RestroAI features: 8 AI insights every Indian restaurant owner gets daily

RestroAI doesn't send everything it can compute. It picks 3-5 things that actually matter for the day. But across all the days and all the restaurants, the kinds of insights it surfaces fall into eight categories. Here's what each looks like, what triggers it, and how owners typically respond.

1. Sales anomaly detection

What it does: flags when daily sales (revenue or order count) deviate significantly from your baseline for that day of the week.

What "significant" means: more than ~15% deviation from your rolling 4-week same-day-of-week baseline. So a slow Tuesday compared to other Tuesdays, not a slow Tuesday compared to Saturdays.

Example WhatsApp message:

"Yesterday was 22% below your last 4 Saturdays. Customer count was normal — order value dropped from ₹420 to ₹325. Check if combo deals or upsells slowed."

How owners typically respond: investigate the cause. Sometimes it's weather, a local event, a staff change, or a specific promo that didn't run. Sometimes the dashboard reveals a real issue — a particular dish dropping out, or a payment glitch.

For the broader analytics context, see our reference post on restaurant analytics metrics that matter.

2. Menu margin alerts

What it does: tracks per-dish margin against your set target. When a dish's margin drifts below target for a sustained period, RestroAI flags it.

Example:

"Butter chicken margin slipped to 38% from 45% target this week. Cost of cream is up; consider a 5% price adjustment or a smaller portion size."

How RestroAI calculates margin: sale price minus ingredient cost (from your menu setup) minus an allocated overhead percentage. If you haven't set ingredient costs, RestroAI uses a default category-level estimate and flags this with lower confidence.

Action: price adjustment, portion review, supplier conversation about ingredient cost, or in some cases removing the dish entirely.

3. Bestseller predictions

What it does: forecasts demand for tomorrow based on day-of-week patterns plus calendar context (weekends, public holidays, festivals).

Example:

"Tomorrow is Saturday + start of a long weekend. Last 3 long weekends, paneer tikka ran out by 9 PM. Prep 1.5x today."

Festival awareness: RestroAI's calendar covers Diwali, Holi, Eid, Christmas, regional festivals (Pongal, Onam, Bihu, etc.), and major public holidays. The festival logic improves as RestroAI sees more years of your historical data.

Action: prep adjustments, ingredient pre-orders, additional staff for the shift.

4. Customer behavior tracking

What it does: monitors repeat customer rate, loyalty patterns, and churn risk based on the customer phone numbers Restrofi captures during ordering.

Example:

"12 customers ordered for the second time this week — your repeat rate is climbing. Send a thank-you offer?"

How RestroAI defines "repeat": same phone number within a 30-day window. The threshold is configurable.

Action: targeted WhatsApp loyalty offer, manual thank-you message, or systematic loyalty program enrolment.

5. Waste detection

What it does: flags menu items that haven't sold in N days. The default thresholds are 7 days for cooked items and 14 days for shelf-stable.

Example:

"Mushroom soup hasn't sold in 8 days. Consider removing it from the digital menu before next week's prep."

Action: menu pruning, ingredient sourcing reduction, or repositioning the dish (better description, better placement, combo bundling).

6. Peak-hour shift detection

What it does: tracks the hour of day that drives the most revenue, and flags when it shifts by more than 15 minutes over a 30-day window.

Example:

"Your busiest hour shifted from 1 PM to 1:45 PM over the last 30 days. Check kitchen staffing for that window."

Why this matters: kitchen staffing schedules, ingredient prep timing, and reservation policies are all built around your peak. If peak drifts and you don't notice, you're under-staffed at the new peak and over-staffed at the old one.

Action: review staff schedules, prep windows, and any reservation timing rules.

7. Daily summary

What it does: the always-on baseline message — orders, revenue, AOV, top dish, slowest dish.

Example:

"Yesterday: 47 orders, ₹18,400, AOV ₹391. Top dish: paneer tikka (12 orders). Slowest: dal makhani (1 order)."

This is the message everyone gets every day, regardless of plan. Free plan users get only this; paid plans get this plus the other 7 categories layered on when relevant.

8. Festival and seasonal readiness

What it does: 5-7 days before major festivals or seasonal events, sends preparation reminders based on your historical performance during the same event.

Example:

"Diwali week is in 5 days. Last Diwali, you ran out of paneer items on Day 2. Stock check now."

Built-in calendar: major Indian festivals plus your historical performance during each. The first year, RestroAI uses general patterns; from year two, it uses your specific data.

Action: ingredient ordering, menu specials planning, staff scheduling, marketing pushes.

Which features are in which Restrofi plan

Feature	Free	Starter (₹499/mo)	Professional (₹1,499/mo)
Daily summary message (#7)	✅	✅	✅
Sales anomaly alerts (#1)	❌	✅	✅
Menu margin alerts (#2)	❌	✅	✅
Bestseller predictions (#3)	❌	✅	✅
Customer behavior tracking (#4)	❌	✅	✅
Waste detection (#5)	❌	✅	✅
Peak-hour shift detection (#6)	❌	❌	✅
Festival readiness reports (#8)	❌	❌	✅
Multi-outlet consolidated view	❌	❌	✅
Hindi/Hinglish/English	✅	✅	✅

For the full pricing breakdown, see RestroAI pricing.

How RestroAI prioritizes which insights to send

Every night, RestroAI runs all 30+ signals across all 8 categories. But it doesn't send everything — that would be notification spam, and notification spam gets muted within a week.

The selection rule is brutally simple: would the owner do something differently if they knew this? If yes, surface. If no, drop. The result is a daily message that almost always contains something you'd want to act on.

This is why a busy day might surface 5 things, and a quiet day might surface just the daily summary plus one observation. The volume varies; the quality bar doesn't.

Where to go next

If you're new to RestroAI, start with What is RestroAI? for the orientation. If you want to understand the underlying pipeline, How RestroAI works walks through the nightly batch architecture. And if you're trying to pick a plan, RestroAI pricing lays out the tiers.