Seasonal Demand Forecasting for Diwali and Peak Planning

The Cost of Diwali Inventory Planning That Starts Too Late

For a brand selling into Blinkit, Zepto, Swiggy Instamart, or modern trade, a festival peak compresses a month of demand into roughly two weeks. Channel POs get larger and more frequent, DC appointment slots get scarcer, and every supplier you rely on is simultaneously serving everyone else's festival build. The brands that perform well through it are usually not the ones with the cleverest forecast — they are the ones that converted a rough forecast into supplier POs, confirmed inwards, and allocated warehouse stock weeks before the window opened.

How FilFlo Actually Forecasts: LTMA and the Retail Forecast Workbench

A lot of software in this category claims machine-learning demand prediction. We'll be precise about what FilFlo ships instead, because operators deserve to know what the number on the screen actually is.

The forecasting method in the product is LTMA — the last-three-month average of primary shipments, computed per distribution point. It is deliberately simple: what you shipped to each retailer or distributor over the past three months, averaged, is the baseline expectation for next month. Simple has a virtue that matters at 6 a.m. on a dispatch dock: everyone on the team can see why the number is what it is, and everyone can argue with it.

The Retail Forecast workbench sets that baseline against two imported data sets: secondary sales — actual sell-through at the retailer, uploaded via the Secondary Sales Import — and opening stock at each distribution point. From primary shipments in, secondary sales out, and opening stock, the workbench computes a projected closing stock per retailer-SKU. That single derived number is the heart of seasonal planning, because it distinguishes two situations that look identical in your own shipment data: a retailer who stopped ordering because demand died, and a retailer who stopped ordering because they are still sitting on your last big shipment.

Beverage brands like Jimmy's Cocktails, selling through modern trade, quick commerce, and general-trade distributors at once, run exactly this workbench: primary shipments per distribution point, secondary-sales imports for sell-through, and a planner walking the retailer-SKU grid before committing the festival build. It is manual where manual is honest — the seasonal judgement stays with a human who knows that Navratri moves differently in Ahmedabad than in Kolkata — and automated where automation is safe: the arithmetic, the reconciliation, and the flags.

The Days of Cover Formula: Safety Floors, Reorder Points, and MOQ

Forecasting tells you what demand to expect; replenishment math tells you when to buy and how much. FilFlo's replenishment is consumption-based and runs on three rules:

In the product this shows up as Procurement Alerts: each low-stock SKU appears with a severity (Critical for zero stock, Reorder Soon otherwise), current stock, days left — e.g. "10.5 days left" — and an editable suggested quantity, with the default supplier and last unit price alongside. The buyer multi-selects alerts and creates purchase orders in bulk. Two clicks from "we're going to run out" to a PO the supplier can acknowledge.

Seasonality Is a Run-Rate Adjustment (Plus a Lead-Time Adjustment)

Here is the practical trick that makes the formulas above seasonal: every input is adjustable, and only two of them change with the season. Ahead of Diwali, the planner raises the daily run-rate on festival SKUs — if last October ran 2.5× the trailing average and this year's channel commitments are bigger, set the run-rate accordingly. At the same time, stretch the inward TAT, because every supplier and transporter is slower in the pre-festival crunch. A higher run-rate and a longer TAT both push the safety floor up, and the reorder math produces the festival build automatically — weeks early, because that is when the now-larger floor gets breached.

The same logic runs in reverse after the peak. Drop the run-rates back down in the week the festival window closes and the alerts stop suggesting purchases — which is precisely when the panic-buying instinct is strongest and most expensive. Monsoon works identically for the categories it touches: adjust run-rates for the demand shift and stretch TATs for the logistics reality of August freight.

After the Peak: Fill-Rate Scorecards Show Where the Plan Missed

A seasonal plan is only as good as the post-mortem it gets. FilFlo grades the season with the fill-rate funnel: every order line is tracked through Ordered → Approved → Fulfilled → GRN Received, and any quantity lost between stages carries a named reason — short supply or out of stock at approval and fulfilment, damage received or short received at GRN. Computed per channel, this becomes a scorecard that answers the only question that matters in November: where exactly did we leak?

The distinctions are actionable. Losses tagged out of stock are a buying failure — the run-rate adjustment was too timid or the PO landed late. Losses tagged short supply with stock on hand are an allocation failure — inventory sat in the wrong warehouse or was held for the wrong channel. Losses tagged damage received are a packaging and freight failure that no forecast improvement will fix. Each line of the scorecard feeds a different next-season correction, and the Sales Loss report attaches the rupee value — short quantity × PO rate — so the corrections get prioritised by money, not by anecdote.

Alongside the funnel, the Sales Flash report (day/week/month channel × period matrix with percentage deltas) shows the peak unfolding in real time, and per-warehouse Days On Hand shows which locations are draining fastest — useful when the mid-peak question becomes "we have stock, but is it in the right city?"

What to Do in Your Operations 6 Weeks Before Peak Season

If you sell into Blinkit, Zepto, Swiggy Instamart, or distributors, peak season planning is not just about having stock — it's about having the operational setup to absorb a surge in channel POs without compliance failures or fill-rate drops.

Frequently Asked Questions