Part of our guide hub: Inventory software
Most businesses do not decide to replace their inventory spreadsheet. They discover it has already failed: usually during a stock count that does not reconcile, or when a customer is promised something that turned out to be sitting in another branch, sold last week.
Excel is genuinely good at inventory up to a point. Here is where that point is, what to move to, and the requirements that make or break an implementation in the Pakistani market specifically.
The five signals you have outgrown the spreadsheet
1. More than one person edits it
The moment two people record stock movements, you have a synchronisation problem. Shared cloud sheets soften this but do not solve it: two people editing the same row minutes apart still produces a wrong number, and nobody can tell afterwards which entry was right.
2. You have more than one location
Two branches, or a shop and a warehouse. Now every transfer has to be recorded twice, and the discipline required to always record both sides does not survive a busy week. Multi-location is the most common single trigger for moving to software.
3. Stock counts disagree by more than a rounding error
If physical count and recorded count differ by 5% or more, your numbers are not usable for purchasing decisions. At that point you are buying on instinct while paying for the illusion of a system.
4. You have batches, expiries or serial numbers
Pharmaceuticals, food, cosmetics, electronics with warranties. Tracking batch and expiry in a spreadsheet is theoretically possible and practically hopeless once you have a few hundred SKUs. Expired stock discovered at count is a direct, measurable loss.
5. Someone spends a day a week on stock admin
The salary cost of maintaining the spreadsheet exceeds the software subscription long before most businesses notice. This is the calculation that usually settles the argument.
What to look for, in priority order
Multi-location with transfers as first-class operations
Not “we support branches”: actual transfer documents with dispatch and receipt as separate steps, and visible in-transit stock. Stock that has left branch A and not yet arrived at branch B is a real state and it must exist in the system, or the differences show up as shrinkage.
Units of measure that convert
You buy in cartons, store in cartons, sell in pieces. You buy fabric in rolls and sell in metres. The system must hold the conversion and apply it consistently in purchasing, costing and sales. Systems without proper UoM handling get worked around by creating duplicate SKUs, which destroys your reporting.
Costing method that matches how you actually price
Weighted average, FIFO, or specific batch cost. In a market with meaningful currency movement, the difference between costing methods materially changes reported margin. Pick deliberately, and make sure the system can show landed cost (including customs, freight and clearing), not just invoice price, or your imported-goods margins are fiction.
Barcode support that works on the devices you own
Most businesses do not need dedicated scanners; a phone camera through the app is adequate at moderate volume. What matters is whether your existing barcodes are readable and whether you can generate labels for items that arrive without them.
Reorder points and purchase suggestions
Minimum stock level per item per location, with a purchase suggestion when it is breached. This is the feature that converts inventory software from record-keeping into something that actually saves money, and it is the one most often left unconfigured after go-live.
Integration with how you sell
If you have a retail counter, inventory and POS must be the same stock ledger, not two systems synced nightly. If you sell online, the storefront must decrement the same stock. Our note on why cloud POS is replacing legacy billing covers the multi-branch version of this problem.
The Pakistan-specific requirements
- Tax and invoicing compliance. Sales tax invoice formats, and integration with FBR digital invoicing where it applies to you. Systems built abroad frequently need customisation here, and it is not optional.
- Multi-currency landed cost. If you import, you need purchase in USD and stock valued in PKR at the rate that actually applied, plus duty and clearing distributed across the consignment.
- Credit sales and recovery. A large share of B2B trade runs on credit. Inventory that cannot see customer balances and ageing is only doing half the job.
- Connectivity tolerance. If your warehouse loses internet for two hours, can goods still be received? For anything outside major cities, ask this explicitly.
- Urdu-capable printing. If your delivery challans or labels need Urdu, confirm it renders correctly on your actual printer before purchase, not in a PDF preview.
Cloud or on-premise?
Cloud for almost everyone now: no server to maintain, accessible across branches, backups handled, and the subscription is predictable. The dependency is connectivity, and that is a smaller risk than it was five years ago.
On-premise still makes sense for businesses with genuinely unreliable connectivity, or with data policies that prohibit external hosting. Understand what you are taking on: server hardware, backups that someone actually verifies, and a person responsible when it stops.
The middle path (cloud with offline-capable local operation that syncs when connectivity returns) is the right answer for distribution and warehouse businesses, and it costs more. Decide before you shortlist, because it filters most vendors out.
Why implementations fail
Rarely the software. Almost always one of these:
- Dirty opening data. Duplicate SKUs, five spellings of one supplier, inconsistent units. Migrating this as-is guarantees the new system is wrong on day one and blamed for it by week two. Clean before you migrate; it is the highest-return week available.
- No physical count before go-live. The opening balance must be real. Starting from a spreadsheet figure everyone privately doubts means nobody trusts the system’s numbers afterwards.
- Partial adoption. If some stock movements are recorded and some are not, the data is worse than the spreadsheet was, because it looks authoritative. This has to be all-in from a fixed date.
- Nobody owns it. One named person responsible for stock accuracy, with the authority to insist on process. Without that role, accuracy decays within a quarter.
Build or buy?
Buy, unless your operation is genuinely unusual: manufacturing with multi-level bills of materials, complex consignment arrangements, or an industry-specific traceability requirement. Inventory is a well-solved problem, and a configured off-the-shelf system beats a mediocre custom build in almost every case. The threshold is covered properly in custom software versus off-the-shelf, and if you are weighing a full ERP instead, what an ERP really costs is the piece to read first.
Frequently asked questions
When should a business move from Excel to inventory software?
When more than one person records stock movements, when you have more than one location, when physical counts differ from records by more than about 5%, or when you track batches and expiries. Any one of those is enough.
How much does inventory management software cost in Pakistan?
Cloud systems typically run 10,000 to 60,000 PKR per month depending on users and locations. Custom builds start considerably higher and are justified mainly by unusual operational requirements rather than by scale alone.
Do I need barcode scanners?
Not initially. Phone-camera scanning through the app handles moderate volume well. Dedicated scanners pay for themselves in high-throughput warehouse receiving, not at a retail counter.
How long does implementation take?
Four to ten weeks for a single or two-location business, longer for multi-branch with integrations. Data cleanup is the largest variable and the part most often underestimated.
Ezitech builds and implements inventory, POS and ERP platforms for retail, distribution and manufacturing businesses across 10+ industries. See our inventory management system or tell us how many locations you run.