AIDA flashcards
AIDA · Build recall with a focused review session.
32 cards. Read the term and the definition here. The deck at the top of this page is optional.
- Combined ratio
- Combined ratio equals incurred losses plus underwriting expenses divided by earned premium; values below 100% indicate underwriting profit before investment income.
- Loss ratio
- Loss ratio equals incurred losses divided by earned premium; it measures claims cost relative to premium and is a core underwriting profitability indicator.
- Expense ratio
- Expense ratio equals underwriting expenses divided by written premium; it measures the cost of acquiring and servicing policies.
- Earned premium
- Earned premium is the portion of written premium that corresponds to insurance coverage already provided during the accounting period.
- Written premium
- Written premium is the total premium charged for policies issued during a period, regardless of how much coverage has been earned.
- Claims leakage
- Claims leakage is the difference between the amount paid on a claim and the amount that should have been paid under policy terms, including avoidable overpayment.
- Fraud score threshold
- A fraud score threshold is the cutoff above which a claim is referred for investigation; it is calibrated to balance missed fraud against investigation cost.
- Special investigation unit
- A special investigation unit is a claims team that investigates suspected fraud, often triggered by analytics alerts, and coordinates with legal and law enforcement.
- Model drift
- Model drift is the deterioration of a model's predictive performance or calibration over time due to changes in populations, behavior, or data definitions.
- Out-of-time validation
- Out-of-time validation tests a model on data from a later period than the training data to estimate how performance may degrade after deployment.
- Gini coefficient
- The Gini coefficient measures a model's discriminatory power; it is related to the area under the ROC curve and ranges from zero to one in credit-risk style uses.
- Data quality dimensions
- Data quality dimensions include completeness, accuracy, timeliness, consistency, validity, and uniqueness; each dimension can be measured against an agreed acceptance threshold.
- Telematics
- Telematics uses vehicle sensors and communications to capture driving behavior, which insurers may use for usage-based rating, claims reconstruction, and risk monitoring.
- Next-best-action
- Next-best-action analytics ranks customer offers by expected incremental value, combining propensity, value, and eligibility rather than propensity alone.
- Claims triage
- Claims triage uses analytics to route each claim to the appropriate handling path, such as fast track, standard review, or investigation, based on risk and complexity.
- Proxy discrimination
- Proxy discrimination occurs when a model uses a variable correlated with a protected characteristic, producing disparate impact even without using that characteristic directly.
- Data strategy
- A data strategy defines how an organization will acquire, govern, and use data to achieve business goals, including architecture, ownership, quality, and funding.
- Use-case prioritization
- Use-case prioritization ranks analytics initiatives by business value and delivery feasibility so that the first wave balances quick wins and strategic impact.
- ROI formula
- Return on investment equals cumulative benefits minus cumulative costs, divided by cumulative costs; it expresses net value as a percentage of the investment.
- Net present value
- Net present value discounts a project's future net cash flows at the organization's cost of capital and subtracts the initial investment; positive NPV supports approval.
- Payback period
- Payback period is the time required for cumulative benefits to equal cumulative costs; it measures liquidity risk but ignores cash flows after the payback point.
- Benefits realization
- Benefits realization is the process of defining baselines, targets, owners, and measurement cadence for expected benefits, then verifying that realized value matches the business case.
- Data governance
- Data governance assigns decision rights and accountability for data quality, access, privacy, and model risk through roles, policies, and oversight forums.
- Data steward
- A data steward is accountable for a data domain's quality, definitions, and issue resolution, working with business owners and technical teams.
- Executive sponsor
- An executive sponsor provides strategic authority, funding, and escalation for a data initiative, helping remove organizational barriers and align priorities.
- Pilot success criteria
- Pilot success criteria are pre-agreed measurable thresholds for value, adoption, quality, and risk that determine whether an initiative should scale, revise, or stop.
- Change management
- Change management addresses stakeholder incentives, skills, process redesign, and communication to increase adoption of data-driven workflows and sustain new behaviors.
- Build-versus-buy decision
- Build-versus-buy compares internal development against vendor acquisition using total cost, speed, differentiation, control, and integration risk; the choice depends on strategic context.
- Data maturity model
- A data maturity model assesses capabilities such as data quality, governance, analytics, and culture across defined stages, producing a baseline for roadmap priorities.
- OKR
- OKRs pair an ambitious objective with measurable key results; for data strategy they translate outcomes such as faster claims or better retention into tracked targets.
- Balanced scorecard
- The balanced scorecard tracks financial, customer, internal process, and learning perspectives, helping data initiatives show value beyond cost reduction.
- Model risk management
- Model risk management identifies, validates, monitors, and controls risks from models, including poor data, flawed assumptions, misuse, and unclear ownership.