Using Sales Data to Plan Production Instead of Guessing

Production planning in many small manufacturing businesses works like this: the owner asks the sales manager what is selling, the sales manager says everything, the production head makes what is easiest, and three weeks later one SKU is out of stock while another fills the godown. The cure is not an expensive forecasting system. It is a simple, repeatable plan built from data you already collect in gamitrack. This guide shows you how.

The three numbers behind every production decision

For each SKU you need three things:

1. Demand rate: how many units move per day or week.
2. Stock cover: how many days current stock will last at that rate.
3. Lead time: how many days a production run takes to become sellable stock at the branch.

If cover is shorter than lead time plus a buffer, you are already late. If cover is several multiples of lead time, you are overproducing.

Step 1: Get demand from orders, not from opinions

Open Orders and look at order lines by SKU across the last eight to twelve weeks. Separate out:

• unusual bulk orders (one-off, not repeating),
• orders affected by a scheme or a price change,
• weeks disturbed by holidays or road closures.

What remains is baseline demand. A simple average per week is fine to start. Look at the trend as well: three weeks of steady increase deserves a different reaction from a single spike.

Step 2: Use sell-out where you have it

If you sell through dealers, order data shows sell-in, which can be inflated when dealers stock up before a scheme or price rise. Check Dealer sales for sell-out, meaning what dealers actually supplied to shops. When sell-in runs well ahead of sell-out for several weeks, dealers are building stock, and a slowdown in orders is coming. Combine this with Dealer stock, where monthly counts show expected closing stock against counted stock, to see how much is sitting in dealer godowns.

This is the single most valuable habit for a manufacturer: do not produce for the dealer's godown, produce for the shop's shelf.

Step 3: Read stock and cover

Open Stock and check, per SKU, quantity on hand across branches and the low-stock list. Compute cover:

Stock cover (days) = total sellable stock / average daily sales

Use sellable stock only. Batches close to expiry are not really available; see batch and expiry control.

A hypothetical example with invented numbers: SKU A sells 40 cartons a day and you hold 520, so cover is 13 days. Lead time is 7 days. You have six days to start the run, which means scheduling it this week. SKU B sells 5 cartons a day and you hold 400, so cover is 80 days; no production needed.

Step 4: Build the weekly production table

Put the SKUs in a table and decide the action:

SKU	Avg daily sales	Stock	Cover (days)	Lead time	Action
A (example)	40	520	13	7	Schedule run now
B (example)	5	400	80	10	Skip
C (example)	18	150	8	6	Rush, check branches for surplus

The numbers are invented; the point is the habit. Do this table every week at the same time. Include a column for transfers: before producing, check whether another branch holds surplus. See moving stock between branches and warehouses.

Step 5: Decide batch sizes

Batch size is a trade-off between machine set-up cost and the risk of holding too much. Start from demand and cover: a run should be sized so that cover after the run is between your minimum and maximum target, for example 20 to 45 days. Never make a batch so large that it will not sell before the expiry date. Create the batch in the system on receipt with its expiry date so that stock keepers can run first-expiry-first-out.

Step 6: Add the known events

Overlay events that change demand:

• festivals and wedding seasons (see Dashain-Tihar season planning),
• planned schemes, which pull demand forward (see designing dealer schemes),
• monsoon disruptions to supply routes,
• new SKU launches, which need a separate plan with a smaller first run.

Step 7: Close the loop

Each week, compare plan with reality. Where did cover end lower than planned? Where did stock build up? Write one sentence about why. After a month you will see patterns such as "the western region always sells faster than we think". Adjust the baseline for that region and the next plan improves.

Who should be in the planning meeting

Production head, sales manager, stock manager and the owner or a delegate, for thirty minutes weekly. The sales manager brings sell-out and field notes (competitor moves, no-order reasons that mention stock-outs). The stock manager brings cover and expiry. The production head brings capacity and lead times. Decide, record and move on. This can be the second half of the weekly sales meeting.

Common mistakes

• Planning on sell-in only.
• Ignoring stock in other branches.
• Changing the plan daily. Fix the plan for your lead time and adjust weekly.
• Not recording write-offs. If you do not see waste you cannot fix the plan.

What to do when data is thin

If you only started using gamitrack recently, you may not have twelve weeks of history. Use what you have, and fill the gap with the best of your old records, such as the Excel sales sheet you migrated from. Be conservative: plan shorter runs more often until the pattern is clear. A new SKU should start with a small run sized to a few weeks of expected demand, and the second run should be decided from the first weeks of real orders, not from the launch excitement.

Reporting the plan to the team

Share a one-page summary after each planning meeting: which SKUs are in production this week, which are short, and which are overstocked. Officers use it to steer shop conversations, and dealers appreciate an honest note about expected availability. When people see the plan reflects the data, they bring better data next time.

Next steps

All the inputs above come from the same workspace you use to run orders and stock. See the plans on the pricing page and the reports articles in the Help centre.