DSO, DPO, WIP, inventory, close time — then EPM vs. ERP↔SaaS integration vs. a finance transformation. Move the dials, spend the change capacity, and see which number moved and which one was load-bearing.
"DSO is a project. DPO is a project. Close time is a project. Pick the platform with the best ROI deck, add up the line items, and the benefits stack." Every metric has an owner, a target and a vendor.
The levers share the same cash cycle, and the tools don't move metrics — they raise the ceiling on policies you already control. How much of each lever you can actually pull depends on your data and process foundation, and on a finance team with a finite capacity to absorb change.
A day of working capital is a one-time stock of cash, not a recurring gain — its recurring value is only the carry (9% cost of capital here). That's why a DSO number looks heroic in a deck and modest in the P&L.
Month 12: the programme loses its sponsor's attention. Either discipline lapses for a few months and resumes, or the finance team reverts to the old habits for good.
Every collections, supplier and inventory lever is realised at 30% + 70% × capability. With today's systems you get roughly half of what the policy promises; integration, planning and process maturity raise the other half. Tools don't pull the lever — they let you pull it harder without breaking something.
Billing errors → disputed invoices → late payment. Aggressive dunning on a dirty ledger chases the wrong customers, so level-3 collections costs sales and discounts before it gains days. Integration and foundation shrink the dispute rate and make the same policy cheaper.
DSO, WIP and inventory days are yours once released. Days stretched from suppliers beyond contract are a loan: suppliers reprice (0.25–1.2% of purchases) and claw back up to half within two years.
Three projects at once cost more than three times one project, because the same people carry them. Wins refill capacity; stalls are the way programmes actually end. Sequencing mostly buys headroom, not extra value — the thing it protects is the run.
EPM's cash value runs through data it doesn't create; transformation pays back mostly by raising the foundation under everything else; integration unlocks the policies you already have, so it's worth more than its own claim. Valuing each in isolation gets all three wrong in different directions.
The score is cumulative net cash over 36 months after programme cost and carry, with borrowed float at 50%. A programme that releases cash in month 30 and costs a stall in month 31 is not a good programme.