.” On a major job, a site supervisor learns in week two that the main steel delivery will be about three weeks late. The supervisor tries to solve it alone and raises it in week six. By then, the only options are overtime and air freight, costing about $14,000, with the risk of a late-completion claim from the customer.

Had it been raised in week two, the business could have sourced part of the steel elsewhere, resequenced the installation or agreed a revised date with the customer at little cost.

The owner reconstructs the last four surprises in the business and finds that, on average, problems were known internally about three and a half weeks before reaching the owner. In each case, the person who knew had reasons to wait.

The owner makes four changes:

• An early warning item in the weekly toolbox meeting: anything that might go wrong, no solution required.
• The person raising a concern is not automatically its owner. The owner or a manager assigns it.
• A reply within two working days to any concern raised, even if only to say what will happen next.
• Recognition for early warnings in performance conversations.

Six months later, the number of early warnings raised has risen sharply, which the owner reads as a sign of health, not of worsening operations. On the next two significant problems, the time from first internal knowledge to the owner’s knowledge is about a week.

How this applies to a small Australian business

In a small business, the owner’s reactions are the channel. A single irritated response can shape what staff report for months. Practical steps:

• Reconstruct the timing of your last few surprises.
• Ask for early warnings without requiring solutions.
• Separate raising a concern from owning it.
• Reply to every concern within a stated time.
• Ask about exceptions rather than general updates.
• Notice your own reaction when bad news arrives, and what it teaches.
• Remember legal obligations: workers must be able to raise work health and safety concerns, and some reports, such as certain misconduct disclosures, have specific legal protections. Check guidance from Safe Work Australia, the Fair Work Ombudsman, ASIC or an adviser as relevant.

The articles on running a weekly business review and key-person dependence cover related practices.

Signals worth watching

• Status that jumps from fine to serious in one step.
• Problems arriving with solutions already prepared, suggesting they were held while a response was assembled.
• News reaching you first from a customer, supplier or auditor.
• A fall in concerns raised after a change of manager, often misread as improvement.
• The same kind of problem recurring at the same stage of successive jobs.

Common mistakes

• Assuming an open-door statement creates an open channel.
• Requiring solutions before concerns can be raised.
• Making the reporter the owner by default.
• Leaving concerns without a reply.
• Reading fewer escalations as better performance without checking.
• Relying on personal inspection instead of fixing the channel.

Frequently asked questions

Will this flood us with trivial concerns? Some, at first. A short, regular forum for early warnings, with quick triage, handles volume well. The cost of hearing a few trivial concerns is small compared with the cost of hearing an important one too late.

What if people raise concerns to avoid responsibility? Separating raising from owning does not remove accountability. Once a concern is assigned, the owner is accountable for handling it. The aim is to stop the act of raising a concern from being punished.

How do I know if my business has this problem? Reconstruct the last few surprises. If problems were known internally weeks before they reached you, the channel is slow.

Should early warnings be anonymous? Usually not for operational issues, because follow-up questions matter. For sensitive matters such as misconduct, a confidential route may be appropriate, and some situations have legal requirements.

Questions to ask

• For our last three surprises, how long were they known internally before I knew?
• What happened to the last three people who brought me unwelcome news?
• What must someone have in hand before raising a concern here?
• Do our reports have a place for what could still go wrong?
• If concerns raised halved next quarter, would we read it as improvement or failure?
• What might we be unable to hear that would change a decision already made?

Bringing it together

A business always makes decisions on a picture assembled from what people choose to tell it. The question is whether that picture was designed or left to chance. Bad news is an early-arriving fact, most valuable when it is least welcome. Measure latency and fidelity, stop requiring solutions as the price of raising a concern, separate reporting from fault and ownership, reply within a stated time and ask about exceptions. The quality of an owner’s information is not a measure of the team’s honesty. It is a measure of what the business has made honesty cost.

---

Source: KEVOS notes, drawing on general management research including Edgar Schein’s work on organisational culture and Irving Janis’s work on group decision-making. Examples and figures in this article are illustrations. This article is general information, not legal or employment advice.