What makes inventory software difficult to adopt?

Ask a small business why it stopped using its inventory app and the answer is rarely a missing feature. The numbers stopped matching the shelf, so people went back to checking the shelf.

In brief

• Inventory software is hard to adopt because it front-loads the cost: a complete item list, a counted opening stock and a record of every movement, before it gives anything back.
• Stock records drift even in large companies. In a study of close to 370,000 records at a large retailer, 65% did not match the physical count.
• The exceptions, such as returns, damage, samples and transfers between branches, are daily work, and most systems make them slow to record.
• Adoption gets easier when software starts from the existing spreadsheet, counts stock in parts, and takes movements from the bills and challans a business already makes.

That sequence is the adoption problem in miniature. In why SMBs still depend on spreadsheets for inventory, we argued that spreadsheets win because they bend to the business. This note looks at the other side: what inventory software demands before it gives anything back, and why those demands are so hard for a small business to meet.

Why is inventory software hard to adopt?

Because it front-loads the cost. Before its first report is useful, the system needs three things most small businesses do not have in one place: a complete list of items, a counted opening stock, and the habit of recording every movement. Each is reasonable on its own. Together they amount to a project, and a small business has nobody whose job is to run it.

Plenty of businesses never start. In a US survey by Wasp Barcode, a company that sells inventory systems, 43% of small business owners said they tracked inventory with pen and paper or spreadsheets, or did not track it at all (Wasp Barcode). The businesses that do start run into the same three demands, in order.

The item list is never finished

Software needs an item master: one row per product, with a unit, a code and, for GST invoicing in India, often an HSN code as well. Real stock does not arrive that tidy.

The same product is bought by the box and sold by the piece. A supplier renames an item. Sizes and colours turn one product into twelve. Loose goods are sold by weight. Each of these needs a decision about how to model it, and the decisions have to be made before the software can be used at all. In a spreadsheet, nobody has to decide anything: a new row appears when a new item does.

The opening count stops the business

A system that works stock out from movements needs a starting point. That means counting everything, usually on a day the business is closed or slow, and entering it before trading resumes. For a shop or a small distributor with a few thousand items, it is one of the largest single pieces of work the business will do all year, and it has to be done before the software has shown it is worth doing.

The count also has a short shelf life. It is only correct until the next movement nobody recorded.

One missed entry is enough to break trust

Stock on hand is arithmetic: opening stock, plus receipts, minus issues, plus or minus adjustments. If any movement goes unrecorded, the number is wrong, and nothing in the system says which movement was missed.

This is not only a small-business weakness. When Nicole DeHoratius and Ananth Raman examined close to 370,000 inventory records across 37 stores of a large retailer, they found that 65% did not match the physical count at the time of audit (Management Science, 2008). That was an established chain running inventory systems at scale.

If records drift even there, software for a small business has to assume its numbers will sometimes be wrong, and make wrong numbers easy to notice and cheap to correct. Most systems assume the opposite: that a record is right until someone proves otherwise.

The exceptions are the job

Inventory software is usually designed around the clean path: goods come in on a purchase bill and go out on a sale. A small business spends much of its day off that path.

Indian GST rules give a useful list of what else happens. Rule 56(2) of the CGST Rules requires most registered businesses to keep stock accounts that record the opening balance, receipts and supplies, and goods “lost, stolen, destroyed, written off or disposed of by way of gift or free sample”, along with scrap and wastage (CBIC).

To that list, add returns from customers, stock moved between branches, goods sold before the purchase bill has been entered, and items borrowed from the shop next door to complete an order. Each is ordinary. In many systems, each needs a separate form, a permission, or a workaround such as a sale that is blocked because the system believes the stock is not there.

When recording an exception takes longer than handling it, people stop recording it. That is where the drift begins.

Nobody owns the numbers

Larger companies have a storekeeper or an inventory controller. In a small business, stock is handled by whoever is free: the owner, a counter salesperson, a helper in the godown. Recording movements is a second job for all of them, and it competes with the first job, which is serving the customer in front of them.

Software priced and designed around named users with defined roles quietly assumes an organisation the business does not have.

How adoption unravels

Put these together and adoption tends to fail in a predictable order:

1. A few movements go unrecorded, usually the exceptions.
2. The stock figure on screen drifts away from the shelf.
3. Staff learn to check the shelf instead of the screen.
4. With nobody relying on the numbers, fewer movements get recorded.
5. The drift gets worse, and the spreadsheet comes back.

This is our reading of the pattern, not a measured result. It does explain why abandonment rarely shows up as a complaint about the software. The software still runs. Its data stopped being worth trusting.

What makes inventory software easier to adopt?

If the cost of adoption sits at the start, the design job is to move it later and spread it thin. From what we have studied so far, that points to six commitments.

1. Start from the file that exists. Import the current spreadsheet as it is, odd columns and notes included, instead of asking for a clean item master first.
2. Count in parts. Begin with the items that sell most, then count the rest shelf by shelf over the following weeks. This is cycle counting, and it replaces the single stock-take day.
3. Capture movement from work already done. Sales bills, purchase bills and delivery challans already describe most movements. Stock should follow from them, not from a separate entry.
4. Accept first, reconcile later. Let a sale go through when the system thinks stock is zero, flag it, and resolve it afterwards.
5. Make exceptions one step. Damaged, returned, sample, moved to branch: each should take seconds, from a phone.
6. Show confidence, not just quantity. Tell the user when an item was last counted, so a doubtful number looks doubtful.

What this means for the lab

Read together with the spreadsheet note, the brief for inventory software at SETITUP LABS is getting sharper: keep the spreadsheet’s tolerance for messy, partial entries, take the reconciliation work away from the operator, and never make a business stop trading in order to start. The problems reach the lab largely through SETITUP, which builds custom systems for businesses, and inventory is one of the territories we are studying most closely.

Frequently asked questions

Why do small businesses stop using inventory software?

Usually because the stock figures stop matching the shelf. A few unrecorded movements, often returns, damage or transfers, make the numbers drift. Staff then check the shelf instead of the screen, fewer entries get made, and the software is dropped because its data is no longer trusted, not because a feature is missing.

Is a spreadsheet good enough for inventory management?

For a single location with one person updating stock, often yes. Spreadsheets break down when several people edit the same file, when a second location opens, or when you need to know when and why a quantity changed. At that point you need a movement history, not just a current count.

What is cycle counting?

Cycle counting means counting a small part of your stock on a rolling schedule, for example one shelf or category a day, instead of counting everything at once. It keeps records accurate without closing the business for a full stock take, and it is a practical way to build an opening stock gradually.

Do GST rules require businesses in India to keep stock records?

For most registered businesses, yes. Rule 56(2) of the CGST Rules requires registered persons, other than composition taxpayers under section 10, to keep stock accounts covering the opening balance, receipts, supplies, goods lost, stolen, destroyed, written off or given away as gifts or free samples, and the closing balance. This is a summary, not tax advice.

How accurate are inventory records in practice?

Less accurate than most people assume. A study published in Management Science in 2008 examined close to 370,000 inventory records at a large retailer and found that 65% did not match the physical count at audit. Software should expect records to drift and make the drift easy to find and fix.

Sources

1. Nicole DeHoratius and Ananth Raman, “Inventory Record Inaccuracy: An Empirical Analysis”, Management Science 54(4): 627–641, April 2008.
2. Central Board of Indirect Taxes and Customs, “Rule 56: Maintenance of accounts by registered persons”, CGST Rules, 2017.
3. Wasp Barcode Technologies, “6 Inventory Management Best Practices”, citing its State of Small Business Report, a survey of US small businesses.
4. SETITUP LABS, “Why SMBs still depend on spreadsheets for inventory”, 18 June 2026.