Sell the stock you hold at the best margin.
One engine, six parts that share your data and rules.
01Why Zyntel
They discount stock that didn’t need discounting, and run out of stock people wanted. Both come from decisions made late, on too little information, with only one tool.
Zyntel finds out why it’s slow first. A visibility problem needs creators, not a markdown. A stock-in-the-wrong-store problem needs a transfer.
Separate a price problem from a visibility problem.
Reorder quantities and order-by dates are worked back from supplier lead times and the festivals coming to each region.
Stop sell-outs before Diwali, not after.
If an event that clears this category is days away, the engine says wait. Discounting now gives margin away for sales you would have made anyway.
Wait costs nothing. Discounting costs margin for good.
02Six levers
Zyntel asks one question of every product, every day: is it in trouble, or about to be? Then it recommends the cheapest fix that works, and explains why in plain words.
Every product, every run
Is it selling faster than we hold, or holding more than we can sell?
The engine compares expected demand, including events ahead, with stock on hand and on order.
Buying path
A reorder quantity and an order-by date, worked back from the supplier’s lead time. Capped by open-to-buy, storage and shelf life.
Promoting path · cheapest fix first
Send it where it sells: another store, region or warehouse.
Put it in front of more people through creators and affiliates.
A festival or season is coming that clears it at full price.
Step down the category’s ladder, above the margin floor.
Clear to an outlet, bundle or liquidator and free the space.
Keeps marginCosts margin for good
03Inventory intelligence
Health and risk for every product, scored from sell-through, age, days of cover and margin. Slow movers surface weeks before they become dead stock.
| Product | Health | Cover | Sell-through | Lever |
|---|---|---|---|---|
| Slim denim, blackFF-DEN-0874 | Overstocked | 61 days | 34% | Move |
| Graphic teeFF-TEE-5510 | Ageing | 96 days | 41% | Discount |
| Cotton kurtaFF-KUR-2291 | Selling out | 9 days | 88% | Buy |
| Silk-blend sareeFF-SAR-1150 | Not seen | 48 days | 22% | Get seen |
| Quilted jacketFF-JKT-3302 | Event ahead | 70 days | 18% | Wait |
Decide with reasons
Every recommendation shows why, the second-best option, doing nothing, and the rules it followed. Approve it, pick another, or reject it with a reason the engine learns from.
Graphic tee, summer drop
Discount 20%
Why: 96 days in stock against a 12-week selling life. No event ahead clears this category. Moving it won’t help: every store is overstocked. First step of the fast-fashion ladder (20 / 40 / 60).
| Option | Units sold | Margin | Left after 10 days |
|---|---|---|---|
| Discount 20% · recommended | 1,120 | ₹3.9L | 720 |
| Get seen · 3 creators | 610 | ₹3.1L | 1,230 |
| Do nothing | 240 | ₹1.4L | 1,600 |
Rules applied: margin floor ₹212 · brand limit 30% · category ladder · tightest wins
04Two ways to run it
Zyntel owns a live cart engine. If you already have a promotion tool, take the recommendations instead and act in your own systems.
Live cart
Every order passes through the engine, so no order goes below the margin floor and an offer stops the moment its budget or stock runs out.
Recommendations only
The engine decides product, lever, depth, place and timing, and sends the list to your systems every day as a file or feed. Nothing in your checkout changes.
Mix them: online orders through the live cart, store tills on recommendations. Or start on recommendations for a season and move to the live cart one category at a time. The decisions are the same either way.
06Retail calendar
Festival dates shift every year, Eid is confirmed by moon sighting, and Dhanteras is a few auspicious hours. We keep the calendar for every state, and your team adds its own local events.
| When | Event | Moves | Lever | Engine says |
|---|---|---|---|---|
| 17–21 Oct | Durga PujaWest Bengal, Assam | Ethnic wear, footwear | Wait | Hold 3 lines at full price |
| 6 Nov | DhanterasAll states | Jewellery, appliances | Buy | Order-by date passed for 1 line |
| 8 Nov | DiwaliAll states | Gifting, home, ethnic | Move | Shift 2,400 units north |
| Nov–Feb | Wedding seasonPeak dates by region | Ethnic, jewellery | Buy | Buy ahead on 14 lines |
05The platform
07Built for your teams
A daily worklist of what to reorder, move or hold, with order-by dates and draft purchase orders.
Promotions, markdowns and creator campaigns timed to stock and events, inside brand limits.
Margin floors, budgets and approval levels set before the spend. Every decision logged with its cost.
Start with a file upload. Add feeds and the live cart when you’re ready. One API, separate test and live.
08Questions
It reads your sales, stock, what’s on order, the festival calendar, weather and sport, and decides what to do with each product: buy more, move it, get it seen, wait, discount or exit. It recommends the cheapest fix that works, inside your rules.
No. You can run on recommendations only: the engine sends a daily list or file and your own systems apply it. Or use our live cart for some channels or categories and recommendations for others.
At least 100 SKUs and whatever sales history you have. More history sharpens forecasts, but you can start with a few months.
Yes. We maintain festival dates by state, plus wedding season, payday and school cycles, weather and sport. You can correct dates and add your own events.
You choose per campaign or client: a fixed fee, per sale, or per sale after the return window closes. Rate limits and return windows are set by category.
No. Rules come in layers (break-even, category, segment, brand agreement, product override) and the tightest one always wins.
09Walkthrough on your data
We load a sample of your catalogue, sales and stock, run the engine on it, and walk you through every recommendation, the reason behind it and what it would have earned.
Go live with as few as 100 SKUs and any length of sales history.