Why does information flow matter in business operations?
Most business processes are sequences of decisions and information handovers.
Sales needs to know, for example:
- what is in stock;
- what is already reserved;
- when new goods will arrive;
- which prices it may offer;
- which commitments it may make to customers.
Purchasing needs to know:
- what is required and in which quantities;
- when it is needed;
- which orders must be fulfilled;
- which stock is already reserved;
- which supplier circumstances have changed.
The warehouse needs to know:
- what to prepare;
- which order takes priority;
- what has arrived;
- what may be issued;
- and which discrepancies have occurred.
If any of this information is interrupted or changes without the next decision point knowing, an operational blind spot emerges.
The process often fails at an earlier information handover, rather than where the problem becomes visible.
Which information flow blind spots are most common?
1. Information does not reach the next participant
Data is created, but the employee or system concerned does not receive it.
For example:
Purchasing already knows the supplier is three days late, but sales still sees the original delivery date.
2. Information arrives too late
Information technically reaches its recipient, but only after the decision has been made.
The data exists, but is no longer useful for that business decision.
Timing is part of information's value.
3. Several versions of the same information exist
For the same product, for example:
- ERP shows 430 units;
- Excel shows 390;
- the warehouse list shows 410;
- the salesperson considers 350 saleable.
This is more than a data discrepancy.
The business has no clear rule defining which information is authoritative for which question.
4. The same word carries different meanings
Typical examples:
- “ready”;
- “reserved”;
- “available”;
- “approved”;
- “confirmed”;
- “deliverable”;
- “closed”.
One department may understand a “confirmed order” to mean that the customer accepted it.
Another may understand that stock is secured.
A third may assume the delivery date is final too.
The same expression does not guarantee the same business meaning.
This is particularly serious in ERP integration and AI implementation.
5. Information changes, but the change does not propagate
Data has reached the right place once.
Later, it changes.
The question is no longer whether the original data was transferred.
It is:
has the change reached every point where decisions may still rely on the original information?
This is a common hidden information flow defect.
6. There is no clear information owner
If it is unclear:
- who is responsible for data accuracy;
- who may change it;
- who checks it;
- and who decides when there is a discrepancy,
the problem can easily become ownerless.
Each department may assume another has handled it.
7. Information remains in informal channels
Significant business activity often takes place outside ERP or other official systems.
Instead, it happens:
- in email;
- in Teams;
- in WhatsApp;
- in Excel;
- by phone;
- in face-to-face discussions;
- on paper;
- or in an employee's memory.
Informal channels are not inherently wrong.
The problem arises when business-critical information exists only there and subsequent participants cannot access it reliably.
How can information flow blind spots be uncovered?
The investigation should not begin with:
“Which systems does the company use?”
A more useful question is:
“Which information is needed to make this decision correctly?”
Then trace that information back.
1. Select a critical piece of business information
For example:
- committed delivery date;
- available stock;
- customer price;
- order status;
- payment status;
- purchase requirement;
- expected arrival;
- production readiness.
Then examine:
- where it originates;
- how it is calculated;
- who records it;
- which system stores it;
- who uses it;
- who may change it;
- and which decisions depend on it.
2. Follow the complete information lifecycle
Examine information flow as an event chain:
creation → recording → processing → transfer → use → modification → feedback
At every point, ask:
- What happened?
- Who did it?
- In which system?
- Based on which data?
- When?
- What was the result?
- Who received it?
- How do we know they received it?
This reveals where information is lost, delayed or changed.
3. Examine handovers as well as systems
The greatest information flow risk often lies outside any single system.
It lies at the boundary between participants or systems.
For example:
sales → purchasing
purchasing → warehouse
warehouse → sales
ERP → online shop
CRM → ERP
ERP → management report
employee → employee
For every handover, record:
- the source;
- the recipient;
- the information content;
- the channel;
- the time;
- the expected response;
- and any acknowledgement.
Operational risk often appears at the boundaries between units, rather than within them.
4. Check the meaning of information
Checking whether the same field or expression exists in multiple systems is insufficient.
Check whether it means the same thing.
For example, “stock” may mean:
- physical stock;
- book stock;
- available stock;
- stock available for reservation;
- saleable stock;
- expected stock.
If two systems use “stock” for different concepts, integration may be technically flawless but commercially wrong.
Matching fields do not guarantee matching meanings.
5. Check timing
Information is often correct only at a particular point in time.
Ask:
- when it was created;
- when it changed;
- when it reached the next participant;
- and when the decision was made.
Information can be accurate yet unusable because it arrives late.
For example:
the supplier delay information is entirely correct.
But if it reaches sales after a delivery date has been promised to the customer, information flow has failed operationally.
6. Find parallel “truths”
Pay particular attention to questions answered from several data sources.
For example:
“When will this product arrive?”
The answer may be found:
- in ERP;
- in a purchasing spreadsheet;
- in a supplier email;
- in a Teams message;
- or in the buyer's personal notes.
The question is not simply which system is best.
Establish:
which source is authoritative for which claim.
Kyntrax does not necessarily seek one universal single source of truth.
A business can have several authoritative sources for different business claims.
The critical requirement is clarity about these distinctions.
7. Examine feedback
Sending information does not establish that the handover is complete.
Sending an email establishes only that the message was sent.
It does not prove that the recipient:
- read it;
- understood it;
- processed it;
- updated the relevant system;
- or changed the next decision accordingly.
For critical information, it may therefore be important to examine transfer and processing separately.
Which evidence should you rely on?
Reconstruct information flow using more than interviews.
| Area examined | Examples of evidence | Verification question |
|---|---|---|
| Information creation | order, event, measurement, document | Where and when was the information created? |
| Data source | ERP, CRM, Excel, email, document | Which source is authoritative for the question? |
| Handover | system event, message, interface, document | Did the information reach the correct recipient? |
| Timing | timestamp, status change | Did it arrive before or after the decision? |
| Meaning | field definition, business rule | Does the same concept mean the same thing to everyone involved? |
| Change | change log, new status | Did the change propagate to every affected point? |
| Processing | subsequent transaction or decision | Did the recipient actually use the information? |
| Responsibility | role, process owner | Who owns data accuracy and updates? |
Evidence aims to reconstruct how information travelled from one point in the company to another, or where that journey broke down.
Illustrative example: supplier delay
Suppose a supplier reports on Monday that:
goods previously promised for Friday will arrive the following Tuesday.
The buyer receives the information.
At first glance, the matter seems resolved:
the company knows about the delay.
But does it really?
Examine:
- Has someone recorded the new date?
- In which system?
- Has the order's expected arrival date been updated?
- Has the warehouse been informed?
- Has sales been informed?
- Does the salesperson still see the old date?
- Is a customer order affected?
- Has the customer commitment changed?
- Has the customer been informed?
- Who is responsible for ensuring this happens?
Every department may have performed its own task correctly.
Yet the failure still occurs.
The blind spot is in the connections between them.
Why does a dashboard not solve this automatically?
A dashboard can show only what the underlying information system knows.
If:
- the data is incorrect;
- updates are late;
- its meaning is unclear;
- information is not transferred properly;
- or important information remains in a separate spreadsheet or email,
a new dashboard merely presents the incomplete operational picture more attractively.
Understanding information origins and journeys is therefore essential before introducing management reports or dashboards too.
Better presentation cannot replace reliable information flow.
How do information flow blind spots relate to ERP problems?
Very closely.
ERP is often blamed because:
- it displays incorrect information;
- a status is missing;
- something is not updated in time;
- or users work from other sources.
But ERP may simply display the consequences of an earlier information flow problem.
System faults and information faults are therefore different.
An ERP investigation must establish:
- where data originates;
- how it enters the system;
- which rule changes it;
- and how it moves onwards.
Why is this particularly important before AI implementation?
Because AI also works from information.
An assistant or agent cannot reliably support business decisions unless it knows:
- which source is authoritative;
- what each item of data means;
- how current it is;
- how it relates to other data;
- and what to do when sources conflict.
If AI is asked:
“When can we deliver 500 units to this customer?”
it must connect several kinds of information:
- physical stock;
- reservations;
- expected arrival;
- customer priority;
- transport capacity;
- order status.
If any of these information chains is flawed, AI can give an extremely convincing but commercially incorrect answer.
Therefore:
reliable information flow is a prerequisite for reliable AI.
Is asking colleagues enough?
No.
Interviews matter.
They reveal how participants understand operations.
But memory and actual process behaviour do not always match.
Compare reported practices with:
- real transactions;
- system data;
- documents;
- event times;
- emails;
- spreadsheets;
- and direct observation.
The gap between:
“this is how we usually do it”
and:
“this is what happened in this specific case”
is often the blind spot itself.
When is information flow adequate?
Digitalising all data is not enough.
Connecting all systems is not enough either.
A critical information flow is adequate when it is clear:
- where information originates;
- what it means;
- which source is authoritative;
- who owns it;
- who uses it;
- when it must be updated;
- how it reaches the next participant;
- how changes are handled;
- and whether the next decision point can be shown to have used the right information.
What should change after a blind spot is found?
Not every blind spot requires a new system.
The appropriate intervention may be:
- clarifying a concept or business rule;
- assigning responsibility;
- process changes;
- defining a data owner;
- changing a system setting;
- integration;
- an automatic notification;
- a shared data source;
- eliminating a manual handover;
- or new technology where needed.
Choose technology according to the cause of the blind spot.
The Kyntrax approach
Kyntrax examines more than the process steps that exist.
It also asks:
how does the next participant know what to do?
That is the essence of information flow.
The Kyntrax Mirror therefore examines together:
- the process;
- the information;
- the data source;
- timing;
- responsibility;
- decision points;
- exceptions;
- and handovers between systems.
The aim is not to create as many flowcharts as possible.
It is to understand whether the company's decisions actually rest on the right information.