Methodology

Behavioural Data from Real Financial Decisions

Wealth Intelligence data is captured from consumer financial modelling environments operated across the TFE Group network. These are real-world planning tools used by UK households to make genuine financial decisions — including ISA provider selection, pension drawdown strategies, protection calculations, and investment allocations.

Every data point represents an actual decision event: a user comparing providers, modelling a withdrawal strategy, or evaluating a product. This is not survey data, panel data, or scraped data. It is first-party behavioural intelligence captured at the point of decision.

Sample Composition and Coverage

Our dataset currently tracks thousands of UK households actively modelling financial decisions each month, across 50+ financial providers. The sample spans multiple product categories including ISAs, pensions, protection, and investments.

Behavioural modelling datasets do not require the same sample sizes as opinion surveys. Because we capture actual decision events rather than stated preferences, smaller samples produce statistically robust signals. Each observation represents a real financial planning interaction with multiple data points captured per session.

Coverage Areas

• ISA provider selection, deposit amounts, and wrapper-type preferences
• Pension drawdown strategies, pot sizes, and withdrawal timing
• Cross-wrapper capital allocation between savings and retirement products
• Protection calculations, income cover, and risk modelling
• Investment platform selection and portfolio construction signals
• Switching intent and provider comparison behaviour

Addressing Sampling Bias

Any behavioural dataset carries potential biases. We take active steps to identify and mitigate these where possible, and are transparent about known limitations.

What We Do

• Track demographic distribution (age, income, region) and report any skews in our data
• Publish confidence indicators alongside key metrics so readers understand signal strength
• Report month-on-month deltas rather than relying on absolute figures alone
• Include sample sizes at the metric level where relevant
• Flag any metrics where sample sizes fall below reporting thresholds

Known Considerations

Our sample is drawn from digitally engaged consumers using online financial planning tools. This skews towards users who are actively planning their finances, which may differ from the broader UK population. We consider this a feature rather than a limitation: our data captures the behaviour of people who are actively making financial decisions, which is precisely the audience most relevant to our institutional clients.

What the Data Contains

Each monthly intelligence report is built from aggregated behavioural events across the reporting period. The underlying dataset is structured around decision events, each capturing multiple dimensions of a single financial planning interaction.

Aggregated Behavioural Events

• Provider selection and preference ranking
• Capital allocation decisions and deposit sizing
• Withdrawal timing and drawdown strategy modelling
• Switching intent and transfer signals
• Product wrapper preferences and cross-product flows
• Demographic segmentation (age band, income band, region)

Derived Intelligence

• Provider market share rankings with month-on-month momentum
• Behavioural persona clusters with commercial action recommendations
• Accumulation and decumulation index scores
• AUM leakage signals and acquisition opportunity indicators
• Competitive switching heat maps and provider comparison flows

Data Privacy and Anonymisation

No personally identifiable information (PII) is captured at any stage of data collection. All data is anonymised at source — our consumer planning tools do not require registration or personal details to use.

The data we collect is purely behavioural: what people do, not who they are. We capture choices, preferences, and modelling parameters, never names, email addresses, or account numbers.

Data Quality and Validation

Before any data enters our reporting pipeline, it undergoes automated quality checks to ensure integrity and consistency.

• Outlier detection to filter anomalous or bot-generated interactions
• Session completeness checks to ensure meaningful engagement
• Temporal consistency validation across reporting periods
• Cross-metric coherence checks to identify data pipeline issues
• Confidence scoring applied to all published metrics