What do your Swiggy and Zomato offers really cost? Working out each coupon

Offers are one of the most useful levers a restaurant has on Swiggy and Zomato. A well-chosen coupon brings in new customers, fills a quiet afternoon and lifts a dish people have not tried. The question worth asking every month is not whether to run offers, but which ones are earning their place — and that needs one figure per offer that most restaurants never work out.

Three kinds of order

Every completed order in a month falls into one of three groups, and each tells you something different:

• Not discounted — the customer paid the full menu price, with no restaurant-funded discount.
• Couponed — a discount the customer applied with a code. Zomato records it in the order's discount construct; Swiggy as the coupon type applied by the customer.
• Discounted without a coupon — a discount that ran without a code: buy-one-get-one, freebies, item-level offers, and membership offers such as Zomato Gold or Swiggy One exclusives.

The figure to work out: effective discount

For each offer, add up the discount you funded on its orders and divide it by the item total of the same orders. That is its effective discount: what the restaurant gave up on the orders the offer actually ran on. Count the promo and non-promo discount columns together: the platforms record them separately, and both are discount the restaurant funded.

Effective discount = restaurant-funded discount on the offer's orders ÷ item total of those orders

A worked month

Here is one sample month on one platform, 2,030 completed orders in all. The figures are made up, but they add up:

Across the whole month the discount is 13.4% of the item total — a number that sounds modest. On the orders that carried an offer it is 25.6%, and TRYNEW and FLAT100 together account for four fifths of the money. That is where a small change makes the biggest difference.

Keep, trim or retire

Nothing in a settlement file can tell you how many customers a smaller offer would keep. What the file can tell you is the size of the trade-off, so you can weigh it with your own sense of your customers:

• Trim a large offer a little. Taking a tenth off TRYNEW's discount brings its effective rate from 30.0% to 27.0% and saves about ₹3,780 if orders hold. If each of its orders pays out about ₹190, the change still comes out ahead unless more than 19 of its 420 orders stop coming.
• Retire a small, deep offer. SAVE60 ran on under 2% of orders at 40% effective — well above the 25.6% the month's offers average. A coupon like that is worth a second look.
• Watch discount without a coupon. Here it is 16% of the discount. When it grows to a quarter or more, it is worth checking which item offers and freebies are running, because they are easy to forget once switched on.

Doing it for every outlet

For one outlet on one platform, a spreadsheet and the order-level export will get you there. For several brands, two platforms and a new month every four weeks, the grouping by coupon is the slow part — and the part that is easiest to get wrong when the platforms place the same discount in different columns.

MenuMetrics' Discount burn does this from the settlement files you already upload: every completed order grouped by coupon, the effective discount of each, and suggestions to trim or retire — with the saving and how many orders it could afford to lose, worked out for you.