The brief and the plan

01The problem

This is the capstone of the Business Analyst track. There's nothing new to learn here. Instead you'll do the whole job a junior business analyst is hired for, from a vague request to a recommendation the board can approve.

The company is Shieldline Insurance, a motor insurer with branches in Lagos, Abuja, Port Harcourt, Ibadan and Kano. Its managing director has sent you this:

"Our motor claims take far too long. Customers are leaving us, complaints are piling up and the regulator has started asking questions. In April, Lagos tried some changes to the claims process. IT wants us to buy a new claims system for ₦280 million instead. I need to know what's really wrong, whether the Lagos changes worked, and what we should do. The board meets in six weeks."

Notice what the brief contains: a symptom ("too long"), two consequences (customers leaving, complaints), one experiment (Lagos) and one solution already on the table (a new system). A BA's first job is to slow down and frame the problem before anyone buys anything.

02The concept

The arc of a BA project

Use the tools you know: SQL, Excel or Power BI. The lessons show SQL, which answers every question here; the same steps work as pivot tables or DAX measures.

A problem statement, not a solution

A good problem statement says who is affected, what happens, how much (with a number), and why it matters to the business. It doesn't name a solution: "We need a new claims system" is a solution, and it closes off the options before you know the cause.

Scope

Say what's in and what's out. In: motor claims from submission to payment, the five regions, July 2025 to June 2026. Out: underwriting and pricing, non-motor products, and choosing a system vendor.

Definitions before numbers

Write these down before you calculate anything. Without them, two analysts get two different "average days" from the same data and the board stops trusting both.

Stakeholders

Plot each stakeholder by influence and interest. Manage closely those high on both, keep satisfied those with high influence but less interest, and keep informed those with high interest but little influence.

03Example

The data has seven files:

A first look at the baseline:

SELECT COUNT(*) AS claims,
       SUM(outcome = 'Paid') AS paid,
       ROUND(AVG(CASE WHEN outcome = 'Paid' THEN julianday(closed_at) - julianday(submitted_at) END), 1) AS avg_days_to_settle,
       ROUND(100.0 * SUM(outcome = 'Paid' AND julianday(closed_at) - julianday(submitted_at) > 30) / SUM(outcome = 'Paid'), 1) AS pct_paid_over_30_days
FROM claims
WHERE submitted_at < '2026-04-01';

claims  paid  avg_days_to_settle  pct_paid_over_30_days
  2239  1904                25.2                   22.9

And the stakeholders, by influence and interest:

SELECT influence, interest, GROUP_CONCAT(role, '; ') AS stakeholders
FROM interviews
GROUP BY influence, interest
ORDER BY influence, interest;

influence  interest  stakeholders
High       High      Managing director; Head of claims
High       Low       Compliance officer
High       Medium    Agency manager; Finance controller; Head of IT
Low        High      Claims officer; Customer (accident damage)
Medium     High      Claims manager, Lagos; Senior assessor

The high-influence, high-interest people are the managing director and the head of claims: manage them closely. The finance controller, head of IT, agency manager and compliance officer have high influence but less day-to-day interest: keep them satisfied, and expect objections. The finance controller's fraud worry and the head of IT's preference for a new system will both matter later.

04Walkthrough

1. Download the dataset below and load it into your tool.
2. Run the two queries above, or build the same in Excel or Power BI.
3. Read all ten interviews. For each, write one line: what they care about, and what they'd need to support a change.
4. Write your scope (in and out) and your definitions.
5. Write the problem statement (the task below).

05Practice

06Check your understanding