Training Data
How do we define a problem: My experience tackling an issue
One important lesson I learned today is that before attempting to solve a problem, we must first understand the company’s current situation and reality.
For example, while working on a ticket regarding the time lag between courier partners marking an order as delivered and our internal system reflecting that update, I initially focused on the technical issue itself (How do we calculate the lag? What is the mean time lag?). However, by continuously asking the question “Why?”, I eventually realised that the core problem was not the lag itself, but rather the company’s current lack of visibility of that lag. That became the true reason why this ticket exists.
Peeling Back Layers with “Why?”
⏱️
THE SYNDROME
Courier marks an order as
delivered, but the system does not immediately mark the same order as delivered
WHY?
➔
📊
DEEPER
ISSUE
Data is fragmented. Thus,
there
is a lack of visibility into how large or frequent the time difference between the courier marking an
order as
delivered and the system marking it as delivered.
WHY?
➔
🎯
TICKET’S TRUE PURPOSE
We need to know whether
there are
late invoices being issued
Figure 1 — Asking “Why?” repeatedly reveals the true reason a ticket exists.
Naturally, many additional questions began to surface. Why does the lag happen? Is it caused by the current data pipeline architecture? How can the delay eventually be reduced or eliminated? While these are important questions, I learned that they are beyond the scope of the current ticket. This highlighted another important lesson: staying focused. If every possible problem is tackled at once, the original task loses clarity and becomes difficult to complete effectively.
🎯 Key Lesson: Not every surfacing question belongs in the same ticket. Staying focused on the
primary objective is what makes a task completable.
After understanding the company’s current situation, the next step is understanding the expectation — what the company believes it should be able to do. In the context of this ticket, the expectation is straightforward: the company should have visibility of the time when an order is delivered and when the tax invoice is issued.
Once the current situation and expectations are clear, we can then identify the business impact of the problem. In this case, this lack of visibility is concerning because tax invoices are required to be sent to customers on the same day that they receive their orders, in accordance with Tax Authority regulations. Timely tax invoice issuance is therefore important not only for operational visibility, but also for compliance and operational reliability.
The 4-Phase Problem Definition Framework
↓
↓
↓
↓
After looking at the available data, I realised that it is not possible to achieve the Initial Objective since the data is too fragmented. Thus, I looked back at the Current Situation and Expectation, and discussed with the team on how to proceed. We then concluded that we simply need to know the number of late invoices.
Figure 2 — My structured approach to defining a problem before attempting
to solve it.
Finally, after understanding the current situation, expectations, and business impact, we must properly scope the objective. The objective should remain simple, clear, and focused. For this particular ticket, the objective is not to identify the root cause of the lag or design a long-term solution for eliminating it. Those are separate problems that may require additional investigations or future tickets. The primary objective of this ticket is simply to find out when an order is delivered and when the corresponding tax invoice is issued.
Scope Clarity: What This Ticket Is (and Isn’t)
✅ In Scope
✓
Find out the delivered orders
✓
Find out the corresponding tax
invoice
✓
Establish if the invoice has been
issued late
or not
✓
Quantify the problem so it becomes
measurable (eg. the number of invoices were issued late)
❌ Out of Scope
✗
Identifying the root cause
✗
Investigating data pipeline
architecture
issues
✗
Designing a long-term solution to
prevent late
invoice issuance
✗
Solving all related problems at
once
Figure 3 — A clearly scoped objective keeps the work actionable and measurable.
Overall, this experience taught me that defining a problem properly is an important part of problem-solving itself. First, we understand the company’s current reality. Second, we define what the expected state should be. Third, we identify related problems without losing focus on the primary task. Finally, we scope the objective clearly and precisely so that the work remains actionable and measurable.
Key Takeaways
🔍
Understand Reality First
Ground yourself in the company’s current
situation
before jumping to solutions.
🎯
Define the Expectation
Clarify what the company believes it should
be able
to do — that gap is the problem.
⚠️
Know the Business Impact
Understand why it matters — compliance,
operations,
or reliability — without losing focus.
✅
Scope the Objective Clearly
Keep the objective simple and focused so the
work
stays actionable and measurable.
Figure 4 — Defining a problem properly is itself an important part of
problem-solving.
