Inventory is cash frozen on a shelf. The goal isn’t maximum stock — it’s the right stock, with minimum capital tied up. Most East African retailers operate at the wrong end of this trade-off: too much of what doesn’t sell, too little of what does, and a stocktake every December that closes the shop for two days and produces numbers nobody trusts. This guide covers the four levers that actually matter, the seasonal planning calendar that gets you ahead of Ramadan and Christmas without panic-buying, and the day-to-day discipline that quietly turns a 3%-of-revenue shrinkage problem into a 0.5% rounding error. Built around NinoPOS patterns; the principles apply to any modern POS.
The four levers
Get these four right and the rest follows. Get any of them wrong and you’ll feel it within a quarter.
1. Reorder points
For each SKU: (average daily usage × lead time in days) + safety stock. Most retailers either don’t set them or set them once and never revise. Both are wrong. Average daily usage drifts with seasons; lead times stretch during peak; safety-stock forecasts get noisier as your customer mix changes. Revisit reorder points quarterly, with a fast-track review for any A-class SKU whose usage shifts more than 30%.
2. ABC classification
Not every SKU deserves the same attention. Sort by annual revenue contribution:
| Class | % of SKUs | % of revenue | Treatment |
|---|---|---|---|
| A | ~20% | ~70-80% | Tight stock control, monthly cycle count, dedicated reorder-point review |
| B | ~30% | ~15-20% | Standard reorder logic, quarterly cycle count |
| C | ~50% | ~5-10% | Loose limits, semi-annual count, delist candidates |
The C-class is where most stockrooms hide their dead capital. A quarterly C-class purge — markdown anything that hasn’t sold in 90 days — usually frees more cash than any pricing tweak.
3. Cycle counting (instead of annual stocktake)
Annual stocktakes are theatre. They close the shop for two days, produce variance numbers nobody can act on, and tell you about problems six months after they happened. Cycle counting replaces them: count 5–10% of SKUs per week on a rotating schedule, prioritised by ABC class. Variance per cycle is small enough to investigate immediately. Year-end becomes verification, not excavation.
4. Bank-to-inventory reconciliation
Every month, compare cost of goods sold (from NinoPOS Inventory) to bank-side supplier payments. Persistent gaps mean one of three things: theft, miscounted receiving, or supplier short-shipping. None are fun, all are findable if you check monthly. Catch them at month-end, not at year-end.
The seasonal planning calendar
East African retail has predictable peaks: Ramadan, Eid, back-to-school in January, mid-year tax-related buying, Christmas, election cycles in Kenya and Tanzania. None of these are surprises. Most retailers still get caught out because they treat seasonal planning as marketing’s problem, not operations’.
| Weeks before peak | Action |
|---|---|
| 10 weeks before | Pull last year’s POS sales for the same window. Adjust for any structural change (new products, lost suppliers, new competitor down the road). |
| 8 weeks before | Calculate forecast volume per SKU. Identify the 20 fast-movers that will drive 70% of seasonal revenue. Confirm supplier capacity for each. |
| 6 weeks before | Place purchase orders. This is non-negotiable. Local suppliers run out during peak; import lead times stretch. Don’t wait for the rush. |
| 4 weeks before | Set up the seasonal display. Train staff on the seasonal SKU prices and modifiers. Print signage. |
| 2 weeks before | Top up safety stock on fast-movers. Verify reorder points have been temporarily raised. |
| 1 week before | Final inventory count on the top 20 SKUs. Lock down till PINs to prevent staff price overrides during chaos. |
| Peak week | Cycle count daily on top 5 SKUs. Watch the dashboard. |
| 1 week after | Compare actual vs forecast. Document gaps for next year’s plan. |
East Africa-specific challenges
The textbook framework above is universal. Three things in East African retail bend it:
- Imports run on long lead times. A 60-day port-to-shelf cycle isn’t unusual. Safety stock for imports has to absorb 14–21 days of forecast error, not the 2–7 you’d use for local supply.
- Mobile money muddies receiving. Suppliers paid by M-Pesa often don’t produce a formal invoice. Without a paper trail, your inventory cost basis drifts. Force every supplier payment through the Payments module with a receiving record attached.
- Power and connectivity outages distort your sell-through data. A till that was offline for three hours on a Saturday afternoon will show those hours as zero sales. NinoPOS’s offline mode handles this — see our offline-first POS guide for the mechanics.
How NinoPOS Inventory ties this together
The NinoPOS Inventory module exposes the four levers as one workflow. Reorder points are per-SKU, per-location; ABC classification runs automatically off rolling 90-day sales; cycle counts schedule themselves on a class-aware rotation; and the dashboard surfaces variance trends so you spot drift before it costs you a quarter.
For multi-store retailers, the same logic runs across every location with consolidated reports. For restaurants, recipe-level deduction means a sold burger automatically deducts the bun, patty, sauce — see our restaurant POS landing page for the recipe engine.
Common inventory mistakes
- Counting once a year. The drift accumulates invisibly for 11 months and surfaces all at once. Cycle counting beats it every time.
- Treating all SKUs equally. Without ABC classification, you spend the same energy managing a fast-moving sugar SKU and a niche hardware tool. The fast-mover suffers; the niche tool gets attention it doesn’t deserve.
- Ordering for season at 2 weeks’ notice. Suppliers run out. Lock in 6 weeks before, not 2.
- Ignoring the C-class. Dead stock isn’t neutral — it’s capital frozen on a shelf. Mark it down or delist it.
- Not reconciling bank-side to POS-side monthly. Theft and short-shipping are findable. They’re also expensive if found at year-end.
- No FEFO discipline on expiry stock. Newest items at the front mean expired items at the back, in landfill, costing you the full purchase price.
Ready to run inventory like the discipline it is?
- Start a 30-day free trial: Try NinoPOS free
- Compare plans: pricing page
- Want help setting up reorder points and cycle-count schedules for your specific stock? Contact our team.
Frequently Asked Questions
What does "inventory turnover" actually tell me?
How many times in a year your stock cycles through the shop. High turnover = capital working hard, less risk of dead stock. Low turnover = cash frozen on shelves. Calculate as cost of goods sold ÷ average inventory value. For East African retail, 6–10 turns/year is healthy; below 4 is a red flag.
How often should I do a full stocktake?
Don’t. A full annual stocktake closes the shop and produces noisy results. Do cycle counts instead — count 5–10% of SKUs per week on a rotating schedule, prioritised by ABC class. Year-end becomes a verification, not an excavation.
What’s a sensible reorder point formula?
(Average daily usage × lead time in days) + safety stock. For most fast-moving SKUs in Tanzania/Kenya retail, lead time is 5–14 days for local suppliers and 30–60 days for imports. Safety stock = your forecast error in days, usually 2–7 days for local, 14–21 for imports.
How do I handle expiry-dated stock (pharmacy, food, cosmetics)?
Track batch + expiry per SKU, not just SKU-level totals. The POS should sell oldest-first (FEFO — First-Expired-First-Out) automatically. Set alerts at 60 days, 30 days, and 7 days before expiry so you can discount before you write off.
What should I prioritise during Ramadan and Christmas?
Two-week buffer of fast-movers ahead of the season. Lock in supplier orders 6 weeks before, not 2 — local suppliers run out and import lead times stretch during peak. Use last year’s POS data, adjusted for any structural change in your customer mix.
How does ABC classification work in practice?
Sort SKUs by annual revenue contribution. The top 20% (A-class) usually contribute 70–80% of revenue and need tight stock control + frequent counts. The middle 30% (B-class) get standard treatment. The bottom 50% (C-class) get loose limits — sometimes the right answer is to delist them entirely.
Can NinoPOS handle multi-location inventory?
Yes. NinoPOS Inventory tracks stock per store with transfers, location-specific reorder points, and consolidated reports. A sale at any location deducts from that location’s stock; transfers between stores are auditable.
How do I reduce shrinkage?
Cycle counting catches it early; cashier-level role permissions prevent fraudulent voids; matching M-Pesa SMS confirmations to POS sales catches missing-reference leaks. The combined effect typically takes shrinkage from 3% of revenue to under 1% within a quarter.
Should I use a barcode scanner for everything?
For shops above ~200 SKUs, yes. The time saved at the till and the accuracy gain on stocktakes pay back a USD-30 1D scanner in a week. Above ~500 SKUs, a 2D scanner that reads QR codes too is worth the small premium.
When should I delist a slow-moving SKU?
When it occupies more shelf space than its margin justifies. A simple rule: if a SKU hasn’t sold in 90 days and isn’t a strategic must-stock, mark it down 30–50% to clear; if it still doesn’t move in another 30, delist. Holding it is costing you more than the markdown.