Insurance companies run on data: pricing risk, paying claims, keeping customers renewing and collecting premiums. Analysts in insurance work on all of these, and domain projects in this area stand out on a resume.
How insurance makes money, briefly
Customers pay premiums. The company pays claims when insured events happen and covers operating expenses. It stays healthy when premiums (plus investment income) exceed claims and expenses over time.
Core KPIs
Claims (loss) ratio
claims ratio = claims incurred รท premiums earnedThe most watched metric. A rising ratio means the company is paying out more relative to what it earns. Analysts break it down by product, region, channel, customer age and hospital or garage.
Expense ratio and combined ratio
The expense ratio compares operating expenses with premiums. Combined ratio = claims ratio + expense ratio. Above 100% means underwriting itself is losing money.
Renewal (retention) rate
Share of policies due for renewal that actually renew. Losing customers at renewal is expensive because acquiring new ones costs more.
Claim settlement time and ratio
How quickly claims are settled, and what share of claims are paid. Both affect customer trust and regulatory standing.
Premium collection and receivables
For commercial policies, how much billed premium has been collected, and how old the unpaid amounts are (ageing buckets such as 0โ30, 31โ60, 61โ90, 90+ days).
Fraud indicators
Unusual patterns such as claims soon after policy start, repeated claims from the same provider, or amounts near limits.
Typical analyst questions
- Why did the health claims ratio jump this quarter?
- Which customers are least likely to renew, and why?
- Which garages or hospitals have unusually high claim amounts?
- Which commercial accounts have overdue premiums?
A sample analysis approach
For a claims ratio spike:
- Confirm the spike after cleaning (late-reported claims can distort recent months).
- Break the ratio down by product, region, provider and claim type.
- Separate frequency (more claims) from severity (bigger claims).
- Look for concentration: a few providers or a few large claims.
- Recommend actions: provider review, pricing changes, tighter checks.
Project ideas for your portfolio
Our insurance live projects let you work through these exact situations:
- Health claims ratio spike
- Motor policy renewals
- Commercial premium receivables and renewals
- Motor claims performance dashboard in Power BI
See all of them on the insurance projects page.
Skills to show
SQL for joins across policies, claims and payments; cohort and ratio calculations; Power BI dashboards; clear explanation of frequency vs severity.
Practise what you just read
Reading helps; writing the code yourself is what sticks. Start here: