Inventory Control: A Step-by-Step Setup Guide

Most small and mid-size distributors do not lose money because they lack an inventory system. They lose it because nobody owns the count, the reorder points are guesses, and the spreadsheet is three days behind reality. This guide walks through the classification, counting cadence, and formulas that

Your sales rep promises a customer forty units. The system says you have forty-three. The picker finds nineteen. Somebody drives to a competitor, buys at retail, and ships it anyway to protect the relationship.

That sequence costs more than the margin on the order. It costs the expedited freight, the picker's afternoon, the controller's reconciliation time, and a customer who now double-checks everything you say. Most companies treat it as a one-off. It is almost never a one-off.

Weak inventory control rarely announces itself as a software problem. It shows up as stockouts on your best movers while cash sits in slow movers nobody has touched in fourteen months. This guide covers how to fix the process first, then the tooling, in an order that actually works for a team of five to fifty people.

What Inventory Control Actually Covers

Inventory control is the set of rules and routines that keep your recorded quantities matching physical reality, and keep those quantities inside a deliberate range. It is narrower than inventory management, which also includes purchasing strategy, vendor terms, and demand planning.

The distinction matters because teams often buy a planning tool while the underlying count is still wrong. Forecasting on bad data produces confident, expensive mistakes.

The Four Numbers You Need Per SKU

Before any system discussion, you should be able to produce these four figures for every item you stock:

On hand — what is physically in your building right now.
Committed — quantities already promised to open orders.
On order — quantities on open purchase orders, with an expected receipt date.
Available to promise — on hand minus committed, plus on-order quantities that land before the ship date.

Most spreadsheet setups track only the first one. That single gap explains the majority of oversell incidents we see when we audit a client's process. If a rep can only see on hand, the rep will sell on hand, committed or not.

Where Spreadsheets Stop Working

A spreadsheet is a perfectly reasonable inventory control tool under specific conditions: one stocking location, under roughly 150 active SKUs, one person entering transactions, and no concurrent order entry.

Break any one of those and the failures start. Two people open the file at once and one set of edits disappears. A warehouse move happens on Tuesday but gets keyed Thursday. Someone sorts a column without selecting the whole range and quantities silently detach from part numbers.

The tell that you have outgrown it is not SKU count. It is when people stop trusting the file and start walking to the rack to check. Once physical verification becomes the default, the spreadsheet has become overhead rather than inventory control.

Build the Count Discipline Before You Buy Software

The annual wall-to-wall count is a financial exercise, not an operational one. It tells you how wrong you were over twelve months; it does not help you ship correctly in March.

Cycle counting replaces it: small, frequent counts of a subset of items, performed during normal operations. Done properly, it produces a running accuracy figure you can actually manage against.

Start With ABC Classification

Not every item deserves equal attention. Rank your SKUs by annual cost of goods sold — unit cost multiplied by annual units moved — then split them:

A items: roughly the top 20% of SKUs driving about 80% of your COGS. Count monthly.
B items: the next 30% of SKUs. Count quarterly.
C items: the remaining long tail. Count twice a year.

Those percentages are conventional starting points, not laws. Pull your own numbers and adjust. A distributor with heavy seasonality may promote items into A class for two quarters and drop them back afterward.

Add a fourth bucket that classification alone misses: anything with chronic discrepancies, high theft exposure, or a short shelf life. Those get counted on the A cadence regardless of dollar value.

A Cycle Counting Cadence That Sticks

Here is a cadence that survives contact with a busy warehouse:

Pick a daily window. Thirty minutes before shipping starts, or right after the last truck leaves. Movement during a count is the primary source of false variances.
Count locations, not part numbers. Counting "bin D-14, everything in it" catches misplaced stock. Counting "SKU 4471" sends someone looking only where the system says it should be.
Blind count. The counter should not see the expected quantity. Showing it produces confirmation, not verification.
Recount every variance once before adjusting, then require a reason code on the adjustment.
Track accuracy weekly as counts matching exactly, divided by counts performed. Not dollar variance — that nets errors against each other and flatters you.

Reason codes are where the real value lives. After eight weeks you will see the pattern: receiving errors, unrecorded damage, two bins holding the same SKU, or units of measure mismatched between purchasing and picking. Each has a different fix, and none of them is "count more often."

The Math That Drives Reordering

Once counts are trustworthy, inventory control becomes arithmetic. These formulas do not require special software — you can run them in the spreadsheet you already have.

Reorder Point and Safety Stock

Reorder point = (average daily usage × lead time in days) + safety stock

Worked example. You sell an average of 12 units a day. Your supplier reliably delivers in 21 days. Your safety stock is 80 units. Your reorder point is (12 × 21) + 80 = 332 units. When available-to-promise crosses 332, you buy.

For safety stock, a practical starting method: safety stock = (peak daily usage − average daily usage) × lead time in days. If your peak day is 20 units against an average of 12, that gives (20 − 12) × 21 = 168 units. Compare that to what you actually carry. The gap between the two numbers is usually the honest size of your service-level risk.

Review reorder points quarterly at minimum. Lead times drift, especially on imported goods, and a reorder point calculated against a lead time that no longer exists is a stockout with a paper trail.

Turnover and Days on Hand

Inventory turnover = annual COGS ÷ average inventory value

Days on hand = 365 ÷ turnover

If your COGS is $4.8 million and average inventory is $800,000, turnover is 6.0 and days on hand is about 61. Whether 61 days is good depends entirely on your category, so benchmark against your own trend first.

For external context, the U.S. Census Bureau publishes monthly inventories-to-sales ratios for manufacturers, wholesalers, and retailers. It is a free, reliable reference point for whether your sector is building or drawing down stock.

Calculate turnover by ABC class, not just company-wide. A blended number hides the real problem, which is almost always C items quietly absorbing working capital while A items run thin.

Moving From Spreadsheets to a System

When you do buy or build, the discipline you established determines whether the implementation works. Software enforces a process; it does not invent one.

The Requirements Checklist

Any serious inventory control system should handle:

Barcode or QR scanning at receiving, putaway, picking, and shipping. If you are assigning your own item identifiers, GS1 US maintains the standards for barcodes and company prefixes.
Location-level quantities, not just warehouse-level totals.
A full transaction log — who moved what, when, and why — with no silent edits.
Native cycle counting with blind counts and reason codes.
Unit-of-measure conversions between purchasing, stocking, and selling.
An API, so your accounting system and online store read one set of quantities.

Off-the-shelf tools cover most of this well. The common reason teams move to custom inventory management software is a workflow the market does not serve: lot and serial traceability, consignment stock at customer sites, kitting with variable components, or rules that differ by branch.

A Realistic 60-Day Rollout

Weeks 1–2: label every bin and rack location, and clean your item master. Delete dead SKUs, merge duplicates, fix units of measure. This is unglamorous and it is the step teams skip and later regret.

Weeks 3–4: load balances for A items only, run them in parallel with the old process, and compare daily. Do not migrate everything at once.

Weeks 5–8: bring in B and C items, train every person who touches stock, and switch off the spreadsheet on a stated date. Leaving both running indefinitely guarantees neither is trusted.

One note on build capacity. If you decide a custom path makes sense, a Mexico-based nearshore team works in U.S. time zones, which means your warehouse supervisor can get on a call at 9 a.m. Central during the two weeks when questions actually matter. That overlap tends to matter more on operational projects than hourly rate does.

Where to Start This Week

Pick your twenty highest-COGS items. Count them blind tomorrow morning. Calculate accuracy as exact matches divided by items counted, and write the number on a whiteboard where the team can see it.

If that figure comes back below 95%, you do not have a forecasting problem or a purchasing problem. You have an inventory control problem, and it is the one worth fixing first, because every downstream decision inherits its error.

Then run turnover by ABC class and reorder points on your A items. Two afternoons of work will tell you more about where your cash is trapped than a quarter of reporting. If the answer is that your process has outgrown what a spreadsheet can enforce, you can get a quote for your project and talk through what a system would need to do differently.