Customer analytics to win repeat orders

Every new customer an Indian restaurant wins through aggregator discounts or marketing pays off only after they return. Customer analytics (cohorts, RFM and lifetime value) is how good operators build repeat orders that make the economics work.

What are the economics of retention?

The idea is not new but is often under-used: winning a new customer costs more than keeping an existing one. Restaurants that spend heavily on Swiggy and Zomato promotions, discount offers and social ads to get first-time orderers feel this in cash.

Take an illustrative case. If you spend ₹100 in effective acquisition cost (promotion discounts, commission on a discounted first order and attributed marketing) to win a delivery customer who never returns, the economics are poor. If that customer orders four times in six months, the cost is spread across four orders and the picture changes. The ratio of one-time customers to repeat customers shows whether your restaurant has a sound model or a leaky bucket.

Why is customer identity a problem in India?

Swiggy and Zomato do not share phone numbers, emails or persistent customer IDs with restaurant partners. You receive anonymised order identifiers, not customer-level records.

So for a delivery-only restaurant, knowing that the same person ordered three times in two months is not possible from aggregator data. Proxies such as the same delivery area, time pattern and order value are imprecise.

Customer analytics is therefore most powerful where you have a direct ordering channel: a branded app, a WhatsApp ordering flow or a direct web ordering page, where identity is captured. Even if direct orders are a minority, they create the customer data asset that makes retention analytics and targeted marketing possible.

How does cohort analysis work?

Cohort analysis groups customers by the month of their first order and tracks how many of each group order in later months. With direct order data, it shows whether retention is improving or worsening, and whether customers won in a promotion behave differently from those who found you on their own.

How do you build a cohort table?

Rows are the month of first order. Columns are later months. Each cell shows the share of that cohort who ordered at least once in that month. A healthy table shows curves that flatten at a loyal-customer floor, since people who order steadily for months rarely stop altogether. That floor is a durable revenue base. The shape of the curve, how steeply it falls at first and where it levels, shows the quality of the customers you win and how well your retention efforts work.

What is RFM analysis?

RFM sorts customers on three scales: recency (how recently they ordered), frequency (how often) and monetary value (how much they spent). For restaurants it creates segments you can act on, so you can send targeted messages instead of blasts.

Which RFM segments matter?

• Champions ordered recently, order often and spend a lot. Thank and reward them. Do not market to them aggressively. A small benefit such as a dessert voucher keeps their loyalty cheaply.
• Loyal customers order often with decent recency but lower spend. They are candidates for upselling and referral programmes.
• At-risk customers used to order often and have missed their usual window. This segment is the best use of targeted effort. A personal reactivation offer converts far better than the same offer sent cold.
• One-time buyers ordered once and never returned. Find out why. For direct channel customers, a short survey or an offer to return within two weeks can help.
• Lost customers had high past frequency and have gone quiet. Win-back offers can bring back some, but set realistic expectations that depend on why they left.

How do you calculate customer lifetime value?

Customer lifetime value (CLV) is the net revenue you expect from a customer over the whole relationship. A simple version is average order value times orders per month times expected months as a customer times net margin.

Here is an example with assumed numbers. A delivery customer has an average order of ₹400, orders 1.5 times a month, stays 14 months and gives a 12 percent net margin. CLV is ₹400 × 1.5 × 14 × 0.12, about ₹1,000. You can rationally spend up to roughly that amount in total to win and keep the customer, which is more than many operators think is right for retention.

CLV also shows the leverage of retention. If better retention stretches the average life from 14 to 18 months, CLV rises to about ₹1,300. Across 10,000 customers that is a gain of nearly ₹29 lakh in lifetime revenue.

How should you use WhatsApp for retention?

WhatsApp is the main messaging platform in India, and the WhatsApp Business API suits retention messages better than email or app push, which need app installs or are often ignored.

Build programmes around triggers, not schedules. Useful triggers are:

• a customer’s expected next order date passing with no order (at-risk trigger)
• a birthday or anniversary, if you collect it
• a poor rating, which should fire an immediate service recovery message
• a welcome message after the first order with an incentive to order again within two weeks

Access to the API usually goes through a business solution provider. Costs are modest next to the revenue recovered if even a small share of the customers you contact come back. Always follow consent rules under India’s DPDP Act for marketing messages.

How do loyalty programmes pay off?

Point systems, cashback on direct orders and tiered membership are worth it when designed well and destroy margin when not. The central question: how much of the benefit goes to customers who would have ordered anyway, and how much changes behaviour?

The most efficient designs reward frequency, not only spend. Giving everyone a flat cashback on every order hands away margin without changing behaviour. A meaningful reward, such as a signature dish or a credit, for customers who order five times in a month targets the behaviour that creates the most profitable customers.

Use analytics to measure the return. Compare order frequency of enrolled and non-enrolled customers, allowing for self-selection, measure extra order value against the discount given and compare CLV between cohorts that joined before and after launch.

How does Restrologic help?

Our restaurant analytics includes customer analytics: cohort retention from direct channel data, RFM segments and CLV set to your own margins. For chains with both direct and aggregator orders, we build a joined customer view where identity signals allow. It connects to WhatsApp providers for trigger-based campaigns. The Restrologic POS QR menu and e-bill can help capture customer details with consent. To get started, contact us.