Your Approval Process Should Not Depend on Chasing People
A purchase request needs approval.
The employee sends a message to their manager.
No response.
A few hours later, they follow up on WhatsApp.
The manager says finance also needs to approve it.
Finance asks for a supporting document.
The employee sends the document by email.
The next morning, someone asks:
"Has this been approved yet?"
Nobody is completely sure.
This may sound like a communication problem.
Often, it is actually a workflow problem.
The business has an approval policy, but the process for executing that policy exists across messages, inboxes, calls, spreadsheets, and people's memory.
As the company grows, this becomes more than an inconvenience.
It creates operational latency.
Approval delays are often invisible
A delayed order is relatively easy to notice.
A delayed approval is different.
The work may simply sit somewhere between two people.
There is no obvious error.
Nothing has technically failed.
Someone is just waiting.
Consider a discount request.
A salesperson wants to offer a customer a 12 percent discount.
They need approval from a manager.
The manager is in meetings.
The salesperson sends a message.
Three hours pass.
The customer is waiting for confirmation.
Eventually, the manager approves it.
From a reporting perspective, the process succeeded.
From the customer's perspective, the company took three hours to make a relatively simple decision.
Multiply that across purchasing, expenses, leave, pricing, hiring, vendor payments, inventory adjustments, refunds, contracts, and operational exceptions.
The cumulative delay can become significant.
The real problem is not approval itself
Businesses need controls.
Not every employee should be able to approve every expense, discount, refund, or purchase.
Removing approvals entirely would often create more risk, not less.
The problem is how approvals are coordinated.
A healthy approval process should answer several questions immediately:
Who owns the next decision?
What information do they need?
What authority do they have?
How long has the request been waiting?
What happens if they do not respond?
Who approved the final decision?
If answering those questions requires searching through messages or asking several people, the workflow is carrying unnecessary friction.
Informal approvals work surprisingly well at first
A five-person company can run many processes through conversation.
"Can I buy this?"
"Yes."
"Can we give this customer a discount?"
"Go ahead."
"Can we refund this order?"
"Approved."
The decision-maker is nearby, context is shared, and request volume is low.
As the organization grows, the same model begins to break.
There are more employees.
More departments.
More requests.
More policies.
More managers.
More exceptions.
More financial exposure.
The company tries to preserve control by adding approval steps.
But if the underlying workflow remains informal, every additional control can also create another waiting point.
This is similar to the broader scaling problem discussed in 7 Signs Your Business Has Outgrown Spreadsheets.
The problem is not that spreadsheets are inherently unsuitable.
The problem begins when employees spend significant time moving information between systems simply so management can understand what happened.
WhatsApp and email are communication tools, not workflow engines
There is nothing inherently wrong with approving something through email or chat.
The problem begins when those conversations become the system of record.
Imagine a purchase request approved through WhatsApp.
Later, finance needs to know:
Who requested it?
What was the amount?
Which budget did it belong to?
Who approved it?
Was a second approval required?
Was the purchase actually completed?
Where is the invoice?
The answer may now be distributed across a chat conversation, an email attachment, an accounting entry, and someone's memory.
The approval happened.
The organization did not necessarily capture the process around it.
Approval workflows need context, not just buttons
Automation is sometimes approached too simply.
Request.
Approve.
Reject.
Done.
Real business decisions usually need more context.
A manager approving a purchase may need to see the requested amount, supplier, department, budget availability, previous purchases, supporting quotation, and reason for the request.
A sales manager approving a discount may need deal value, margin, customer history, standard pricing, and the salesperson's justification.
A finance leader approving payment may need the purchase order, invoice, delivery confirmation, payment terms, and previous approval history.
A useful workflow puts the decision and its relevant context together.
Otherwise, the company simply automates the notification while employees continue searching elsewhere for the information required to decide.
Not every request should follow the same approval path
One of the biggest opportunities in workflow design is conditional routing.
Suppose every purchase requires director approval.
That may make sense for a BDT 1,000,000 equipment purchase.
It probably does not make sense for a routine BDT 2,000 operational expense.
Instead, businesses can define authority thresholds.
For example:
Routine requests below a defined amount may be approved by a department manager.
Larger requests may require finance.
Requests above another threshold may require a director.
Certain categories may always require additional review.
Exceptions may be escalated regardless of amount.
The exact rules depend on the business.
The principle is more important:
**Control should be proportional to risk.**
When every request receives the same level of scrutiny, senior people become bottlenecks for decisions that do not require senior judgment.
Good automation separates routine decisions from exceptions
This is where workflow automation becomes more useful than simply digitizing a form.
Suppose a company processes 300 expense requests each month.
Most are routine.
A smaller number are unusual because they exceed a threshold, violate policy, lack documentation, or belong to a restricted category.
A well-designed system can make routine requests easy to process while directing attention toward the exceptions.
That means experienced managers spend less time confirming predictable decisions and more time reviewing situations where their judgment actually matters.
The same principle applies to many operational workflows:
Discount approvals.
Purchase requests.
Refunds.
Inventory adjustments.
Leave requests.
Vendor onboarding.
Payment authorization.
Contract review.
Hiring requests.
Credit limits.
Automation should not eliminate human judgment.
It should reserve human attention for decisions where judgment creates value.
Every approval should leave a usable history
A structured approval process creates something informal communication rarely provides well:
an audit trail.
The organization should be able to see:
Who submitted the request.
When it was submitted.
What information was provided.
Who reviewed it.
What decision was made.
When the decision happened.
Whether the request was modified.
Why an exception was approved.
What happened afterward.
This is useful beyond compliance.
Approval history helps businesses understand how their own operations work.
Management can identify where requests repeatedly become stuck.
Finance can see which categories generate unusual exceptions.
Operations can discover whether certain teams consistently require escalation.
Policies can be improved based on actual behavior instead of assumptions.
This connects directly with the broader principle behind our article on When Your Business Reports Arrive Too Late to Be Useful.
Information creates more value when it becomes visible while someone can still act on it.
Escalation should be part of the process
Every organization has situations where the assigned approver is unavailable.
They may be travelling.
In meetings.
On leave.
Managing a large queue.
An informal process usually handles this through more chasing.
"Did you see my message?"
"Can you approve this?"
"Who can approve while she is away?"
A structured workflow can define what happens automatically.
For example, if a request remains pending beyond a certain period, the approver receives a reminder.
If it remains unresolved longer, it can escalate.
If the approver is unavailable, delegated authority can apply.
Urgent categories can follow different response targets.
The objective is not to bombard managers with notifications.
It is to prevent important work from disappearing into someone's inbox.
Approval data can reveal process problems
Once approvals become structured, management can measure them.
Average approval time.
Requests waiting beyond SLA.
Approval volume by department.
Rejection rate.
Most common exception reasons.
Requests repeatedly returned for missing information.
Approval workload by manager.
Frequency of escalations.
These metrics can reveal something important.
Sometimes the approver is not the bottleneck.
Perhaps employees repeatedly submit incomplete requests.
Perhaps policy is unclear.
Perhaps approval thresholds are too restrictive.
Perhaps one manager receives requests that should be delegated.
Perhaps the company has added controls over time without removing obsolete ones.
Good operational visibility should help management improve the process, not simply measure how slowly it operates.
Do not automate a bad approval policy
This is an important warning.
If a company requires six approvals for a routine decision that realistically needs two, software can make those six approvals move faster.
It does not make the policy good.
Before automating an approval workflow, ask:
Why does this approval exist?
What risk does it control?
Who actually has enough context to make the decision?
Can routine cases be handled through predefined rules?
Which situations genuinely require escalation?
Could any approval step be replaced by visibility rather than permission?
That last question is particularly useful.
Sometimes senior management does not need to approve a transaction.
They simply need to know it happened.
Changing a step from mandatory approval to notification can preserve visibility without creating another operational queue.
Start with the workflows people chase most often
Businesses do not need to automate every approval process at once.
Look for the sentences employees already use:
"I'm waiting for approval."
"I sent it yesterday."
"Can you remind him?"
"Who needs to approve this?"
"I thought finance already approved it."
"Please send the approval screenshot."
Those conversations are process signals.
Choose one recurring workflow and map it from request to completion.
Identify the requester, decision-makers, information requirements, business rules, thresholds, exceptions, escalation path, and final system of record.
Only then decide what technology should support it.
Depending on the organization, that may involve an ERP workflow, CRM automation, internal business application, system integration, or a purpose-built approval module.
Neura Solutions helps businesses design custom software, workflow automation, ERP and operational systems, and integrations around processes that have outgrown manual coordination.
The technology should follow the operating model, not the other way around.
A mature approval process makes waiting visible
The objective is not to eliminate approvals.
It is to eliminate uncertainty around them.
An employee should not need to chase three people to discover who owns the next decision.
A manager should not need to search through chat history to understand what they are approving.
Finance should not need screenshots to reconstruct authorization.
Leadership should not discover approval bottlenecks only after customers, suppliers, or employees complain.
A good workflow makes the request, context, owner, decision, status, and history visible.
Because in a growing business, "waiting for approval" should describe a temporary workflow state.
It should not describe how the company operates.