Digital inventory management is the use of software to keep stock records, so each sale, delivery and correction changes the figure at once. It replaces the stock book, the credit book and the year-end count sheets. The gain is a stock figure you can read at any hour without walking to the shelf.
The switch itself, more than the software, decides whether going digital works. This guide shows what changes record by record, what the move costs in hours and materials, and the order to do it in. It ends with the switch worked for one hardware shop in Mbarara, in Uganda shillings.
What changes when stock moves from paper to a system?
Each paper record gets a replacement that updates itself, except the count, which stays. The USAID Logistics Handbook sets out what any stock record must hold: the stock on hand, the losses and the adjustments, product by product. Paper and software both hold them; software keeps them current.
| Paper record | What it does on paper | What replaces it | What changes for you |
|---|---|---|---|
| Stock book or stock cards | A running balance per product, added up by hand | Each product's stock per warehouse, moved by every sale, purchase, return and adjustment | The balance is current after each sale, with no adding up |
| Delivery notes in a file | Proof of what arrived | A purchase saved as received, which adds the stock | Stock rises when the delivery is entered, not when someone remembers |
| Credit book | Who owes what | Customer records with a balance and a credit limit | The till can refuse a sale over a customer's limit |
| Receipt book and daily sales sheet | The day's takings | Sales and payment reports, split by method | Cash in the drawer is checked against cash taken |
| Year-end count sheets | Once a year, the only figure you trust | A count sheet printed from the system, then a correction for each difference | Counting becomes a check you run on one shelf a week |
| Reordering by eye | Someone notices an empty shelf | An alert level per product, and a list of what is low | You still decide what to order and from whom |
Two things don't change. Goods still have to be counted, because records drift with theft, breakage and wrong entries. Someone still decides what to buy: the system raises the alert, and the owner places the order.
The Handbook also states the trade plainly. An automated system can reduce arithmetic errors and flag missing information, but it needs hardware, electricity and training. Its gains must be weighed against those costs, which is the next question.
What does it cost to go digital?
The biggest cost of going digital is time, not software: listing products, counting them and labelling them. The rest is a plan, a device you may already own and some label sheets. Each line below gives what decides the cost and how to work it out for your shop.
| Cost | What decides it | How to work it out |
|---|---|---|
| Software | Products, users and branches on the plan | The Starter plan is free, with no card needed. Your plan sets how many products, users and branches you can have: see pricing, shown in your currency and billed in US dollars |
| Devices | Whether a phone is enough | A phone camera scans barcodes. A USB scanner and a receipt printer are optional; Kenyan prices for both are in inventory management tools for small business |
| Labels | Products with no maker's barcode | Labels needed ÷ 40 per A4 sticker sheet, printed on an ordinary printer |
| Listing | How many products you sell | Products ÷ products typed per hour |
| Counting | Products and people to count them | Products ÷ (pairs counting × products a pair counts per hour) |
| Training | How many staff use it | Staff × days of practice × hours a day |
| Internet | Where and how you sell | Your usual data bundle; check what keeps working when the internet drops before you choose |
Set that against what paper costs you already. A stock book hides its cost: stock it says you hold but don't, found only at the year-end count.
In what order should you switch from paper?
Switch in seven steps around one fixed cut-off: the night you count everything and close the stock book. Everything before the cut-off is preparation. Everything after it happens only in the system.
1. Fix your codes and units
Give every product one code, and use the maker's barcode where the product carries one. Decide the unit you count in and the packs you buy in, such as a carton of 24 pieces.
2. Build the product list
Put every product in one spreadsheet: name, code, cost, selling price, category, unit and an alert level. Leave the quantity empty for now. The columns are set out in how to build an inventory list.
3. Count everything on the cut-off night
Close, count every product, and write each count into the list. Take the count, never the stock book's balance: the book carries every mistake made since the last stock-take. The cut-off and the count sheets are set out in the year-end inventory count.
4. Load products, stock and debts
Load the product list first, then the counted quantities, then each customer's balance from the credit book. Enter what you owe each supplier too, so your first payment settles a real debt.
5. Give each person a login
Each person gets a login and a role before the first day, so every sale and correction carries a name. Train them on a quiet day, as in how to train employees on a new POS system.
6. Go live, and close the book
From the next morning, every sale goes through the system and nothing goes in the old book. A parallel book sounds safe, but every sale must then be written twice, the two soon disagree, and nobody knows which to believe.
Keep the old books. In Uganda, the Tax Procedures Code Act sets a five-year retention period for records, as PwC's Uganda tax summary notes.
7. Label, set alerts, and count again in two weeks
Label what has no barcode, set each product's alert level, and count one category two weeks after go-live. Trace each difference to its cause before you correct it. The reorder point formula sets each alert level from daily sales and the supplier's lead time.
For example: one hardware shop's switch, in hours and shillings
For example, a hardware shop in Mbarara, Uganda, moves 420 products off paper in two weeks, for about 43 person-hours of work. The figures are illustrative, not a real shop's.
It sells nails, paint, plumbing parts and tools from one counter and a store room. Of its 420 products, 260 carry a maker's barcode and 160 don't, such as loose bolts and cut pipe.
| Task | How it is worked out | Person-hours |
|---|---|---|
| Type the product list | 420 products ÷ 60 an hour, by the owner | 7.0 |
| Count on the cut-off night | 420 products ÷ (2 pairs × 50 an hour) = 4.2 hours, for 4 people | 16.8 |
| Load products, stock and customers | Three files | 1.0 |
| Print and stick labels | 160 products × 12 on the shelf = 1,920 labels, or 48 A4 sheets of 40, stuck at 4 a minute | 8.0 |
| Train two staff | 2 people × 5 days × 1 hour | 10.0 |
| Total | 42.8 |
The count needed the owner, two staff and one helper hired for the night. Its 420 products and three users set the plan it needs, which the pricing page shows.
What the cut-off count found
The count found 63 of the 420 products, 15%, different from the stock book. The book overstated the shop's stock by UGX 1,340,000 at cost.
| Result | Products | Value at cost |
|---|---|---|
| Fewer on the shelf than the book said | 41 | 1,870,000 |
| More on the shelf than the book said | 22 | 530,000 |
| Net: the book overstated stock by | 63 | 1,340,000 |
Had the owner typed the book's balances into the system, all 63 products would have started wrong. The first month's reports would then have blamed the new system for the old book's errors.
The credit book held 38 customers owing UGX 2,860,000 between them. Each balance went in as that customer's opening balance, so each first payment reduced a real debt.
Two weeks later
The owner counted the plumbing shelf, 70 products, and found 3 differences. Two were sales rung up under a lookalike product, and one was a delivery saved as ordered instead of received.
Each was corrected with an adjustment and a note. The causes were fixed too: the lookalikes got distinct names, and deliveries are saved as received on arrival.
What common advice about going digital gets wrong
Much of the advice on going digital starts with choosing software. Three of its usual lines cost shops money.
- Choose the software first puts the steps in the wrong order. Codes, the product list and the count decide whether the switch works, and any system loaded with the old book's balances starts wrong.
- Going digital ends counting mixes up adding up with checking. A system removes the arithmetic, not the theft, breakage and wrong entries, so count one shelf a week and fix each cause.
- Buy scanners and RFID tags spends money a small shop rarely needs to. A phone camera reads barcodes, and an ordinary printer makes labels for the rest. When RFID pays, and when it never will, is worked out in RFID vs barcodes.
How to move your shop onto ArmPOS
ArmPOS runs in the browser on a phone, tablet or computer, so the switch starts with a device you own. These steps follow the seven above, each with its guide.
- Make units and categories first. An import finds each product's unit by its name or short name, so the units must exist: Categories, brands and units.
- Import the product list from an XLSX or XLS file of up to 20 MB. Codes must be unique, and nothing is added until the whole file passes: Import products from a spreadsheet.
- Load the count. Products › Opening Stock takes a file of product_code and qty for the warehouse you pick, and adds it to what that warehouse holds.
- Bring in the credit book. The customer file takes each customer's opening_balance: Import customers and suppliers. The supplier file takes no balance, so enter what you owe in each supplier's Opening Balance (Previous Dues): Record a purchase from a supplier.
- Make a role, then a login for each person. Untick All Warehouses to limit someone to their own: Add your team.
- Label what has no barcode, 40 to an A4 sheet or on stickers. For the rest, scan the maker's barcode into Code Product: Print barcode labels.
- Set each product's Stock Alert. Product Quantity Alerts lists what is at or below it: Reports to check every day. Alerts can also reach your phone: Get alerts on your phone.
- Count again in two weeks. Download a count sheet by warehouse and category, then enter each difference as an adjustment with a note: Count and adjust stock.
If the internet drops while the POS is open, it shows Offline mode and keeps the sales on the device until the connection returns. Card and mobile money payments need the internet: The POS screen.
A small maker that also labels materials and finished cartons should read barcodes in manufacturing. A business that turns materials into products, or runs several companies, may need an ERP for its inventory instead.
Frequently asked questions
Is a spreadsheet digital inventory management?
It is a first step: the records are digital, but every sale must still be typed in by hand. A system that sells and keeps stock from the same records stays current without retyping.
How long does it take to switch from paper?
For a shop of a few hundred products, about two weeks, most of it preparation before the cut-off night. The example above took about 43 person-hours for 420 products.
Do I still need to count stock after going digital?
Yes. Records drift with theft, breakage and wrong entries, and a weekly count of one category finds the drift while its cause is still fresh.
Can I start with just a phone?
Yes, with a system that runs in the browser and scans with the camera, as ArmPOS does. Add a USB scanner when the queue at a busy till makes it worth the cost.
What do I do with the old stock book?
Close it on the cut-off night and keep it with your other records. Write nothing in it after go-live, or you will have two figures for the same shelf.
Pick your cut-off night first, then build the product list before it. When the list is ready, create a free ArmPOS account and import one category of products to try the steps above.
