Designed for the Edges

All Sectors

• The Problem

  Across retail financial services, inclusive design is frequently treated as an accessibility add-on — adjustments made for edge-case customers after the main product experience has been built. This produces two problems simultaneously: the standard journey remains optimised for a fictional average customer, and the adjusted journey for customers with additional needs becomes a second-class experience. Consumer Duty makes both a compliance issue. The FCA’s vulnerability guidance explicitly states that firms should anticipate the needs of vulnerable customers in the design of products and services, not retrofit accommodations afterward — whether the journey in question is opening a current account, applying for a mortgage, choosing a drawdown option, or notifying a claim.

• The Pattern

  The structural move is to test customer journeys against edge-case users and extreme scenarios systematically — not starting from the edges but stress-testing against them to ensure the design does not break for atypical but foreseeable customers:

  Edge-case stress testing

  Before any journey goes live — whether it is an account application, a mortgage affordability check, a drawdown decision tool, or a claims notification — test it with the customers who are hardest to serve: highest cognitive load, lowest digital literacy, most emotionally compromised, least stable circumstances. The test is not whether the journey works for them with adjustments — it is whether the primary journey works for them without adjustment. Where it breaks, that is a design flaw in the main journey, not an edge-case accommodation need

  Genuine alternative pathways

  Phone, paper, branch, and assisted digital as real equivalents of the primary channel, not degraded fallbacks. Alternative pathways should offer the same functionality, the same information, and produce the same outcomes. If the phone or branch journey cannot do what the digital journey can — open the same product, change the same details, access the same guidance — it is not an alternative, it is a lesser service. This becomes acute where firms are reducing physical channels: if the digital substitute does not match the function being withdrawn, customers reliant on the original are excluded by design

  Brief-stage inclusion

  Inclusive requirements must enter the product development process at the brief stage, not the QA stage. If vulnerability testing only happens after the product is built, the structural decisions have already been made and only surface-level adjustments remain possible. The brief should specify who the edge-case users are — the variable-income borrower, the digitally low-confidence retiree, the recently bereaved customer, the customer with cognitive impairment — and what success looks like for them

• Indicators It's Working
  • Customer understanding scores for complex product features are comparable across customer segments — not materially worse for older, less digitally literate, lower-numeracy, or financially stressed cohorts

  • Alternative pathway usage — phone, branch, paper, assisted digital — is monitored and iterated with the same frequency as the primary digital channel

  • New product briefs include an inclusive design requirement and edge-case testing plan before development begins, with edge-case profiles defined for the specific product context

  • Complaint and query themes related to confusion or inaccessibility decline across all segments, not just those flagged as vulnerable

• In Practice

  • A retail bank, planning a programme of branch closures, originally treated the assessment as a property and footfall exercise: which sites had declining transaction volumes and viable alternatives within a defined radius. Pressed by FCA expectations on vulnerable customers reliant on physical channels, it stress-tested the closure plan against specific edge-case profiles — older customers with no digital banking, customers with cognitive impairment supported by a familiar branch team, small businesses with regular cash deposits, and customers in areas with poor mobile signal. The assessment surfaced communities where the digital and telephone alternatives did not, in practice, replicate what the branch did: cash handling, in-person identity verification for vulnerable customers, and informal support for customers who could not navigate menus or apps. The bank redesigned the programme to retain a smaller core estate, extended Post Office partnership functionality, and built a dedicated phone line with longer opening hours and trained staff for affected customers. Subsequent complaint volumes and access-related queries from the affected communities ran materially below those seen in earlier closure waves where edge-case profiles had not been used. The edge case had revealed what the standard footfall analysis was hiding.

  • A wealth platform reviewed its drawdown decision journey and stress-tested it against a profile drawn from the FCA’s retirement income work: a customer in their late sixties with limited investment experience, modest numeracy confidence, and a defined-contribution pot that represented most of their retirement savings. The journey was built around a cashflow modelling tool that assumed the customer could interpret real-terms returns, sequence-of-returns risk, and probabilistic outcome ranges. Tested with customers matching the profile, the tool produced confident-looking outputs that the customers themselves could not explain back. Many abandoned the journey; others completed it but selected options inconsistent with their stated objectives. The platform redesigned the primary journey to lead with plain-English scenarios anchored in monthly income, surface a guidance pathway and a regulated advice referral at the points where customer comprehension typically broke down, and require an explicit understanding check before any irreversible drawdown action. Completion rates improved across all customer cohorts, not only the lower-confidence cohort the redesign had targeted. The decision quality of the cohort the journey had previously been quietly failing improved most.

• Watch-Outs
  Common failure modes

  The failure mode is treating inclusive design as a checklist — producing a list of accessibility features that satisfies a review process without changing the underlying design logic of the primary journey. A second is treating it as a one-time exercise at product launch rather than an ongoing discipline: journeys that worked inclusively at launch can drift as they are iterated without vulnerability considerations maintained. A third is testing only within the firm’s existing customer base, which self-selects for people who could already navigate the journey — the customers excluded by the design, including those who lapsed, abandoned, or never started, are the very people not available to test it. A fourth, particularly relevant where firms are withdrawing physical or assisted channels, is assuming that the digital substitute is functionally equivalent without testing it against the customers who relied on what is being withdrawn.