Quality, Forecasting, Inventory, and Supply Planning

8.1 Creating Product Specifications, Quality Standards, and Acceptance Criteria

A product specification is the written definition of what the company intends to sell. It should describe the product in enough detail that a supplier, quality inspector, warehouse partner, retailer, and customer care team can understand what is acceptable and what is not. Without written specifications, every production issue becomes an argument about expectations.

Specifications should cover the finished product, packaging, components, labeling, shipping case, and any critical inputs. For a food product, this may include formula, ingredient suppliers, allergen controls, net weight, sensory profile, nutrition targets, shelf life, lot coding, and packaging seal integrity. For a cosmetic, it may include formula, color, fragrance, viscosity, fill weight, packaging compatibility, pump function, microbiological limits, and stability requirements. For hard goods, it may include materials, dimensions, tolerances, finish, durability, safety features, and fit.

Quality standards define how good the product must be. Acceptance criteria define how the company will decide whether a batch, shipment, or unit passes. A standard might say that the pouch must seal reliably. The acceptance criteria should specify the seal test, sample size, acceptable failure rate, and what happens if the batch fails. The more concrete the criteria, the easier it is to make decisions under pressure.

Founders should distinguish between critical, major, and minor defects. Critical defects: issues that can create safety, legal, regulatory, or serious customer harm. Major defects: issues that make the product unsellable, unusable, misleading, leaking, damaged, or materially below brand promise. Minor defects: cosmetic or low-impact issues that may not affect use but should still be tracked. This hierarchy helps the team avoid treating every problem equally.

Every approved specification should have a version number, approval date, owner, and change history. If a supplier changes a component, artwork, formula, material, fill level, case pack, or pallet configuration, the specification should be updated and reapproved. Quality begins with clarity, and clarity begins with controlled documents.

When possible, maintain a golden sample: the approved physical reference that shows the intended product, packaging, color, finish, fill level, scent, texture, or assembly. Photos help, but physical references are often more persuasive during disputes. Store golden samples carefully and date them by production version.

8.2 Forecasting Demand for a New Product with Limited Data

Forecasting a new CPG product is uncomfortable because the founder has more conviction than data. That does not mean the company should guess. It means the forecast should be built from explicit assumptions, updated frequently, and translated into production and cash decisions.

Start with demand drivers. A direct-to-consumer launch forecast may depend on email list size, expected conversion rate, paid media spend, CAC, site conversion, average order value, reorder timing, and inventory available for promotion. A retail forecast may depend on door count, weekly units per store, promotional calendar, distribution timing, replenishment cadence, and retailer purchase order behavior. An Amazon forecast may depend on listing conversion, review count, ad spend, search rank, stock position, and FBA availability.

Build three scenarios: conservative, base, and upside. The conservative scenario should be plausible, not pessimistic theater. It answers, “What if demand is slower than expected and cash is tighter?” The upside scenario answers, “What if demand works and we need inventory quickly?” The base scenario should reflect the founder’s best current view. These scenarios should drive order quantities, reorder timing, cash needs, and marketing commitments.

Early forecasts should be reviewed weekly after launch. The founder should compare forecasts to actual sales, actual orders, traffic, conversion, wholesale purchase orders, retail sell-through, returns, cancellations, and out-of-stock days. When actual results differ from forecast, the question is not who was wrong. The question is which assumption changed.

Limited data should be supplemented with leading indicators. Email signup quality, preorder behavior, repeat purchase intent, review sentiment, sampling conversion, retailer reorder speed, subscription attach rate, and customer questions can all improve the forecast. The founder should avoid false precision, but also avoid operating blindly. A forecast is not a prediction to defend. It is a decision tool to improve.

The launch calendar should also be reflected in the forecast. Press, influencer drops, email campaigns, retail resets, holidays, founder appearances, and promotions can all create demand spikes. Forecasting without the marketing calendar is incomplete because supply must be ready before demand is created.

8.3 Managing MOQs, Safety Stock, Reorder Points, Stockouts, and Obsolescence

Inventory management is a balancing act between service level and cash risk. Too little inventory creates stockouts, lost sales, unhappy customers, poor marketplace ranking, and damaged retail trust. Too much inventory creates storage cost, expiration risk, discounting, write-offs, and trapped cash. The right answer depends on demand variability, lead time, MOQ, shelf life, margin, and the cost of being wrong.

MOQs often force founders to buy more than they want. A manufacturer may set MOQ by production run, flavor, color, size, packaging artwork, component, or raw material. The founder should understand the real constraint. Sometimes a smaller run is possible if the founder accepts a higher unit cost, standard packaging, fewer variants, or a shared production window. Do not assume the first MOQ is the only possible structure.

Safety stock is extra inventory held to protect against demand spikes or supply delays. It is not a random cushion. A simple early-stage approach is to hold enough safety stock to cover expected sales during supplier delay risk. For example, if the product sells 100 units per week and a supplier delay could add three weeks, the founder may need at least 300 units of safety stock, adjusted for growth and variability.

The reorder point is the inventory level that triggers the next purchase order. It should reflect expected demand during lead time plus safety stock. If demand is 100 units per week, total reorder lead time is eight weeks, and safety stock is 300 units, the reorder point is roughly 1,100 units. This calculation is simple, but many founders miss it because they focus on current inventory, not inventory at the future arrival date.

Stockouts should be managed deliberately. If inventory will run out, decide whether to reduce paid media, prioritize high-margin channels, reserve units for subscriptions, pause wholesale expansion, raise bundle thresholds, or communicate clearly with customers. Obsolescence should also be monitored. Slow-moving SKUs, seasonal packaging, expiring goods, old formulations, and discontinued components should be identified early enough to sell through, bundle, donate, liquidate, or write off with control.

Track weeks of supply by SKU, not only total inventory value. A company can look well stocked in aggregate while the hero SKU is at risk and slow variants are piling up. This is why SKU-level reporting matters from the first production run.

8.4 Quality Control, Defect Management, Batch Records, Holds, and Recalls

Quality control is the process of confirming that the product received matches the approved specification. It should happen at multiple points: supplier qualification, sample approval, pre-production, in-process checks, finished goods inspection, warehouse receiving, and customer feedback review. Waiting until customers complain is not quality control; it is damage control.

For every production run, the company should retain core records. These may include purchase order, approved specification, production date, lot or batch number, component lots, certificates of analysis, inspection results, release approval, deviations, and shipment details. Batch records matter because they connect a product in a customer’s hand to the materials, production run, supplier, and quality decision behind it.

Defects should be logged in a consistent way. Capture SKU, lot code, order number, customer complaint, photo, defect type, severity, quantity affected, channel, root cause, corrective action, and resolution. Over time, patterns matter more than isolated anecdotes. Five complaints about leaking caps may indicate a packaging compatibility issue. A spike in broken units from one warehouse may indicate handling or packing failure.

A quality hold is a decision to prevent inventory from being sold or shipped until an issue is understood. Holds can apply to a lot, batch, SKU, shipment, warehouse location, marketplace inventory pool, or retailer shipment. Founders should not fear holds. A disciplined hold can prevent a larger recall, customer harm, or brand damage. The hold process should state who can place a hold, who can release it, what evidence is required, and how inventory is physically or digitally blocked.

Recalls are serious and should be handled with qualified legal, regulatory, technical, and communications support. The founder should still prepare in advance. At minimum, the company should know how to identify affected lots, locate inventory by channel, notify partners, stop shipments, communicate with customers, process returns or replacements, and document corrective actions. The best recall plan is one you hope never to use but can activate quickly if needed.

After any meaningful quality issue, require a corrective action. The answer may be supplier retraining, revised packaging, tighter inbound inspection, new transit tests, revised specifications, or a supplier change. A defect that is only refunded, not fixed, will usually return.

8.5 Checklist: Lightweight S&OP Process for an Early-Stage CPG Company

Sales and operations planning, usually called S&OP, is the recurring process that connects demand, supply, inventory, cash, and decisions. Large companies can make S&OP bureaucratic. Startups need the opposite: a short, disciplined meeting that forces the team to see the same facts and make trade-offs before problems become emergencies.

In an early-stage company, S&OP can be a weekly or biweekly 45-minute meeting. The founder, operations lead, growth lead, finance lead, and sales lead should review demand, inventory, production, cash, and risks. The meeting should produce decisions, not just updates. Examples include whether to place a purchase order, reduce paid spend, prioritize one channel, approve air freight, delay a SKU, or reserve inventory for a key account.

Lightweight S&OP Checklist

• Demand review: Compare actual sales, orders, sell-through, traffic, conversion, subscriptions, retail reorders, and marketplace velocity against forecast.
• Inventory review: Check on-hand units, available-to-sell inventory, committed inventory, inventory in transit, inventory on hold, and weeks of supply by SKU.
• Supply review: Review open purchase orders, supplier lead times, production slots, component risks, quality issues, inbound freight, and warehouse receiving dates.
• Cash review: Confirm upcoming supplier deposits, freight bills, duty payments, storage costs, inventory financing needs, and expected receipts from channels.
• Risk review: Identify potential stockouts, excess inventory, expiring goods, retail penalties, packaging delays, supplier failures, and customer-impacting quality issues.
• Decision log: Record decisions, owners, due dates, assumptions, and triggers for revisiting the decision.

The founder should keep the process lightweight but non-negotiable. A simple forecast, inventory report, purchase order tracker, cash view, and risk list are enough to start. The discipline is what matters. In CPG, the company wins when demand creation, product quality, supply reliability, and cash planning move together.