Preboot for restaurants
An AI that knows your menu — and your regulars — takes the order while the line rings busy. Journeys that resell no-show tables, win back drifting regulars, and catch the bad night before the 1-star review. And Loop behind them, doing the commission math, the menu calls, and the branch reviews — arriving with decisions, not dashboards. Approval-gated, reversible, on the record.
Sound familiar?
The phone rings and nobody can reach it.
Every missed call during the rush is an order that went to whoever answered.
Every no-show is a table I turned someone away for.
Published averages put reservation no-shows around one in seven bookings — and higher on peak nights, exactly when every table counts.
I don't know which dish brings people back.
Kitchen instinct is real. But it can't read three months of orders.
Your food already brings them back. The rest of the experience should work just as hard.
The tables
Three journeys that defend the covers — tonight's bookings, the regulars, and the reputation.
Confirm-or-release reminders run before every service, designed to cut no-shows 30–45% — and the freed prime table is rebooked the same night.
Describe your no-show policy in one sentence — the journey is built in minutes and retuned when results slip.
Reads your reservation system through custom integration: released tables are offered to tonight's waitlist automatically.
Win-back journeys run 24/7 on drifting regulars, designed to lift reorder frequency 15–30% — without a single blast to the whole list.
Regulars past their usual order gap are flagged daily — caught two or three weeks before they're truly gone.
Separates 'went off the food' from 'moved away' — before you spend discount margin on the wrong fix.
Feedback is asked after every meal, while it's warm — and unhappy diners are pulled to a human in minutes, before the public 1-star.
Every reply is mined — dish gripes, wait times, staff mentions — themes from all your feedback, not a sample.
A complaint theme rising at one branch is flagged the week it starts, not in the quarterly review.
Loop, watching
Loop watches orders, covers, and complaints across every branch — and arrives with the call already made: reasoning, expected outcome, and risk attached.
Slipping prep times, a cooling branch, a stale offer — surfaced unprompted, even after weeks of being ignored.
Every alert lands with a recommendation, the expected outcome, and the risk of another quiet week.
Approve the fix inside the alert and Loop executes it — reversible, audited, done by lunch.
Menu change, courier delays, or a competitor opening? The causes come ranked with evidence — before any money moves.
Orders, chats, campaigns, and feedback already live in one queryable warehouse — no data team to hire.
It lands as a decision: the fix, the expected recovery, and the cost of doing nothing.
A brief, not a hunch: which customers to migrate first, the commission saved per month, and the risk — shown before you act.
Direct orders are taken end-to-end in chat — the 15–30% app commission stays in your margin.
QR-in-the-bag journeys convert app customers to WhatsApp — the next order arrives commission-free.
The front line
The AI absorbs the peak so nothing rings out — and catering leads stop waiting for next-day callbacks.
The AI takes the full order knowing the regular's usual and allergies — missed peak orders go to near zero.
WhatsApp, Instagram, and Messenger orders land in one inbox — nothing rings out, nothing is taken twice.
The menu bot upsells sides as it takes the order — designed to lift the average ticket 8–15% — and tricky requests hand off with context.
Headcount, date, budget, and dietary needs are captured in minutes — your events manager gets a qualified brief, not a missed call.
The bot asks the right questions in order and hands off complex events with everything attached.
Instagram DMs and WhatsApp inquiries queue in one owned inbox — no catering lead answered twice, or late.
Run it on facts
The part no chatbot ever touched: seasons, menus, and staffing — decided on evidence, on a cadence that keeps itself.
Every year, a month out: last season's numbers, a staffing plan, and campaign drafts arrive unprompted.
This year's dates, demand trends, and competitor iftar offers are researched and folded into the plan.
Seasonal campaigns are drafted, targeted from last year's orderers, and launched only on your approval.
'Margin per dish, by branch, by daypart' — answered in chat. Menu-consultant answers, no retainer.
What diners ask for and never find — the dishes your menu is missing, mined from the whole inbox.
Kill, reprice, or promote — per-dish calls arrive as briefs with the expected margin impact.
Shift targets and bonuses are set from actual service behavior — gut-feel incentives become ones your staff call fair.
Reply times, resolutions, and upsell outcomes per staff member — from real conversations, not annual reviews.
Order and message traffic by hour and branch in one chart — the staffing case makes itself.
Before & after
The remaining five moments from our restaurant map — run by hand, and run on Preboot.
Slow weeknights
A blanket discount to the whole list
Loop spots the soft night, drafts and targets the offer, launches on your approval — designed to lift off-peak covers 10–25%
New-dish launches
Judged on likes, reach, and gut feel
The message→conversation→order chain captured natively — the launch's revenue is a fact by Friday
Branch comparison
A day of spreadsheets, and excuses
Why one branch trails another — location, staffing, menu mix, or service — ranked with evidence, in one question
The unexplained dip
Stays open for weeks, answered by nobody
Loop messages the branch manager who owns it — the answer folds into a decision brief the same day
The next branch
A broker's pitch and a gut feel
Delivery heat-maps, chat demand, and market trends aligned into a case you can defend
"Published averages" are industry benchmarks from public studies, not Preboot results. Your own numbers will live on your dashboard — that's rather the point.
Done right by default
Your WhatsApp number, protected.
Restaurants blast their whole list and lose their number. Preboot sends to segments built from real behavior — the biggest cause of blocks, designed out.
The menu and the bot, never out of sync.
Change the menu and the bot follows. Every change is validated before publish, so it never invents a dish or a price.
Every AI conversation, auditable.
Session summaries, handoff reasons, reasoning traces — you can read exactly what any customer was told during the rush.
Nothing changes without you.
Campaigns, journeys, the bot itself — every change is previewed, approved by you, and reversible in one click.
Keep them — they're a channel. Preboot owns your direct channel: the customers, the conversation, and the data stay yours, and no commission stands between you and your regulars.
Update it and the bot follows. Changes are validated before publish, so the AI never sells yesterday's menu or invents a price.
Reservations work in chat from day one. POS and reservation systems connect through custom integrations so everything lands where your team works.
Yes — multiple stores and projects are built in, so each branch can have its own number, menu, and reports while you see the whole picture.
It measures what's already on the record: reply times, resolution quality, and upsell outcomes from real conversations — the same metric, the same window, for every branch. Managers see it under roles and permissions, and targets built from it tend to be accepted precisely because they're factual.
No. Loop drafts and proposes — a campaign, a re-timed reminder, a staffing change — and executes only after you approve, right inside the alert. Every change is previewed, reversible in one click, and on the record.
Message us on WhatsApp — our own number runs on Preboot, so you'll taste the experience before we explain a thing.