Why External Benchmarking Is Essential for Targeted Support Under PS25/22

The Regulatory Case for External Benchmarking

When the FCA finalised PS25/22 and introduced COBS 9B, it created a regulatory framework that fundamentally changes how firms must think about consumer segmentation. The targeted support regime does not simply require firms to group consumers into segments — it requires them to evidence that those segments are built on objective behavioural characteristics, not internal commercial priorities.

This distinction is critical. A firm that segments consumers based solely on its own book data — product holdings, balances, transaction history — is building segments that reflect what consumers do with that firm. It tells the firm nothing about what those consumers do elsewhere, how they compare to the broader market, or whether the segment boundaries are genuinely driven by consumer need rather than product architecture.

What COBS 9B Actually Requires

Under COBS 9B, firms must demonstrate that their targeted support segments are built using:

• Common characteristics: Observable behavioural traits that group consumers with genuinely shared needs and circumstances
• Excluding characteristics: Evidence-based criteria that distinguish one segment from another, ensuring consumers receive appropriately tailored support
• Better position test: Proof that consumers who receive a ready-made suggestion are, on balance, in a better position than they would have been without it

Each of these requirements implicitly demands a market-wide perspective. You cannot credibly claim that your segments capture “common characteristics” if your evidence base is limited to your own provider data. The FCA will rightly ask: common relative to what?

The Internal Data Trap

Most firms begin their targeted support journey by analysing their existing client data. This is a natural starting point, but it creates a dangerous circularity. Your internal data reflects the outcomes of your own distribution, pricing, and product design decisions. Segments derived from this data will inevitably mirror your existing product architecture rather than genuine consumer need.

Consider a pension provider that segments decumulation customers by pot size alone using internal data. The resulting segments will reflect the provider’s own client mix — shaped by years of marketing, default fund selection, and employer scheme wins. A consumer with a £150,000 pot at Provider A may have fundamentally different needs, risk appetite, and cross-provider behaviour than a consumer with the same pot size at Provider B.

Without external benchmarking data, the firm has no way to know whether its pot-size-based segments genuinely capture distinct consumer needs or simply reflect its own distribution history.

What External Benchmarking Provides

External benchmarking data — particularly cross-provider behavioural data — solves this problem by providing an independent reference frame. When you benchmark your segments against market-wide consumer behaviour, you can:

• Validate segment boundaries: Confirm that the characteristics you use to define segments are genuinely discriminating across the market, not just within your book
• Identify blind spots: Discover consumer behaviours that your internal data cannot capture — such as multi-provider holdings, switching intent, or cross-product interactions
• Quantify “better position”: Measure whether your targeted support suggestions actually improve outcomes relative to market benchmarks, not just relative to doing nothing
• Defend against regulatory challenge: Demonstrate to the FCA that your segmentation methodology incorporates objective, third-party evidence rather than relying exclusively on data you control

The FCA’s Supervisory Approach

The FCA has been clear that it will take a data-driven approach to supervising targeted support compliance. Firms should expect supervisory questions along the following lines:

• How did you determine the characteristics that define each segment?
• What evidence supports the claim that these characteristics are “common” to the consumers in each segment?
• How do you know your excluding characteristics are genuinely differentiating rather than arbitrary?
• What independent data sources did you use to validate your segmentation?
• How do you measure whether consumers are in a “better position” after receiving targeted support?

Firms that can point to cross-provider behavioural benchmarks will be in a significantly stronger position to answer these questions than firms relying solely on internal analytics.

Practical Implementation

External benchmarking does not replace internal data analysis — it complements it. The most defensible approach combines:

1. Internal analysis to understand your existing client base and identify preliminary segment hypotheses
2. External benchmarking to validate those hypotheses against market-wide behaviour and refine segment boundaries
3. Ongoing monitoring using both internal and external data to track whether targeted support is delivering better outcomes over time

This combined approach gives firms both the granularity of their own data and the objectivity of independent market evidence — exactly what the FCA expects to see.