Fix the cost data, and the commercial questions become answerable.

The challenge

Fast-growing consumer brands outrun their cost systems in a predictable order. Sales scales first. Fulfillment gets outsourced to a 3PL, then a second one. Component sourcing widens. Somewhere in there the finance function stops being able to answer what a finished good actually costs, and starts answering what it approximately costs, which is a different number that nobody labels as different.

Three specific failures compounded here. Receiving lagged because the process required an invoice before inventory could be received into the ERP, so the ERP and the 3PL databases disagreed by days, permanently. Component costs were not fully burdened, meaning freight and vendor fees sat somewhere other than the item they belonged to. And when the packaging supplier's fees were questioned, nobody in the building could break out what was being charged or why.

Purchase orders are legally enforceable contracts. They were being treated as suggestions, and the variance was landing quietly in an expense account.

The work

Receiving rebuilt against source documents

Inventory now receives at the time of delivery, evidenced by source documentation. That single change closed the standing gap between ERP and 3PL inventory positions and made everything downstream measurable.

The variance account turned into an instrument

Purchase price variance had been an expense line people booked to and nobody read. Activated deliberately, it became the measurement device, capturing per-unit cost variance and unrealized unburdened cost as it happened rather than at year end.

The variance disaggregated by vendor

Three months of PPV were broken out by supplier and root-caused into a named list of who, how much, and why.

Component costs mined from the source

Two quarters of the packaging supplier's bills were reconstructed line by line to establish what was actually being paid and what it was attributable to. That produced, for the first time, a fully burdened view of component cost that could be handed to purchasing as negotiation leverage.

Kitting labor costed into the BOM

Items built through third-party kitting were brought onto the same footing as items received from component suppliers, with labor cost established per touch and built into the bill of materials. Pre-build and on-demand kitting had been invisible in unit cost; they stopped being invisible.

The ERP made the single source of truth

SKU categorization was rebuilt to align with how sales and FP&A actually segment the business rather than how the system happened to be configured, so item management, inventory, and cost modeling finally read from one place. Cycle counting was stood up across both 3PLs, and the receiving process was documented to 22 pages so it would survive staff turnover.

The result

Individual vendor amounts are gross and do not sum to the total, which nets against credits and reversals elsewhere in the period. Figures cover a three-month window and were produced from transaction-level ERP and vendor bill data.

What the cost data unlocked

Getting to trustworthy bottom-up cost was not the objective. It was the precondition. Once component costs were fully burdened and the item master reflected reality, three things became possible that had not been before.

A customer-level P&L, and a sales team that could read it

The business sold into national retail, including Amazon, Target, Walmart, Lowe's, Tractor Supply, and Walgreens, across roughly 30 to 40 accounts and around $200M of revenue, without a view of gross margin by retailer. Program decisions on assortment and trade spend were being made on topline and instinct.

Built on the bottom-up cost data, the company's first customer-level P&L gave gross profit margin by account. It surfaced roughly $200K of recoverable costs onto the P&L and contributed to a 15% gross margin improvement, and it supported the onboarding of a major new retail account.

The part that made it stick was not the model. It was training the sales team to use it, so account managers could dial in mix at their own retailer rather than waiting for finance to tell them what was working. A margin tool that only finance can operate changes reporting. One the commercial team runs changes behavior.

Logistics renegotiated from a costed position

Third-party logistics contracts were renegotiated using the cost model and volume growth as leverage, reducing logistics costs by 10% while scaling fulfillment capacity to support new retail accounts. Renegotiating freight without knowing your true landed cost is guesswork; with it, the conversation changes.

Systems evaluated against what the business actually needs

An enterprise ERP fit-gap analysis was run across the ERP, MRP, BI, 3PL, and contract manufacturing stack, structured around requirements to operate the business over a two-to-three-year horizon rather than a feature comparison. Delivered with an external systems consultancy, with internal leadership of the process. Interviews spanned retail and DTC sales, customer support, demand planning, purchasing, the 3PL and contract manufacturer liaisons, EDI, accounting, IT, and the executive team, producing gap analysis, requirements, vendor responses, scored fit demonstrations, and a roadmap with investment estimates.

Making it survive the author

Cost fixes decay. Processes that are written down, owned, and understood by the people running them do not. Key processes across the company were documented and mapped: procure-to-pay end to end with explicit ownership at each step across accounting, purchasing, supply chain, demand planning, and logistics; receiving in full; kitting, both pre-build and on demand.

Then the functional leaders were coached to present their own processes to the wider organization. That step is usually skipped, and skipping it is why documentation projects fail. A process a leader has explained in front of their peers is one they own. A process that lives in a consultant's binder is one they tolerate.

Why this generalizes

The instinct in a growing business is to treat purchase price variance as noise: a reconciling number that finance cleans up. It is the difference between what you agreed to pay and what you paid, and in a company with an active sourcing footprint it is one of the few places margin leaks at a rate visible in a single quarter.

Two questions establish whether you have the same problem. Is inbound freight burdened into item cost, or does it sit in a separate expense account? And when a supplier bills above the purchase order, does anything happen? If the answers are "separate account" and "no," the variance is already there. It just has not been named yet.

What it took

The cost and variance work ran roughly four months across seven functions, two third-party logistics providers, and the ERP vendor's technical team. The analysis was the fast part. The durable change was procedural: documented receiving, a standard cost update process that runs, cycle counts that produce actionable numbers, and supplier agreements that treat a purchase order as the contract it is.