2020 · United Kingdom · G - Wholesale and retail trade · Account takeover of customer loyalty accounts via credential stuffing
Analysed 2026-11-07 by Hermes
Open this case in the tool →Account takeover of customer loyalty accounts via credential stuffing
Expertise to Execution
Evidence coverage: 2/6 stages, weight 11/60
Decisive stage: entry (clearest evidence) · Alternative: none defensible
(1) Boots held 14.4 million customer accounts whose points balances were directly spendable money (1 point = 1p in store and online), so the standing expectation was customer-authentication protection against the known, industry-wide credential-stuffing threat. (2) In practice criminals replayed credentials leaked at other sites against boots.com; boots.com itself required only an 8-character password with no complexity rules and offered no two-factor authentication, and enough logins succeeded (under 1% of accounts, up to ~150,000) that Boots suspended points payments brand-wide and replaced stolen balances. (3) The gap sits in the material column: MFA, password strength and breached-password checking were standard specialist remedies, named as such by independent security specialists during the incident, and the inspection evidence shows they were simply not realised in the boots.com account architecture; no source documents a strategy, standard or leadership decision behind their absence, so this is a knowing-doing gap of known expertise not executed (Expertise to Execution), not a doctrinal gap. The classification rests on the entry stage, which is also the load-bearing one; the entry vector itself (credential stuffing using third-party leaks) is documented but is not the DV - the missing customer-account protections are. Detection (own IT security spotted the activity), egress (no data taken from Boots systems), recovery (temporary suspension, points restored) and response (direct notification, replacement of points) show no documented gaps; the containment scoping condition is recorded as an inferred, contributing gap. RM22 (Expertise to Execution lens): the identity controls that protect customer accounts - MFA, credential monitoring, adequate password rules - existed as known specialist remedies but were absent from boots.com login, letting stuffed credentials log in as legitimate customers (causal role: Enabling). RM04 (Expertise to Execution lens): all Advantage Card spend-runs went through one shared points-payment capability, and with no account-level fraud blocking in evidence the only way to stop the takeovers was a precautionary shutdown of point payments for every one of the 14.4 million cardholders, so the concentration amplified the impact far beyond the under-1% of accounts attacked (causal role: Amplifying).
RM22 — Authentication / credential exposureRM04 — Control dependency concentrationRow description matches the incident (Advantage Card records targeted, no financial data stolen, points payments suspended); the case file's Which? URL now redirects to the article URL cited here. The dataset's informationisbeautiful summary calls it a records breach, but Boots itself denied any system compromise - the incident was account takeover via credential stuffing, and reporting varies between ~140,000 and ~150,000 affected accounts ('less than 1%' of 14.4m). The absence of 2FA and the weak password rule rest on Pen Test Partners' contemporaneous inspection of boots.com (security-vendor analysis, not a Boots admission); Boots was not known to have confirmed them. Whether Boots reported the incident to the ICO is not on record (Tesco self-reported its parallel incident). Attribution to the group behind the Tesco Clubcard attack is expert speculation only, so the actor stays unattributed. No regulator action, fine or follow-up investigation against Boots was found, and no article documenting the exact date points payments were restored.