Chapter 2: Information
The version everyone thought they had
Priya leads operations for a regional service business. On Friday afternoon, the finance team sends a spreadsheet showing that one branch has exceeded its quarterly budget. At Monday’s review, Priya explains that the excess comes from emergency repairs approved after a safety inspection. The chief operating officer looks surprised. His copy of the spreadsheet contains a comment stating that the cost was “unplanned local expenditure.” Priya’s copy does not.
The meeting stops while people compare files. There are four versions with almost identical names. Finance added comments for the executive group, removed the inspection report because the attachment was large, and sent branch managers the numerical sheet only. No one intended to mislead the room. The distribution process nevertheless gave one group an interpretation and denied another group the context needed to answer it.
Information problems are often described as communication failures, as if the remedy were simply to send more messages. The deeper questions concern selection, timing, context, and usability. Who receives the material? Which version? What background travels with it? How much time remains to act? Can recipients tell what is authoritative?
Teams need both shared and specialized knowledge. No one can or should know everything. Confidentiality, privacy, legal restrictions, security, and role relevance all justify limits. Yet research on group decision making shows a recurring difficulty: groups tend to discuss information already shared by members more than information held by only one member, even when the unique information is important to the best decision.1 Poor distribution can make this natural bias worse. Good design helps the group discover what it does not already have in common.
Information is not a substance poured equally
“Everyone got the email” sounds fair. It may not mean everyone received the same practical chance to understand and respond. A data analyst receives a model they can inspect; a community representative receives forty columns without definitions. A manager receives a verbal explanation; a remote colleague receives terse minutes. A team receives a policy after leaders have already discussed its likely effects for two weeks.
Fair information access is not identical volume. It is sufficient, relevant, timely, and usable information for a person’s role. Sometimes that requires different formats or levels of detail. The key is a common core: the facts, assumptions, definitions, decisions, and changes that participants need to share before their specialist views can be meaningfully combined.
This chapter examines five recurring problems: selective distribution, missing context, unequal advance briefings, concentrated knowledge, and information released too late to use.
1. Selective distribution of important information
A distribution difference becomes consequential when people participating in the same process receive materially different information without a reason connected to their roles or obligations. The difference may involve recipients, attachments, links, versions, comments, or access permissions.
Begin with need, not symmetry. Did the people being compared need the same information to perform the same part of the process? A legal team may receive privileged advice that cannot be circulated. A manager may receive personal data that should be withheld from a wider group. A security lead may see technical details that would create risk if copied broadly. In each case, others may still need a non-sensitive summary of the implication.
The strongest distribution systems distinguish the protected detail from the shared decision fact. “Legal advice is restricted” does not require “No one else may know that the contract cannot be terminated before June.” “The employee record is confidential” does not prevent a committee from learning that an approved accommodation affects the schedule. Redaction, role-based access, and summarized constraints allow privacy and participation to coexist.
When a difference is discovered, ask four factual questions:
- Which artifact or message differed?
- Who received each version, and when?
- What rule or reason governed the difference?
- What could recipients not do because of what they lacked?
In Priya’s case, the problem was not that executives received comments. It was that the comments framed expenditure in a way branch managers could not see, while the source that explained the expenditure was removed. A common decision pack with a version number, a short note about the safety approval, and links to restricted material would have prevented the conflict.
2. Missing background or constraints
Every summary omits. Good communication depends on omission; no executive paper can contain the full history of a program. The question is whether the missing context would change a reasonable person’s understanding of the choice.
Relevant context often includes the purpose of the decision, previous commitments, assumptions, dependencies, rejected alternatives, known risks, and the source of estimates. Consider a proposal to replace a supplier because its price is 12 percent higher. If the brief omits that the lower bidder has not passed a required safety audit, the headline comparison is incomplete. If it omits three years of ordinary correspondence, the decision may remain perfectly sound.
Missing context is frequently produced by division of labor rather than concealment. The author knows the background so well that they forget it is not shared. A deadline encourages cutting. A template rewards brevity. Teams store decisions across chat, email, slides, and people’s memories. Distributed work makes surrounding circumstances less visible, which can lead colleagues to explain a puzzling action as a personal failing rather than a response to conditions they cannot see.2
A compact decision brief can reduce the problem. It should answer:
- What decision is requested?
- Why now?
- What must be true for the recommendation to work?
- What material constraints apply?
- What alternatives were considered?
- What prior commitment limits the choice?
- Where can a reviewer inspect the supporting record?
The brief need not be long. A sentence under each heading is better than a polished narrative that hides uncertainty. If space is tight, link to the record rather than pretending it does not exist.
Individuals raising a context problem should identify its decision relevance. “You left out my earlier email” sounds like a dispute about recognition. “The current brief assumes a two-week deployment, but the signed client schedule allows six weeks; that changes the cost comparison” explains why the omission matters.
3. Unequal advance briefings
Advance briefing is normal. Leaders prepare chairs. Specialists translate technical issues. Negotiators explore positions. A manager helps a new employee understand who will attend and what questions may arise. These conversations can make a shared meeting more productive.
They become problematic when selected participants use the head start to define the issue, assemble support, or answer objections before others know a decision is active—while the formal meeting is still presented as an open review.
Timing and content both matter. Two groups can receive a paper on the same day while one has already discussed draft versions for a month. One participant may receive a ten-minute logistics call; another may receive the chair’s preferred outcome and coaching on how to support it. The surface fact “both were briefed” does not capture the difference.
Do not assume that every separate briefing should be disclosed in full. People need candid preparation, and some conversations are legitimately confidential. Focus on material changes to the shared basis of decision. If a pre-brief produces a new option, constraint, forecast, or commitment, circulate that result. If it effectively settles the matter, do not call the later forum a consultation.
A helpful norm is the common-core rule: everyone responsible for the same decision receives the same essential facts, criteria, options, and material changes with enough time to use them. Supplementary briefings may vary according to role, accessibility, language, expertise, or responsibility. The core does not.
Managers can make this concrete by marking a document “common decision brief” and maintaining one change log. Participants then know that a side conversation may add interpretation but cannot silently alter the shared facts.
4. Knowledge concentrated in one place
Specialization is one of the reasons organizations exist. It would be wasteful for every person to master every system. Teams often develop a practical map of who knows what, sometimes called a transactive memory system: members rely on different people for different domains while coordinating their expertise.3
Specialization becomes a bottleneck when important work cannot continue, be checked, or be transferred without one person or group. The danger is not merely absence. Concentration can also give the knowledge holder unreviewed power over how facts are translated, which questions are answered, and when a dependency is considered satisfied.
There are several signs:
- only one person can run a recurring process;
- key decisions depend on explanations that are not documented;
- access credentials or source files have no authorized backup;
- requests repeatedly queue for translation through one role;
- the organization cannot distinguish expert judgment from policy; or
- handover has been postponed through several predictable absences.
The remedy is not to strip experts of value or publish sensitive knowledge widely. It is to make concentration deliberate. Identify critical knowledge, its owner, authorized backups, minimum documentation, access controls, and recovery plan. Pair specialists for major cycles. Record not only the steps but the judgment points: what signs cause an expert to stop, escalate, or choose a different path.
Research on knowledge transfer emphasizes that transfer depends on opportunities, characteristics of the knowledge, mechanisms, motivation, and the depth with which recipients engage.4 A document dump on someone’s last day is not transfer. The successor needs practice, feedback, and a chance to retrieve the knowledge under realistic conditions.
Be cautious with the phrase knowledge hoarding. Research often defines knowledge hiding as an intentional effort to withhold knowledge that someone requested.5 That is a claim about intention. In daily use, it is safer to begin with the process: requests go unanswered, documentation is missing, or no backup can perform the task. The cause may be overload, fear of becoming replaceable, poor incentives, confidentiality, or a role never designed for teaching. Each cause suggests a different response.
5. Information arriving too late
Information can be technically available and practically useless. A 200-page contract released at 4:00 p.m. for a 9:00 a.m. approval gives reviewers access without a meaningful review opportunity. A new risk disclosed after a meeting begins may demand a pause, even if the deadline is real.
Late information has several common causes: facts emerged late; an external party delayed; a review took longer than expected; someone forgot; the team worked from different time zones; or leaders were reluctant to share an uncertain draft. Test these before interpreting the delay as strategic.
Then assess avoidability and effect. Who knew what, when? What work did a reasonable review require? Did some participants have more time? Could the deadline move? Did the late material change the recommendation or only confirm it? Was an explicit exception recorded?
Minimum review periods help, but a fixed number of days is not enough. Five pages of familiar material and five hundred pages of novel analysis are not equivalent. A better rule combines a normal period with a material-change trigger. If new information could reasonably change the decision, the chair must either extend review, narrow the decision, or record why acting now outweighs the risk.
Individuals can make a bounded request:
I received the revised forecast forty minutes ago. The change affects two assumptions in my analysis. I can give an initial view today, but I need until noon tomorrow for a checked recommendation. If the decision must proceed, please record that the forecast was not fully reviewed.
This separates refusal from professional judgment. It also gives the decision-maker options.
More information can make the problem worse
When trust is low, organizations often respond by copying everyone on everything. The result is not transparency but noise. People cannot tell which message matters, personal data spreads unnecessarily, and responsibility for reading becomes impossible.
Information quality has at least five dimensions:
- Relevance: does the recipient need it for a defined role?
- Completeness: are the material facts and constraints present?
- Timeliness: can the recipient still act?
- Provenance: can a reader tell where it came from and which version is authoritative?
- Usability: is it in a form the intended reader can understand and access?
A decision system should optimize these dimensions, not message count. One numbered pack with a change log is often more transparent than thirty scattered emails. A short summary plus controlled access to evidence is often safer than attaching confidential records to a wide list.
Information and status
The same information can travel differently depending on who carries it. A concern presented by a senior leader may be treated as strategic; the same concern from a frontline employee may be treated as anecdotal. A spreadsheet from finance may be presumed authoritative while an operational explanation is asked to prove itself line by line. These differences can reflect relevant expertise, but status is an imperfect shortcut for reliability.
One way to reduce the shortcut is to separate the claim, evidence, and source role. A decision paper can state: “Branch repair cost: 84,000; approved after inspection finding S-17; source records: inspection report and maintenance authorization.” Readers can still weigh the competence and responsibility of the source, but the claim does not depend only on the source’s standing.
Anonymous input is sometimes proposed as a solution. It can reduce status cues and fear, particularly in an early idea-gathering exercise. It can also make clarification difficult and hide relevant conflicts of interest. Use anonymity for a defined purpose, not as a universal substitute for a safe culture. In many decisions, a better approach is to gather views independently before discussion, display the evidence consistently, and then invite named expertise where questions require it.
Leaders can also change whose information enters the room by asking deliberately. “What do we know from direct customer contact?” “Which assumption is least supported?” “Who holds information that has not yet been discussed?” These prompts counter the tendency to repeat what everyone already knows. The goal is not to force every participant to speak. It is to make unique, relevant knowledge welcome before the group converges.
Communicating uncertainty without losing authority
People sometimes withhold early information because they fear that uncertainty will be mistaken for incompetence. A forecast owner may wait until numbers are polished, a project manager may avoid sharing a risk that could disappear, and a leader may hold a draft until every stakeholder agrees. The result is a paradox: recipients receive information that looks certain only after their chance to influence it has narrowed.
A mature information system allows calibrated uncertainty. Useful phrases include:
- “This is a working estimate, currently 60–70 percent confidence.”
- “The direction is stable; the amount may change after supplier confirmation.”
- “Two interpretations remain. The decision does not depend on resolving them today.”
- “This risk is low probability but high consequence, so we are sharing it before validation is complete.”
- “Do not rely on this draft for external commitments.”
These statements are not weakness. They tell recipients how to use the information. A traffic-light label can help, but plain language is better than an unexplained color. Specify what is known, what is uncertain, who is checking, and when the next update is due.
The same discipline applies when correcting information. Quietly replacing a file preserves a clean appearance but creates multiple realities. Mark the change, state whether it affects the decision, and notify the people who used the earlier version. A correction culture becomes credible when leaders model it on their own materials.
A case in full: the delayed safety note
Imagine a warehouse team preparing to introduce a new packing line. The project board receives a Friday paper recommending launch on Monday. A short note says that the safety review is “complete subject to routine closeout.” The worker representative receives the paper on Sunday evening and learns from a supervisor that one guarding test failed on Thursday.
Several patterns may be present, but they should not be collapsed too quickly. The representative received the paper late. The failed test is missing from the shared context. Some board members may have received an advance verbal briefing. The safety specialist may hold knowledge that no one else can interpret.
Begin with the decision fact: can the line safely launch on Monday, and who has authority to decide? Ask when the failed test became known, how it was classified, whether a compensating control exists, and what “routine closeout” means under the organization’s safety process. The specialist may explain that the test concerned an optional cover and that a verified interlock provides the required protection. Or the phrase may have minimized an unresolved hazard. The facts must lead.
The immediate response should match the stakes. Where safety may be affected, the board should not preserve the schedule merely to avoid embarrassment. It can pause, obtain the authorized safety judgment, record the evidence, and use the formal escalation route. If a launch proceeds under an exception, the owner, basis, duration, and required follow-up should be explicit.
After the immediate issue, examine the information design. Why could a failed test be summarized as routine closeout without a common definition? Why did the worker representative receive the pack later? Did the specialist have a standard way to flag a material change? Fixing these routes matters even if the line was ultimately safe and every person acted in good faith.
This example also shows why a pattern guide is not a substitute for domain expertise. A process observer can identify late, asymmetric, or incomplete information. Only an authorized and competent safety process can determine whether the equipment is safe.
Team norms that make information usable
Teams can adopt a small set of norms without building a heavy knowledge-management program:
One source for the decision. Working material may live in many places, but the current brief and final record have known locations.
Changes announce themselves. A revised number, assumption, recommendation, or constraint is marked; recipients are not expected to discover it by comparing files.
Questions travel with answers. If a side conversation resolves a question that others share, the result returns to the common record.
Restrictions have reasons. Sensitive detail is limited by role and purpose; the existence and decision consequence of the restriction are explained where possible.
Silence is not acknowledgment. For high-stakes material, the owner requests a specific confirmation or response. For ordinary material, the team does not pretend that copying a mailbox transferred understanding.
Uncertainty has an owner and a date. Open questions identify who is checking and when the group will know more.
These norms shift communication away from performance—appearing informed, polished, and in control—and toward coordination. They also reduce the personal burden on people who repeatedly have to ask whether they received the whole story.
Common traps in an information dispute
The first trap is assuming that possession implies understanding. Sending a technical model without definitions can shift blame to the recipient while withholding practical meaning. Ask what someone needed to know or do, not merely what file existed.
The second is confusing confidentiality with secrecy. Confidentiality is bounded by a legitimate purpose, defined access, and handling rules. Secrecy is simply the absence of explanation. Even when detail cannot be shared, process facts often can: that advice was obtained, that a restriction applies, who may review it, and when it will be reconsidered.
The third is demanding private correspondence to prove equal treatment. Do not ask to inspect colleagues’ inboxes or confidential briefings merely because you suspect an asymmetry. Start with distribution lists, version histories, decision packs, and other records you are authorized to use. If a formal review needs protected material, let an authorized reviewer handle it.
The fourth is treating every late disclosure as intentional. Information systems are messy. Look for repetition, selective effect, prior knowledge, and resistance to correction before escalating the interpretation.
The fifth is solving a trust problem with surveillance. Monitoring who opens every message may create more privacy and relationship costs than the original issue. Design a process in which the critical material is identifiable and acknowledgment is requested only where the stakes justify it.
For managers: build a common spine
Teams do not need one giant repository. They need a common spine for consequential work. For each recurring decision, define:
- the authoritative brief;
- the owner and version number;
- the common core all decision participants receive;
- the restricted annexes and access reasons;
- the normal review period;
- the method for highlighting material changes;
- the place where questions and corrections are recorded; and
- the final decision and rationale.
This structure reduces accidental asymmetry and makes legitimate differences visible. It also creates a better boundary between a working discussion and an official record. Not every chat message needs preservation; the consequential fact should be carried into the spine.
Priya’s company adopts such a process after the budget meeting. Finance issues one numbered workbook, a one-page narrative, and links to supporting records. Restricted comments are labeled and summarized where possible. When an executive asks for a change, it appears in the log. The process does not eliminate disagreement. It prevents disagreement about which reality the group was given.
Practice: reconstruct one information path
Choose a recent decision that surprised someone. Draw a simple timeline with five points:
- when the issue first became known;
- when the first draft or interpretation was created;
- who received it;
- when material facts changed; and
- when each affected role had a realistic opportunity to respond.
For each difference, write a possible legitimate explanation and the evidence that would confirm it. Then identify one process improvement that would work even if no one acted badly: a common brief, a version number, a redacted summary, a review period, or a change trigger.
Information creates the raw material for influence. The next chapter turns to the scarce resource that determines whether the material is ever considered: attention.