Uncovering £25m of revenue at risk in cruise retail pricing

QuantSpark's pricing-conflict analysis surfaced more than 120,000 individual pricing errors across an international travel retailer's cruise division, quantifying £25m of revenue at risk, around 10% of the division's approximately £320m annual revenue. That is the headline: a margin leak the retailer knew existed but could not see, made visible and actionable through systematic conflict detection.

The division ran more than 100 shops across major cruise lines, carrying roughly 100,000 unique products. Prices had to flex by ship, region and shipping lane, and be converted between four trading currencies. That conversion, and much of the rest of the pricing workflow, ran through manual spreadsheet calculations alongside many other manual steps. When margins began to fall, nobody could pinpoint the cause: at this scale, a pricing error is never one bad number, it is thousands of small misalignments scattered across shops, currencies and SKUs, individually too small to notice and collectively large enough to erode the bottom line.

QuantSpark's approach started with process mapping: tracing the end-to-end pricing workflow, from purchase cost through currency conversion to shelf price, to find where errors were most likely being introduced, rather than starting from the falling-margin symptom and working backwards. That mapping pointed to two distinct fault lines. First, margin conflicts: sale prices that no longer aligned with the target margin once matched against known purchase costs. Second, currency conflicts: prices for the same product that had drifted apart across the four trading currencies, most likely a consequence of manual, spreadsheet-based conversion with no systematic cross-check.

Rather than treat every mismatch as equally urgent, QuantSpark built conflict-detection logic tuned to each fault line and ran it across the full product and currency base, converting a manual check into a systematic one. The analysis flagged margin conflicts on 6,000 items and currency conflicts on a further 114,000, more than 120,000 pricing conflicts in total, rolling up into the £25m revenue-at-risk figure. On a roughly £320m revenue base, that is around one-tenth of annual sales sitting inside prices that were quietly wrong.

Quantifying the problem was only half the job; it then had to become something the buying team could act on immediately. QuantSpark isolated the 300 highest-risk products, the conflicts worth the most in revenue terms, for manual correction, giving the buying team a short, ranked list rather than a 120,000-line output. For the much larger volume of lower-risk conflicts, QuantSpark designed automated price recommendations, so correction did not depend on manual review of every remaining SKU. Finally, the same conflict-detection logic was configured to run on an ongoing basis, giving the retailer standing visibility of pricing accuracy rather than relying on another one-off audit next time margins dipped.

This sequence, map the process, detect by conflict type, quantify and prioritise by revenue at risk, then automate and monitor, is what makes the workflow repeatable rather than a one-time forensic exercise. It also matches the shape of the underlying systems: a multi-currency retail pricing system spanning many shops and SKUs, previously reconciled by spreadsheet, now sitting behind a purpose-built conflict-detection layer that keeps running on an ongoing basis.

The value case has two distinct tiers: delivered and indicative, and they should be kept separate. Delivered value is the £25m of revenue at risk identified and quantified, the 300 highest-risk items handed to the buying team, and the automated detection and monitoring logic now in place. Indicative value sits alongside it: given the division's 50-60% margins, QuantSpark projected that resolving the conflicts could recover around 20% of total margin. That figure is explicitly modelled, a projection based on the analysis rather than a confirmed, realised outcome, and should be read as the scale of the prize rather than a delivered result.

For a business built on volume and thin operational margin, the lesson generalises beyond cruise retail: pricing accuracy at scale cannot be maintained by manual reconciliation once product count, currency count and location count all multiply against each other. Conflict detection, run continuously rather than audited occasionally, is what converts an invisible leak into a managed one.