A founder asks, “How many months of inventory should we carry?”
There is no serious answer without another dozen questions.
Where is the product made? How long does production take? Is it seasonal or agricultural? How reliable is the supplier? Is the packaging unique? How long is the ocean transit? Can labor become a constraint? How quickly does the SKU sell? How much cash does the company have?
A product made locally within a week should not have the same inventory policy as an ingredient sourced from Peru that requires 30 days of production and 45 days on the water.
Inventory management is not about minimizing a balance-sheet account. It is about having the right product available for credible demand without immobilizing more cash than the business can support.
Stockouts and overstock are not symmetrical
Excess inventory is painful. It consumes cash, storage and management attention and can eventually require discounting or write-offs.
But if the company has adequate liquidity and a credible sales opportunity, I would rather err somewhat toward availability on the products that matter. A missed retail opportunity may not come back. An out-of-stock hero SKU can damage velocity precisely when the brand is trying to prove itself.
That does not justify speculative buying. It means inventory policy should distinguish between proven demand and optimism.
For core, high-velocity products with long replenishment cycles, more safety stock can be rational. For a new SKU with uncertain demand, large packaging MOQs and no committed distribution, the same safety-stock philosophy can be reckless.
Sales forecasts are an input, not the forecast
Salespeople are supposed to be optimistic. Their forecast belongs in the process, but it should not become the purchase order by itself.
A useful demand plan combines:
- historical sell-in and sell-through;
- retailer and distributor information;
- units per store per week;
- launch timing and door count;
- seasonality;
- industry and category data;
- promotional plans;
- sales-team intelligence;
- known supply constraints;
- base, upside and downside scenarios.
Finance should integrate and challenge those inputs with Operations and Sales. It should not invent a forecast in isolation.
A cheaper unit can be a more expensive decision
MOQs create one of the most common inventory traps.
A supplier offers a meaningful unit-cost reduction if the company doubles its order. On the P&L, the lower cost looks like a margin win. On the balance sheet, the company may have just committed another several hundred thousand dollars to six months of inventory.
The decision should compare the unit savings with:
- cash tied up and financing cost;
- expected sell-through;
- storage and handling;
- shelf-life or obsolescence risk;
- packaging changes;
- demand uncertainty;
- alternative uses of the cash.
The lowest unit cost is not automatically the lowest economic cost.
Complexity has an economic return, and it can be negative
The right question is not whether complexity is bad. Some complexity earns its keep. A new pack size may unlock a channel. A retailer-specific format may create incremental distribution. A low-volume SKU may serve a consumer need that keeps shoppers in the brand.
The mistake is assuming every new SKU is incremental. McKinsey has documented CPG cases where SKU proliferation increased changeovers, specialized inputs, packaging requirements and inventory while sales per SKU deteriorated. Circana makes the same point from the commercial side: rationalization should test incrementality, consumer need and operational efficiency rather than simply cutting the lowest-volume item.
That is the finance lens I would use. What incremental gross profit does this complexity create, what cash does it consume, and what else becomes harder because it exists?
Every new SKU creates a small operating system
Innovation matters. Small brands can move faster than global competitors, and retailers often want emerging brands precisely because they bring new products and formats.
But every SKU can create another ingredient, packaging component, MOQ, production run, forecast, quality requirement, slot on the warehouse rack and opportunity to be wrong.
McKinsey has long distinguished “good complexity,” which creates incremental value, from complexity that erodes profit and supply-chain agility. Circana's 2026 work makes a similar point: more SKUs do not automatically mean more growth, and assortment should be evaluated for consumer incrementality as well as raw sales.
Jeff Church offers a memorable operator example from Suja: he says the company launched 275 SKUs over seven years before a board member challenged whether what management considered disruptive innovation had become unnecessary churn.
The lesson is not “stop innovating.” It is make complexity earn its keep.
A profitable SKU can still be a bad SKU
Suppose a product generates $300,000 of annual revenue at a 45% gross margin. That sounds respectable.
Now add a unique ingredient, a large custom-packaging MOQ, occasional production changeovers, six months of inventory and a retailer-specific requirement. The SKU may still be worth carrying. Perhaps it unlocks an account the company could not otherwise win. Perhaps it attracts a distinct consumer who then buys the rest of the portfolio.
Quantify what can be quantified, then use judgment.
This is why SKU rationalization should not be a spreadsheet sorting revenue from highest to lowest. The review should consider sales, margin, velocity, inventory requirements, operational complexity, incrementality and strategic role.
A new SKU generally needs enough time to establish itself. In many categories, a year can be a reasonable evaluation window. But if no retailer wants it and the consumer proposition is clearly not landing, “give it a year” is not a strategy.
Use reorder points as decision points
SKU discontinuation does not always require a dramatic portfolio meeting.
The next packaging or ingredient reorder can force the right question: do we want to commit another cycle of cash to this product?
That is often a better moment to kill a weak SKU than after another six months of inventory arrives.
Likewise, quarter-end is a natural point to confront slow and obsolete inventory. If the product no longer has a credible path to normal sell-through, protecting the historical margin is not a reason to keep paying to store it. Discount wholesalers, targeted promotions, donations or write-offs may all be better than pretending the inventory is still worth its carrying value to the business.
Safety stock should be a policy, not an argument
Operations naturally wants more buffer. Finance naturally sees the cash cost. Re-litigating that tension on every purchase order is inefficient.
Build policies by product or inventory class using sell-through, lead time, replenishment reliability and strategic importance. Then require justification for exceptions.
If a supplier's lead time doubles, change the policy. If velocity collapses, change it. If a product becomes critical to a major retailer, change it.
The point of the rule is not rigidity. It is preventing “just to be safe” from becoming an unlimited purchasing philosophy.
Large sourcing commitments are financial decisions too
Sourcing should identify suppliers, negotiate options and understand quality, availability and operating constraints. But a commitment to buy $2 million of raw material over the next 12 months is not merely a sourcing decision.
It affects cash, margin, inventory risk and the company's ability to fund other priorities. Someone with company-wide financial accountability has to evaluate and approve it. Depending on the organization, that may be the CFO, COO, CEO or a combination.
The reporting line is less important than the decision right.
The CEO should manage exceptions, not inventory
A CEO does not need to spend every week debating weeks of supply.
The finance and operations teams should manage the system. CEO attention is warranted when inventory becomes strategic: a material write-off, a cash constraint, an important retailer at risk, a major stockout, or a sourcing commitment large enough to change the company's options.
If a $15 million CPG company has $4 million of inventory, the number alone tells you almost nothing. What is finished goods versus raw material versus packaging? How quickly does it turn? What are the lead times? Which products are strategic? How much is slow or obsolete? What demand is the inventory protecting?
The objective is not “lean inventory.”
The objective is inventory that earns the cash committed to it.