It is the second week of September. Your Diwali campaign is booked, the creative is approved, and your manufacturer just told you their dyeing partner is running three weeks behind because every apparel brand in the country placed the same order in the same fortnight. Nothing you do now fixes that. Festive season inventory planning is not a demand problem — it is a calendar problem, read backwards. The Indian festive arc compresses a third of the year’s revenue into eight weeks, and it does it at exactly the moment your supplier has the least capacity and the fewest people at the machines.
Key takeaways
- Work backwards from the sale date. For a manufactured SKU the first irreversible commitment — material and packaging — lands around T-16 weeks.
- Your supplier is busiest precisely when you need them, and their workers take leave around the festivals themselves.
- Size uplift per SKU from last year’s same-window lift, never a blanket 2× across the catalogue.
- Packaging is the long-lead item founders forget — printed boxes and gift sleeves can outrun the product itself.
- Lead-time buffers computed in July are wrong for October. Re-run them with festive-window data.
The Indian festive arc, and why it breaks planning
The season is not one date. It runs Onam in Kerala through Ganesh Chaturthi, then Navratri and Durga Puja, Dussehra, Karva Chauth, and lands on Dhanteras and Diwali — after which wedding season picks up the baton into December. The dates shift year to year with the lunar calendar, so “last October” is not comparable to “this October”. Anchor to the festival, not the month: measure the 21 days ending on Diwali against the 21 days ending on last year’s Diwali.
The planning consequence is uncomfortable: your demand peaks in the same weeks your supplier’s does. A unit that turns a 3,000-piece order in 25 days in June will quote 40 days in September, then lose four to six working days to leave around Navratri and Diwali. Job-work partners — dyeing, printing, embroidery, plating, polishing — are the real bottleneck. They serve dozens of brands from one shop floor and they queue by who booked first.
So the whole exercise reduces to one question: what is the last responsible date for each decision, and what breaks if you miss it?
The working-backwards calendar
T-0 is the day your festive sale goes live — for most brands, the Dhanteras–Diwali window. Everything else is counted back from there in weeks. These are realistic for a small Indian manufacturing brand with domestic suppliers; if you import components or ship to a US 3PL, add four weeks to every row.
| Milestone | What has to be true | What goes wrong if it slips |
|---|---|---|
| T-16 weeks Material commitment | Fabric, base components and packaging committed with a deposit. Colours, sizes and print runs locked. | Mill and printer capacity is gone. You take substitute material, a different shade, or a longer queue — and the substitution cascades into every downstream date. |
| T-12 weeks Purchase order placed | Final quantities per SKU and size, signed PO, agreed delivery date, penalty or top-up terms in writing. | You enter the production queue behind brands who ordered on time. Your “40-day” run becomes 55 days and no one is lying to you. |
| T-10 weeks Production slot confirmed | Written confirmation of which weeks your run occupies, plus the job-work bookings behind it. | A verbal yes turns into “we’ll fit you in”. You discover the slippage at T-6 when the goods do not arrive, with no time to recover. |
| T-6 weeks Goods received | Finished stock physically in, counted, and QC-sampled against spec. | Every remaining task compresses into the same fixed window. Photography, listings and 3PL inbounding cannot be parallelised past a point. |
| T-4 weeks Photography, listings, QC | Shot, retouched, listed with correct variants and inventory counts; QC rejects quarantined and replacements requested. | You launch with placeholder images or wrong variant stock. Ads run against listings that oversell sizes you do not have. |
| T-2 weeks 3PL stocked | Inventory received at the fulfilment centre, bin-mapped, and reconciled against your system count. | Orders land in the first 48 hours of the sale against stock the warehouse has not put away. Your fastest revenue days become your worst dispatch days. |
| T-0 Sale live | Bundles built, gift SKUs live, top-up slot on standby. | Nothing to fix. Whatever you have is what you sell. |
Notice where the pain sits. The first three rows are supplier-side, invisible from your Shopify admin, and they happen before the season feels real — which is why festive misses get diagnosed as forecasting failures when they were scheduling failures. Running these dates as an actual stage-by-stage production pipeline rather than a WhatsApp thread is the difference between catching a slip at T-10 and discovering it at T-6.
Sizing the uplift honestly
The most expensive line in festive planning is a blanket multiplier. “Order 2× everything” overbuys the basics that barely move and underbuys the two SKUs that carry the season. Uplift is a per-SKU number, and last year’s data already contains it.
Worked example. A jewellery brand looks at two SKUs from last season, using in-stock days only — days when a product was sold out cannot tell you what demand was.
- Kundan gift set: quiet baseline 4 units/day. Festive window 34 units/day. Lift factor 8.5×.
- Everyday silver studs: quiet baseline 11 units/day. Festive window 15 units/day. Lift factor 1.4×.
A flat 3× rule would have bought roughly 290 kundan sets and 790 studs — a stockout on the SKU that makes the season and a warehouse full of the one that does not. The arithmetic is not clever. The discipline of doing it per SKU is what pays.
If it is your first festive season
You have no lift factor, and pretending otherwise is how first-timers end up with January cash locked in October stock. Size from a category benchmark — gifting-heavy categories such as jewellery, sweets, home decor and ethnic apparel see the steepest lifts; functional repeat-purchase categories see modest ones — and commit to the lower end of that range. Then pre-negotiate a top-up slot: a second production run, quantity and price fixed in the original PO, that you trigger in the first 72 hours of the sale when real signal arrives. A supplier will hold two weeks of capacity for a brand that asked in July. None will find it for you in October.
If your history is thin, borrow the structure from our guide to demand forecasting methods for small brands rather than inventing a number — a defensible range beats a confident point estimate.
Gifting SKUs and bundles need their own plan
Festive demand is gifting demand, and gifting behaves differently from self-purchase. Price points cluster at round numbers, presentation matters more than spec, and the buyer is often not the user — so size and personalisation risk rises, and return rates with it.
Three rules. Build bundles from components you already stock deep, not your scarcest SKU — a bundle is a demand multiplier pointed at whatever is inside it, and one short component kills the set. Decide whether a bundle is a real SKU with pre-assembled stock or a virtual kit assembled at pick time; the first dispatches faster, the second is far easier to unwind after Diwali. And make sure your system deducts component stock when a bundle sells. A bundle that does not decrement its parts is the single most common cause of festive oversell, and it is a multi-level stock tracking problem, not a forecasting one.
Packaging is the long-lead item nobody plans
Founders plan the product and assume the box will follow. It will not. Printed rigid boxes, gift sleeves, custom mailers, ribbon, foil-stamped cards and thank-you inserts run on their own lead times — often four to seven weeks for a custom print run, plus plate and proofing time, and the same festive congestion applies to printers as to garment units. Minimum order quantities are usually high, so the decision is chunky and hard to reverse.
Put packaging in the T-16 row with the fabric, not the T-4 row with photography. Give it a line in your bill of materials so it is consumed and reordered like any other component, and order for your upper demand case — year-neutral packaging keeps, whereas “Diwali 2026” foil on a sleeve is worthless on 2 November. Keep the dated element to a cheap sticker you can apply late.
Your July buffers are wrong for October
Safety stock and reorder points are computed from average lead time and lead-time variability. Both change inside the festive window. If your safety stock formula is running on a 25-day average lead time observed in June and the real September figure is 40 days with a worst case of 55, your buffer is sized for a supply chain that no longer exists.
Re-run the calculation with festive-window lead times before you place the T-12 PO, not after. And record what actually happens this season — the arrival dates, the promised-versus-delivered gap per supplier — so next year you are planning from your own data instead of a feeling. The same discipline that makes ordinary production planning work is what makes the festive version survivable.
The post-Diwali cliff
Demand does not taper after Diwali. It falls off a ledge, sometimes 60-70% in a week, and then partially recovers into wedding season for the categories weddings care about. Two things follow.
Avoid ordering into the cliff. The dangerous moment is week two of the sale, when things are selling well and a top-up feels obvious. Ask what the arrival date is. If the reorder lands after Dhanteras, you are not buying festive stock — you are buying January stock at festive prices, and you will discount it. Set a hard cutoff date on your calendar after which no festive-coded reorder is approved regardless of how well it is selling.
Plan the leftovers before you order them. Split your festive buy into three buckets at PO stage and size each one deliberately:
| Bucket | Example | Post-Diwali plan |
|---|---|---|
| Evergreen | Core products in standard packaging | Nothing. It returns to normal velocity and sells through at full price. |
| Wedding-carryover | Occasion wear, gifting jewellery, premium home | Re-merchandise for wedding season within a week. Same stock, different story. |
| Festive-coded | Diya sets, dated packaging, festival-print variants | Clear inside 4-6 weeks while intent is warm. Keep this bucket small — it is the only one that truly expires. |
Most brands discover this split in December, when it is already a markdown problem. Deciding it in July makes it a purchasing decision instead.
Run the calendar, not the panic
Festive season stock planning in India rewards one behaviour above all others: making irreversible decisions early, in writing, with dates attached. The uplift maths takes an afternoon. The bundles take a week. What takes sixteen weeks is the supply chain, and it does not negotiate in September. Put the seven milestones on a shared calendar in June, give each an owner, and treat a missed T-10 confirmation as the emergency it is — by T-6 it is not an emergency, it is just the result. If you are also planning a global sale in the same quarter, read our BFCM inventory planning guide next; the two calendars overlap, and the same factory serves both.