The vendor slide with "25 percent efficiency uplift" is not a business case. A CFO needs specific cash lines, sized for your portfolio, with the assumptions written down. In cosmetics, the returns cluster in five places.

1. Regulatory headcount that scales sub-linearly. PIF, CPSR, CPNP/SCPN and MoCRA drafting done from live formulas rather than from scratch. Typical: 30 to 60 percent fewer regulatory hours per SKU per market.
2. Launches on time. Compliance flagged in formulation, not the week before launch. Moving launch-on-time rate from 60 to 90 percent unlocks the revenue that would otherwise slip a quarter.
3. Retired point tools. Standalone PLM, regulatory suite, SDS authoring, DMS, spreadsheet workflows. Typical stack retirement: 50k to 200k GBP/year of licences.
4. Audit and recall cost avoidance. Full lot genealogy and versioned regulatory files make audit prep a query, not a fire drill. One avoided recall usually pays for the platform.
5. Margin visibility. Per-batch cost, per-customer margin, per-SKU yield visible every shift, not every quarter. Enables commercial and scheduling decisions that are otherwise made in the dark.

Illustrative, not a quote. Adjust the assumptions for your portfolio.

• Regulatory savings: 3 FTE at 65k GBP fully-loaded, 40 percent of time recovered = 78k GBP/year.
• Launch acceleration: 15 launches/year, 4 launches recovered from slippage at 120k GBP contribution each = 480k GBP/year.
• Retired point tools: PLM 45k, regulatory 30k, SDS 18k, DMS 12k = 105k GBP/year.
• Audit/recall provision: 60k GBP/year expected value avoidance.
• Margin/yield recovery: 0.5 percent of 40m GBP COGS = 200k GBP/year.

Total annual return: c. 920k GBP. A Worldover Phase 1 implementation and year-one licence sits well under that for this profile, giving a payback inside 12 months and a strong five-year NPV.

Over five years, the like-for-like comparison usually looks like:

• Generic ERP plus point tools: 3 to 6m GBP (ERP licence, SI implementation, PLM, regulatory tool, SDS tool, integration maintenance, internal run team).
• Modern cosmetics-native platform: 1.5 to 3.5m GBP (single licence, vendor-led implementation, retired point tools, smaller run team).

25 to 40 percent TCO saving over five years is the range we see. The gap widens as the portfolio scales, because the point-tool integration tax compounds.

Three rules that keep the business case credible.

1. Show the current-state cost, line by line. Today's licences, today's regulatory hours, today's launch slippage, today's audit prep. Most CFOs are surprised by their own number.
2. Size each return conservatively and separately. No blended "efficiency" numbers. Regulatory hours, launch revenue, retired licences, audit provision, margin recovery each get their own row and their own assumption.
3. Commit the vendor to timeline and total cost in writing. A payback case is only real if the go-live date and the five-year cost are contractual.

• Counting only the licence saving. The regulatory, launch and margin lines are 5 to 10x the licence line.
• Ignoring the internal cost of the current state. The team's parallel time on the old stack is real money.
• Treating "AI" as a soft benefit. Willow drafting PIFs is a specific FTE-hour line, not a vibe.
• Sizing to today's portfolio. Five-year TCO must model the next brand, next market and next 100 SKUs.