Inventory management for small businesses: when spreadsheets stop working
Stock problems rarely announce themselves. They show up as write-offs, urgent orders and a stock figure nobody quite trusts. Here is what to fix first, and why software comes second.
Stock control is one of those areas where a spreadsheet works well for a surprisingly long time, then stops working without anyone noticing. There is no moment where it breaks. The numbers just drift, and the business slowly starts making decisions from figures that are not quite right.
This post covers what that drift actually costs, the signals worth watching for, and why the first fix is almost always a process change rather than a purchase.
The cost runs in both directions
Inventory problems are usually discussed in terms of running out. That is only half of it, and often the cheaper half.
Running out costs you the sale, sometimes the customer, and the premium you pay to get stock in urgently. It is visible and it gets attention because someone is annoyed.
Holding too much costs you quietly. Cash sits on a shelf instead of in the bank, storage fills up, and some of it eventually expires, obsoletes or gets damaged. Nobody escalates this, because nothing appears to be going wrong.
Businesses that only track stockouts tend to overcorrect into overstocking, because the failure they can see is the one they optimise against. Tracking both is what makes the trade-off visible.
Signs your stock process is failing
- Nobody fully trusts the stock figure. People physically check the shelf before promising anything to a customer. That check is the process telling you it does not believe itself.
- Stock counts produce large surprises. Small variances are normal. Discovering that a line is out by thirty percent means the day to day recording is not working.
- You place urgent orders regularly. Occasional urgency is business. Weekly urgency is a planning failure being paid for at a premium.
- Write-offs are routine. Expiry, obsolescence and damage happening on a schedule means you are buying more than you can move.
- Two systems disagree. The stock sheet says one thing, the accounting system another, and reconciling them is somebody's monthly job.
One of these is worth watching. Three or more means the cost is already material, even if nobody has added it up.
Why spreadsheets fail specifically at stock
Stock is harder for a spreadsheet than most tasks, for a reason that is structural rather than about size.
A spreadsheet records a state. Stock is a continuous flow of events: received, sold, returned, damaged, moved between locations, counted. Recording a flow as a state means every event has to be manually translated into a new number, and any missed event silently corrupts everything after it.
It also fails at concurrency. Two people selling from the same stock at the same time, each updating their own copy, will produce two different truths. That is not a discipline problem, it is what happens when a single number is edited from several places at once. The wider version of this is covered in five signs you have outgrown spreadsheets.
Fix the process before buying software
Inventory software applied to an undisciplined process produces accurate records of the wrong thing. These changes cost nothing and are worth doing regardless of what you buy later.
- Record movements as they happen, not at the end of the day or week. Anything recorded later gets recorded wrong or not at all.
- Give every product one code, used everywhere. Not a description, a code. Descriptions get typed differently by different people and become impossible to match.
- Count a few lines every week rather than everything once a year. Cycle counting finds problems while they are small and traceable, and it does not require closing.
- Decide who is allowed to adjust a figure, and require a reason. Unexplained adjustments are how variances get hidden rather than fixed.
- Set reorder points for your top lines. Even roughly. Most stockouts happen on items whose reorder point was never defined.
Businesses that do only this often find the problem shrinks enough that software becomes a smaller, cheaper project. That is a good outcome.
What inventory software should actually do
If the process is sound and the volume justifies it, the features worth paying for are unglamorous.
- Record every movement with a timestamp and a person. This is the core of it, and everything else follows.
- Handle multiple locations properly, if you have them. Stock in the van is not stock in the warehouse.
- Prompt reordering automatically based on reorder points and lead times.
- Show stock value, not just quantity, so the cash tied up is visible.
- Keep a full history, so you can answer what happened to this item rather than only what the total is now.
Integration matters more than features
The most common failure is buying a capable stock system that does not talk to anything else. Sales come in through one system, stock lives in another, and someone reconciles them manually. That is the same problem you started with, now with a subscription attached.
Before buying, check specifically how stock levels update when a sale happens, whether purchase orders flow into your accounting system, and what happens on a return. If any of those needs a person to retype something, the saving is smaller than it looks.
A stock system that does not connect to how you sell is a more expensive spreadsheet.
This is often where custom work earns its place, not by replacing everything, but by connecting the systems you already have so the numbers stay consistent without anyone retyping them. The trade-offs are covered in custom software vs off the shelf.
Frequently asked questions
The practical trigger is usually when more than one person needs to update stock at the same time, or when you hold stock in more than one location. Both create situations a single shared file cannot handle reliably, regardless of how disciplined the team is.
Not necessarily, but it is the single most effective way to reduce recording errors if you handle a reasonable volume of physical items. Typing product codes by hand is where a large share of stock discrepancies originate. For low volumes or high value items, careful manual recording is often sufficient.
Frequently, yes, and it is often cheaper than replacing what you have. If your sales and accounting systems already work well, adding stock tracking that connects to them is usually a smaller project than moving everything onto a new all in one platform.
Cycle counting a small number of lines every week is more effective than a single annual count. It finds discrepancies while they are recent enough to trace, spreads the workload, and does not require stopping operations. Fast moving and high value lines should be counted most often.