A restaurant POS (point of sale) system manages orders, kitchen tickets, billing, and inventory, while an online ordering stack adds your own website or app ordering alongside Zomato and Swiggy. In India in 2026, cloud POS systems cost ₹12,000–₹40,000 per outlet per year, direct online ordering setups run ₹15,000–₹1,00,000 to launch, and aggregator commissions of 18–28% are the number that decides whether owning your ordering channel is worth it. This guide covers what to buy, what to pay, and when direct ordering beats the aggregators.
The modern restaurant stack, layer by layer
A restaurant running well in 2026 typically has five software layers:
- POS and billing: table/token management, KOT (kitchen order ticket) routing to printers or kitchen display screens, split bills, discounts with audit trails, and GST-compliant invoices.
- Inventory and recipe costing: ingredient-level stock, recipe cards that deduct stock per dish sold, daily variance reports, and purchase orders to suppliers. This layer is where profit hides — food cost variance between the recipe and reality runs 3–8% in most Indian kitchens.
- Aggregator integration: Zomato and Swiggy orders landing directly in the POS, with menu and stock-out sync. Manual tablet-juggling causes wrong orders and missed items at peak hours.
- Direct ordering: your own ordering website or app, plus WhatsApp ordering. ONDC has also opened a lower-commission channel worth evaluating for delivery-heavy brands.
- CRM and loyalty: customer database from billing and online orders, repeat-visit campaigns via WhatsApp, and feedback capture.
A single-outlet café needs layers 1 and 3. A cloud kitchen lives on 3 and 4. A multi-outlet QSR brand needs all five plus central kitchen and outlet-transfer management.
What does restaurant software cost in India in 2026?
| Component | Typical cost | Notes |
|---|---|---|
| Cloud POS subscription | ₹12,000–₹40,000/outlet/year | Billing, KOT, reports, aggregator sync tiers |
| Hardware (printer, cash drawer, tablet/PC) | ₹25,000–₹80,000/outlet | One-time; KDS screens extra |
| Aggregator commissions | 18–28% per order + taxes | The number everything else is judged against |
| Direct ordering website (template) | ₹15,000–₹50,000 + PG charges | Payment gateway ~2% per transaction |
| Custom ordering website/app | ₹2–8 lakh one-time | Owned channel, loyalty, no commissions |
| WhatsApp Business API | ₹1,000–₹5,000/month + per-message | Ordering, order updates, campaigns |
On GST: restaurant service is typically 5% without input tax credit (non-AC and AC standalone restaurants alike), with 18% applying in certain cases such as restaurants in hotels above the specified room tariff. Your POS must handle these rates correctly per outlet type, and handle the aggregator model where Zomato/Swiggy collect and remit GST on restaurant services under Section 9(5) — reconciliation reports for this are a genuinely useful POS feature, not a checkbox.
Is direct online ordering worth it against Zomato and Swiggy?
Do the arithmetic on your own numbers. Say your outlet does ₹6 lakh/month through aggregators at an effective 25% (commission, ads, and payment charges combined): that's ₹1.5 lakh/month paid for demand. A direct channel — custom ordering site at ₹3–4 lakh one-time, plus ~2% payment gateway and your own delivery via rider aggregators (₹40–₹80 per delivery) — becomes cheaper than the aggregator at surprisingly modest volumes. If you can shift even 25% of aggregator volume to direct ordering, the build pays for itself in under a year, and the customer data (phone numbers, order history) becomes yours to remarket via WhatsApp instead of renting reach through aggregator ads.
The honest caveats: aggregators bring discovery you cannot replicate, direct channels need active promotion (QR codes on tables and packaging, WhatsApp campaigns, small direct-order discounts funded by the commission you're not paying), and delivery logistics are your problem. The playbook that works in 2026 is hybrid — stay on aggregators for discovery, aggressively convert repeat customers to direct. Chains from Chennai to Coimbatore are running exactly this play, and tier-2 cities like Erode, Salem and Tiruppur are fertile ground for it because delivery radii are small and repeat-customer share is high. Model your build cost in our project cost calculator.
Which POS features actually matter in an Indian kitchen?
Judge products on these, in a live trial during a Friday dinner rush if possible:
- Speed at the counter: billing a 4-item order should take under 15 seconds; every extra tap compounds across 300 daily bills.
- Offline mode: internet drops must not stop billing. Cloud sync when the connection returns — verify this actually works, not just claimed.
- KOT routing: separate printers/screens for tandoor, Chinese, and beverages, with item-level routing and reprint control.
- Aggregator deep integration: auto-accept rules, stock-out sync both ways, and a single menu master. Menu edits in five places is how items stay "temporarily unavailable" for a month.
- Recipe-level inventory: daily food cost report against theoretical consumption. This is the feature that pays the subscription.
- Multi-outlet controls: central menu and price control, outlet-wise P&L, and stock transfers priced correctly.
- Staff controls: role-based discounts, void/reprint logs, and day-close reports that match cash in the drawer.
- Data export: your sales and customer history, downloadable free, any time. Switching POS should never mean losing three years of data.
Anti-features to ignore in demos: AI demand forecasting for a 20-table restaurant, blockchain loyalty, and dashboards no one will open after week two. If a feature will not be touched by the cashier, the kitchen, or the owner at least weekly, it should carry zero weight in your decision — and certainly zero rupees in your negotiation.
Turning billing data into repeat business
Most restaurants sit on a customer database they never use. Every dine-in bill with a phone number, every direct online order, and every aggregator order (partially — aggregators mask customer numbers, which is itself an argument for the direct channel) feeds a CRM that can be worked systematically:
- Segment by recency and frequency. Three lists do 80% of the work: customers who ordered in the last 30 days (protect), 31–90 days (re-activate), and 90+ days (win back). A monthly WhatsApp campaign to the lapsed list — a simple "we miss you" with a time-bound offer — typically brings back 5–12% of recipients, at a message cost of under a rupee each.
- Anchor a loyalty mechanic to direct ordering only. Points or a stamp-card ("6th biryani free") that apply exclusively on your own website or counter orders give customers a concrete reason to skip the aggregator app. Fund it from the 20%+ commission you are not paying on those orders; a 10% loyalty giveaway on a direct order still leaves you ahead.
- Automate the moments that matter. Order-confirmation and delivery-update messages have near-100% open rates; a feedback prompt 30 minutes after delivery catches problems before they become public one-star reviews. Route sub-4-star responses to the manager's phone immediately — service recovery within the hour converts complainers into regulars more reliably than any discount.
- Watch three numbers weekly: direct-order share of total online revenue, repeat-customer share of orders, and food cost variance. A single WhatsApp report to the owner every Monday with these three figures outperforms a dashboard nobody logs into.
None of this needs enterprise marketing software; it needs your POS data, the WhatsApp Business API, and consistent execution. Under the DPDP Act 2023, collect numbers with consent and honour opt-outs — a one-line checkbox at billing and a working STOP keyword keep you clean.
When should a restaurant business build custom software?
For a single outlet, never — mature cloud POS products are excellent value. Custom makes sense in four cases: cloud-kitchen brands running multiple virtual brands from shared kitchens who need consolidated order routing and brand-level P&L; chains (8+ outlets) where per-outlet SaaS fees and rigid workflows start costing more than ownership; franchise models needing royalty computation, franchisee portals, and central purchasing; and brands whose direct-ordering experience is the differentiator — custom apps with loyalty, wallets, and scheduled delivery.
Custom builds in 2026: an owned ordering website with payments and delivery integration at ₹2–4 lakh; a full multi-outlet platform (POS integration, central kitchen, franchise management, customer app) at ₹8–20 lakh, delivered in 3–6 months with AMC at 15–20% a year. Our services page covers how we scope these, and our case studies show comparable platforms.
How NexaEx works with food businesses
NexaEx is a software company based in Erode, Tamil Nadu, serving restaurants, cloud kitchens, and QSR chains across India remotely, with on-site support where hardware and go-live demand it. We don't sell a me-too POS; we build the pieces that off-the-shelf products don't cover — direct ordering websites and apps, WhatsApp ordering automation, aggregator reconciliation tools, and multi-outlet management platforms — integrated with the POS you already run, and you own the code.
For the same buyer-first analysis in other industries, read our clinic management software buyer's guide and the school and college LMS buyer's guide from this series.
Talk to us
Send us your outlet count, monthly aggregator volume, and current POS, and we'll tell you plainly whether a direct-ordering build is worth it for your numbers — contact us or WhatsApp +91 97912 97741. We reply within 24 hours.
Frequently asked questions
How much does a restaurant POS cost in India?
Cloud POS subscriptions run ₹12,000–₹40,000 per outlet per year, with hardware — printer, cash drawer, tablet or PC — adding ₹25,000–₹80,000 one-time. Aggregator integration is often a higher tier. Judge total cost against what the inventory module saves: food cost variance of 3–8% is common in Indian kitchens.
Is a direct ordering website worth it against Zomato and Swiggy?
Usually yes as a hybrid play. At 18–28% effective aggregator cost, an outlet doing ₹6 lakh monthly through aggregators pays roughly ₹1.5 lakh for demand. A ₹2–4 lakh direct-ordering build with ~2% gateway fees pays for itself within a year if you shift even a quarter of volume, and the customer data becomes yours.
What GST rate applies to restaurant billing?
Restaurant service is generally 5% GST without input tax credit, with 18% applying in specific cases such as restaurants inside hotels above the notified room tariff. For Zomato and Swiggy orders, the aggregator collects and remits GST on restaurant services under Section 9(5), so your POS should produce reconciliation reports for that flow.
When does custom restaurant software make sense?
Not for a single outlet — mature cloud POS products are excellent value. Custom pays off for cloud-kitchen groups running multiple virtual brands, chains of eight or more outlets, franchise models needing royalty and central purchasing modules, and brands whose owned ordering experience with loyalty and wallets is the differentiator. Budget ₹2–20 lakh depending on scope.