Restaurant groups

Every brand. Every branch. Every peso.

Accounting, inventory, production, payroll, POS and guest experience for a multi-outlet restaurant group — one Philippine-built system, live in thirty days.

Sales by outlet · yesterday6 outlets
AyalaCebu Business Park₱284,12062.1%
IT ParkApas₱241,88059.4%
MactanLapu-Lapu₱198,34057.8%
BaniladMandaue₱176,90054.2%
TalambanCebu City₱121,45549.6%
Commissaryissues to outlets₱402,610
KUSINA SA CEBUTable 12 · 4 pax · 7:42 PM
2 Lechon Belly760.00
1 Kinilaw na Tanigue285.00
4 Rice60.00
Service charge 10%110.50
VAT 12%132.60
TOTAL1,348.10
Recipe cost₱461.20 · 65.8%
Official Receipt · BIR registered
The question no ledger answers

Five brands. One of them is losing money.

Each brand hits its sales target. One of them is quietly losing four per cent to a commissary it never gets charged properly for, waste nobody logs, and a menu that has not been costed since the last supplier increase. You find out at the quarter, by rebuilding it in a spreadsheet.

Built for this

Three problems your trade has that other businesses never do.

01

Profit by brand, after the kitchen

Share the commissary, the warehouse and head office across the outlets they actually serve, on a basis you choose. Two brands that look profitable on their own costs stop looking profitable the moment they carry the kitchen — and that is the number that decides anything.

  • Five bases — revenue, covers, floor area, headcount or an equal share.
  • It reconciles to zero every period, or it tells you why.
  • Every share opens its working — the driver, the rule and the rows.
See how it works
Commissary, warehouse and head office ₱659,652.39
Bukidnon Roadhouse Cafe₱204,491.24
Tirso’s Deli and Roasters₱171,509.62
Tirsos₱118,737.43
Tisa Lounge₱92,351.33
Totsy’s Cakes and Pastries₱72,562.77
Reconciles to ₱0.00 Allocated out and allocated in agree to the centavo.
02

Food cost you can actually prove

Theoretical against actual, per branch, from two independent sources — what the recipes say should have been used, and what the counts say was. The gap is the number, and it is worthless unless the two sides were measured separately.

  • Recipes and sub-recipes costed from what you actually paid.
  • Yield and trim loss carried, so a plate cost is a real one.
  • Every dish ranked by how well it sells against what it contributes.
See how it works
Yield is 3.8 points under standard, and it costs ₱54 a plate.Line 2, six runs, and the drift is not noise — it started on the run after the supplier changed.
Why it says soStandard and actual come from two independent places: the recipe on one side, what the kitchen actually issued on the other. Both from the same column would agree every time and prove nothing. The variance is posted rather than absorbed, so it is a line somebody answers for.
Yield91.2%standard is 95.0%
Cost per plate+₱54against the March costing
This month₱142,800at current volume
Runs affected6 of 6since 12 September
Recost the recipeDrafted from the prices actually paid. A person approves it.

The supplier price rose 8.4 per cent in March and the recipe was never recosted. That is the whole of it.

03

The registers you already have

We read the Z-reading off whatever is on your counter. If it exports, we connect to it; if it does not, the reading is photographed and read. Nobody retypes a day of sales, and nobody replaces a till that works.

  • Any register — API, file, database or the printed reading.
  • One journal per site per day, with the reading attached as evidence.
  • Senior and PWD discounts captured at the counter and reported from the sales.
See how it works
Kusina Ayala · tonight
18:04Table 12 · dine in6 covers, closed₱4,182.00
18:22Take outPaid by card₱1,240.00
18:41Table 4 · dine inSenior discount, 2 of 4₱2,918.00
19:05Void · Table 7Reason recorded, supervisor−₱860.00

Every one of these posts to the books tonight, not at the end of the month.

Datche AI

What your people actually ask it, on an ordinary Tuesday.

Every answer comes from your own records, runs with the asking person’s permissions, and carries the rows it was read from.

  • “Which brand actually makes money after the commissary?”
  • “What did we waste last week, and where?”
  • “Which dishes lose us money?”
  • “What should I order for the weekend?”
  • “Why did food cost move at Corner Cafe?”
  • “Send me the P&L by brand as a file.”
It cannot make a number up

Every figure comes from a view over your own records — never the model’s arithmetic — and the answer carries the query and the row count behind it. No other ERP sold in the Philippines has this.

16
Brands live
on one set of books
30
Days to live
written into the contract
₱0
Per user
unlimited, every plan
24–48h
To a change
most of it the same day
Questions

The ones this trade asks first.

Do we have to replace our POS?

No. We read whatever you have — through its API, its database, a scheduled file, or the printed Z-reading if it exports nothing at all. Replacing a working till is a cost with no benefit, so we do not ask you to.

Can it handle a commissary that sells to the outlets?

Yes, and it is one of the things the system was built around. Production runs are costed from what was actually issued and what actually came out, transfers move the stock at cost, and the shared cost of running the kitchen is allocated across the outlets it feeds.

What about brands in different companies?

Handled. Each company keeps its own books and files its own returns, and the group consolidates with intercompany taken out. If stock moves between two companies, that is an intercompany transaction and the system posts both sides.

How long before we see our own numbers?

Week two. Your opening balances go in first, and then a real day of your own paper is run through capture in front of you.

We have been burned before. What makes this different?

Almost everybody we talk to has a system project that died, and usually the same way: a long implementation, a partner who left, and a go-live date that kept moving. Ours is thirty days with the date written into the contract, we implement it ourselves rather than through a partner, and if thirty days is not realistic for your business we will say so before you sign rather than in month three.

What happens if it does not work out?

You keep everything. Your data comes out in open formats at any time and at no charge, and there is no license to unwind because we do not sell one. A system you cannot leave is a system you do not own.

Who actually answers when something breaks?

We do — one company, one number, and the people who wrote the code. There is no reseller in between and nobody to escalate to. Datche has been trading in Cebu since 1978 and our own business runs on this.

My people are not technical. Will they actually use it?

That is what the assistant is for. A supervisor asks it how to do something and it walks them through that screen in plain English, with their own permissions and their own data. Most of the training a system like this normally needs is somebody wanting to ask a question — and now there is something to ask.

How much of my staff’s time will this take?

Less than a migration, because there is no migration. Your opening balances go in, and after that the system reads the documents you already produce instead of asking your people to key them again. The week that costs you anything is week two, when your own day runs through it with us in the room.

What about BIR?

VAT, withholding, the 2307s, the summary lists and the books of account are the shape of the system rather than a module added at the end, and we supply the documentation your CAS registration needs. We will also tell you what adopting a computerized accounting system obliges you to do by 31 December 2026 — before you sign, not after.

The Restaurant groups edition 29 pages · 2.4 MB PDF · everything on this page, in print

See it on your own figures.

Bring one ordinary day from your business and we will run it through in front of you — your items, your prices, your branches. About an hour, and no obligation.