Getting Started with Zeneva: A Setup Guide

Get to a first sale before you do anything else

The instinct when setting up any new system is to complete the data first and use it second — load the entire catalogue, get everything perfect, then start. That order is why so many setups stall.

Do the opposite. Complete onboarding, add a handful of products, and put a real sale through the POS in the first fifteen minutes. It confirms the thing works on your device, on your network, with your hands, before you have invested an afternoon in data entry. Everything below assumes you have done that first.

---

Step 1: Onboarding questions

On first login you are asked what kind of business you run — pharmacy, supermarket, fashion, electronics, and so on.

This is not a demographic survey. It determines what your dashboard shows by default: a pharmacy gets expiry alerts surfaced, a fashion retailer gets variant movement, a supermarket gets fast-mover reordering. Answering it accurately saves you configuring the same thing manually later.

If your shop genuinely spans two categories, pick the one that carries most of your transaction lines. The panels are adjustable afterwards.

---

Step 2: Loading your products

Two routes, and the right one depends on where your data currently lives.

Manual entry — for small catalogues and unique items

Inventory → Add Item. Photos, SKU, cost price, selling price, quantity, low-stock threshold.

If you have under a hundred products, manual entry is often faster than preparing a spreadsheet for import, and it gets the fields right on the first pass rather than requiring a clean-up afterwards.

CSV import — for anything larger

Download the template, match your existing export to it, upload. Five columns matter more than the rest:

1. Product name — as staff would search for it, not as the manufacturer prints it
2. SKU or barcode — use the manufacturer barcode where one exists
3. Quantity on hand — count it, do not estimate it
4. Cost price — from the supplier invoice
5. Selling price

Everything else can be filled in later. Those five are what margin, stock value and reorder alerts are all calculated from.

The import mistake that costs the most is not a technical one. In a spreadsheet, sorting a single column without selecting the rest shuffles prices and quantities against the wrong products. Nothing errors. The file imports cleanly. The numbers simply belong to different items, and you find out weeks later when a margin figure makes no sense.

Two habits protect against it: keep an untouched copy of the original file, and after import spot-check ten random products against what is physically on the shelf. A row count matching is not verification — it is exactly what a shuffled file also produces. There is more on this in Excel vs a modern POS.

---

Step 3: The point of sale

Three things worth setting up properly on day one.

Barcode scanning. Your phone camera works. A USB or Bluetooth scanner is faster once you have a queue. Start with the camera, buy a scanner when the queue tells you to — and label your top items by transaction count rather than by revenue, since those are the ones that appear on most receipts.

Offline mode — test it deliberately. Turn data off on the device, complete a sale, force-close the app, reopen it, and confirm the sale survived. Then turn data back on and confirm it synced. This takes two minutes and is the only way to know the behaviour before you need it during an outage rather than after.

Customer phone numbers. Ask at the counter. This is what builds a customer record you can act on later — see understanding your customers with CRM — and it costs nothing while you are already taking payment.

---

Step 4: The storefront, when your counts are trustworthy

Storefront Settings generates a public URL that shares stock with the counter, so selling the last bag of rice in-store immediately shows it as out of stock online.

That shared stock is the reason to hold off publishing until your counts are accurate. A storefront running on unreliable numbers takes orders for things you do not have, which is a worse customer experience than not having a storefront at all. Two weeks of trustworthy counts first, then publish. The details are in the storefront guide.

---

Week one: what to check

Set aside twenty minutes at the end of the first week.

• Reconcile daily takings against cash, card and transfer totals. Do this every evening for the first fortnight — long enough to tell whether the gaps are ordinary noise or a pattern.
• Check ten products against the shelf. Discrepancies this early are almost always import artefacts, and they are far easier to fix now than after a month of transactions has built on top of them.
• Look at what you sold, not just how much. The first week of real sales data is what you set reorder points from in week two.

---

Week two: reorder points and staff logins

Two settings that need a week of real data behind them.

Reorder points should be calculated, not guessed: daily sales × supplier lead time in days, plus a buffer. A product selling 10 a day from a 3-day supplier needs the alert at roughly 35, not at 5. One threshold for the whole shop is wrong for nearly everything in it. See advanced inventory tips for the fuller version.

Staff logins, one per person. This is not optional if you ever want to know who voided a sale, who applied a discount, or whose shift the shortfall happened on. A shared login makes every accountability report in the system unreadable, and shops discover this only when something goes wrong. It takes half an hour.

---

Common stalls, and what unblocks them

Stall	Actual cause	Fix
"Still not finished loading products"	Trying to load everything before selling	Load the top 30 and start
"The numbers look wrong"	Import shuffled, or counts estimated	Spot-check 10 items; recount rather than adjust blindly
"Staff are not using it"	Slower than the old way at first	Barcode the fast movers; the speed is the argument
"Margins make no sense"	Cost prices missing or guessed	Enter from invoices at intake, going forward
"Nobody knows who did what"	Shared login	One login per person

---

Then leave it alone for a month

The most useful thing you can do after week two is stop configuring and start using. Reorder points, categories and thresholds all improve with real data behind them, and real data only accumulates by trading.

Once you have a month of it, the Zen AI Copilot has enough history to say something useful about which products are tying up capital. Before that, it is working from a sample too small to draw from — which is true of any analysis, automated or not.

Beyond operations, if you are looking for capital to scale, Zeneva maintains a business grants directory of verified, active schemes for Nigerian SMEs, and the pricing page sets out what each plan includes.