Can a small café at Maya Berawa earn its rent?

1. Investment decision: pilot and renegotiate

The revised base assumes 75 paid transactions/day, a Rp 109k weighted menu ticket before discounts and a 53.4% contribution margin. Monthly cash surplus is Rp 3.3m after provisional income-tax retention and equipment reserve. Recipe costs and purchase shares remain unvalidated.

Operating break-even before income tax/replacement reserve is approximately 72 transactions/day; cash break-even including those provisions is 73. Opening funding remains Rp 1.12bn. Simple funding recovery at steady trade is 337.5 months. The first-year ramp returns −Rp 116.4m economic cash surplus; free cash remaining while protecting rent renewal is Rp 81.2m. This is weak compensation for the capital exposed.

Arithmetic is verified for the assumptions. Price positioning has operator-menu support. Sales attainability, quoted margins, access and photo conversion remain unproven. No numerical success probability is claimed.

2. What is known, and what remains open

Owner records and supplied photographs/layout were reviewed locally. Existing design files were preserved. No landlord, competitor or hotel was contacted, and no on-site observation was represented as completed.

3. The Berawa market and actual catchment

Maya’s published address is Jl. Pantai Berawa No.168, Tibubeneng. Berawa is the immediate trading area. Openhouse Pererenan and Gigi Susu Batu Bolong are useful brand references, but customers face a different choice at this frontage. [M1]

BPS reported 685,994 direct foreign arrivals to Bali in August 2026, down 1.69% from July. Star-hotel occupancy was 68.19%; non-star/other accommodation occupancy was 43.16%. This supports a substantial tourism market with varying accommodation performance. It does not measure Berawa café demand, resident spending or Maya occupancy. [T1]

The hotel markets itself as a 19-key property. For scale only, an illustration of 19 rooms × 75% occupancy × 1.7 guests × 50% capture yields roughly 12 buyers/day. Those occupancy, guests-per-room and capture inputs are unverified and are not used in the financial forecast. Hotel guests alone are unlikely to support the revised cash break-even of roughly 73 daily transactions. [M1]

4. Your reference brands: what transfers to Maya

For like-for-like price comparison, Copenhagen’s 79k board is approximately 91.6k if its stated 10% tax and 6% service are simply added; the receipt may use a different service-tax base. Confirm receipts before treating posted menu prices as final customer spend. The model uses tax-inclusive Maya customer tickets and assumes no compulsory service charge.

Review complaints: useful hypotheses, not a complaint-rate study

This is a limited review screen, not a representative sample of all current Google reviews. Instagram retrieval was restricted, so follower counts, current bio hours and menu highlights were not used as verified evidence. Third-party pages can mirror the same underlying reviews.

5. The choices customers can make instead

No comparable executed café leases or landlord quotations were obtained. Therefore Rp 40m is not labeled above-market rent for Berawa. The conclusion is narrower: it is difficult to support with the modeled small-format, fully staffed café.

6. Late-night sandwiches: promising geography, weak evidence so far

The test model assumes 13 active nights/month, 40 orders/night, an 80% direct / 20% delivery split and a 75k direct ticket. Two paid crew, security, management oversight, utilities and employment provisions produce Rp 13.9m incremental fixed cost. Cash surplus is only Rp 2.2m/month. At 20 orders/night, the option loses Rp 7.4m. Its base break-even is approximately 34 orders per active night.

Turning on the base night option changes combined monthly cash surplus to Rp 5.5m. That remains a modest return on the opening funding. Treat it as an incremental pilot, not justification for signing the lease.

The timing could invalidate the idea. A 22:00–03:00 service window misses customers leaving at 04:00–05:00. Observe real exits until they finish before choosing hours. Extending the shift requires a new staffing/security cost and written hotel permission. Count the customers who pass the window, see it and can stop—not everybody inside three venues.

Start with two hot sandwiches using shared daytime ingredients, one vegetarian choice, water and a simple non-coffee drink. Target handover within five minutes. Avoid launching a broad fried-food menu or a full all-night café. Validate exhaust, smell, food holding, hotel noise separation and pickup behavior first.

7. Site feasibility and service design

The supplied photos show a narrow entrance, dark finishes and a bar toward the rear. The layout provides a long 43.2m² main room and a separate 11.28m² kitchen reached through the passage. There is no evidence that the whole frontage functions as an open bakery display today. Reflective glazing may hide the product in daylight; check both directions of approach at 07:00, midday and after dark.

• Put visible products, a readable price and ordering at the street service point. Make it obvious that the public can buy without checking into the hotel.
• Keep ordering, collection and the photo position separate. Require a clear route for hotel guests and avoid queues or delivery riders occupying public/shared passage.
• Measure scooter stopping, pedestrian clearance, shade and rain protection. Secured street service does not automatically grant rights to furniture, projection or parking.
• Use sourced pastries with finishing and a short assembly menu initially. Full pastry production introduces equipment, cold storage, skilled labor and waste that this lean budget does not fund.
• Check electrical load, filtration, drainage, hot water, refrigeration, exhaust, grease/waste handling and fire/food operating requirements before any contractor commitment.
• Plan a paid roster: six crew at Rp 4.5m and one manager at Rp 9m produce Rp 36m base monthly wages; THR, employer provision, meals and relief increase payroll to about Rp 47.9m. At three working roles over 11 hours/day, coverage is 231 hours/week before leave. Six crew at 40 hours provide 240; the manager must cover relief and administration. This is a roster feasibility illustration, not an employment-law determination.

The base allocates 12 of 75 day transactions to dine-in: one per seat/day. The other 63 are takeaway or delivery. The 110-transaction upside needs 22 transactions/hour during its assumed busiest two-hour window. Test this with the actual product mix.

The takeaway concept and guided menu

Takeaway coffee and sandwiches with guided customisation. Meals Rp 75–150k, drinks Rp 40–70k, twelve supplementary seats. The menu is inspired by international sandwiches and Copenhagen’s component-choice principle, with a Gigi Susu-inspired recognisable street moment. These are three separate propositions to test: discovery, paid purchase and repeat demand.

Try the guided menu Read the operating and frontage brief

The interactive menu enforces three bread formats, compatible swaps, signature sauce plus three alternate sauces, free removals and at most two priced extras within the Rp 150k sandwich ceiling. Drink spend is additional. Supplier costs, rights and actual preparation capacity remain unverified.

Menu costing contains quantities, purchased-unit costs and yields. Actual supplier quotes and dates are blank. Food margin excludes waste, packaging, labour and transaction/platform charges. Source bread and prepared specialty meats; no bakery or smoker is funded. Katsu cooking still requires confirmed exhaust, power and cleaning capacity.

Extras: cheese 10k, egg 10k, pickles 5k; at most two and a 150k sandwich ceiling. No extras revenue is forecast until measured. Free removals do not reduce price. Bread and sauce swaps require cost, allergen and taste approval; the forecast costs recommended recipes.

Sandwich competition

The 75–150k band is credible to test, not proof customers will switch. Deli’s 82k pastrami particularly challenges Maya’s 150k version: compare portion, bread, filling, convenience and final receipt price. Copenhagen confirms boards/bowls rather than a sandwich builder. Gigi’s historical takeaway/photo coverage supports format inspiration, not transferable demand. [S1] [S2]

8. Transaction economics and opening cash

Open the editable financial model

Switch cases in Inputs!E4. Menu costing holds prices, recipe quantities, sales weights and blank quote fields. Baskets separates drink-only, sandwich-only and both by channel. One operating build feeds launch cash flow. The following comparison records separately recalculated cases, not live multi-case Excel outputs. Night is OFF.

What a customer buys

Base direct mix:45% beverage-only,20% sandwich-only and 35% both. Of beverage-only transactions,15% attach a 35k sourced pastry. Weighted beverage price 50k; weighted sandwich 107.5k. The resulting direct menu ticket is Rp 101k, not 150k on every sale. Meal transactions are 55.0% of direct purchases, with 63.6% drink attachment among meals.

Delivery mix is 10% beverage-only,35% sandwich-only,55% both. A provisional 15% menu uplift creates aRp 149k ticket before merchant discount, excluding customer-paid rider fees. Assumed merchant discount 7% and platform fee 25% require invoice/contract verification. Takeaway discount 2% and direct payment fees 1% are separate assumptions.

Day recipe expense isRp 71.7m/month; packagingRp 9.7m. Recipes use explicit quantities and yields. Downside adds 10% recipe cost; upside reduces it 5%. Additional waste is 4%/3%/2% of ex-tax undiscounted menu sales. No extras, hotel-voucher or viral-conversion revenue is assumed.

Opening cash

Rent expense appears monthly in profit; prepayment appears once in opening funding. Cash flow adds back prepaid expense and retains that amount for next rent renewal. Positive surplus uses a 22% prudent income-tax cash retention assumption, not a statutory tax return. Equipment reserve retains 1m/month. Reserve funding is not investment recovery.

Stress tests on the same base model

At 25m quoted rent, modeled cash surplus isRp 16.3m/month. A four-crew roster at 40m rent returnsRp 11.7m, but requires demonstrated coverage and peak throughput. Do not combine every favourable sensitivity and call it base.

The night column preserves an independent legacy sensitivity:75k direct basket over 13 nights, not a recipe-costed forecast of this launch menu. It remains excluded. Re-cost the actual night menu and shift only after permission and exit-demand observations.

9. Tax and contract points that change the answer

Do not interpret “tax excluded” as automatically adding 10% to rent. DJP describes 10% final income tax on the landlord’s gross building rental income. If 40m is the gross rental consideration, withholding can be deducted from that amount. If the contract guarantees 40m net to the landlord, gross rental consideration becomes 40m ÷ 0.90 = 44.44m. Rental VAT, if properly chargeable, is a separate invoice question. The model assumes net-to-landlord gross-up for prudence and tests an additional 11% unrecoverable rental VAT as a stress, not a legal determination. [X1]

The model uses a 10% customer food/drink tax provision based on Badung’s published PBJT regulation; applicability, exclusions and the current consolidated rules need Bapenda/accountant confirmation. No claim is made that rental VAT, customer PBJT and business income tax are interchangeable. [X4] [X5]

Removing the landlord gross-up changes day cash surplus toRp 6.8m/month. This contract detail matters, but does not establish demand or an acceptable investment return.

A newly formed ordinary PT or CV should not automatically be modeled at the old 0.5% turnover tax. DJP’s explanation of PP20/2026 changes eligible entity categories. The actual income-tax calculation depends on the entity, eligibility, possible facilities, taxable adjustments and depreciation. This is why a provisional cash reserve is shown separately. [X2] [X3]

Badung’s official 2026 UMK is Rp 3,791,002.57/month; the crew wage assumption is Rp 4.5m. Sector classification, employment terms, contribution ceilings and a lawful roster still need a payroll review. The blended employer provision and monthly THR accrual are explicit planning costs rather than a certified payroll computation. [L1] [L2] [L3] [L4]

10. Paid pilot and commitment gates

No field counts, supplier quotes or paid pilot have been completed. The following screens are proposals, not industry benchmarks or observed results.

Seven days of frontage and competitor observation

• Record 15-minute intervals by direction, relevant pedestrians, stopping scooters, delivery pickup, rain, queue obstruction and hotel access during proposed opening hours. Keep hotel guests, residents/workers and destination visitors separate.
• Walk the Copenhagen and Vault routes. Purchase comparable sandwiches and coffees at the reference operators. Record portion, actual receipt tax, packaging and paid-to-handoff time.
• Observe actual nightlife exits until they end. Do not double-count those people again at the frontage or equate club capacity with food demand.
• The base cash break-even mix implies about 61 direct transactions/day plus delivery. Hypothetical 5% street capture would require 1214 accessible passing opportunities/day if all direct buyers came from that flow. This is a required-count illustration, not traffic evidence.

Permitted 21–28-day paid pilot

The workbook’s blank transaction log records receipts, actual tax, discount, recipe, waste, pack, fees, timings,30-minute quality, obstruction and anonymous repeat codes. Assess the forecast against measured daily contribution, not social engagement.

• Proceed only when demand and actual contribution cover costs with an explicit error allowance, access works, and the owner accepts the implied return and recovery period.
• Renegotiate rent, annual prepayment, fit-out grace, permitted trial and break rights when the product works but location terms consume returns.
• Re-test a shorter menu/service arrangement when customisation, katsu preparation or queues undermine margin and speed.
• Reject this site when plausible demand cannot support its economics or essential street/collection rights cannot be secured.

The central challenge is to earn repeat sandwich purchases even when customers do not want a photograph. Choices must add value without enough ordering time, waste or errors to erase their margin.

Sources and research limits

Research checked on 18 October 2026, unless a publication date is specified. Perplexity was used for discovery and then challenged against direct operator pages, menu images, Maps and official statistics/tax/labor sources. Official pages establish what operators publish; they do not prove same-day venue operations. Map times are approximate and were not walked on site.

The work completes a desk feasibility assessment and a financial model. Executed contracts, supplier quotations, hotel occupancy records, first-party review exports and field/pilot measurements were not supplied. Open evidence gaps remain explicit. A current direct Instagram menu check could not be completed because retrieval was restricted.

Local source material: the owner-supplied entrance, bar and pool photographs; source layout; existing-space parameter file; seating notes; and prior investor/design allowance files. These were used for geometry and capital context, not as evidence of market demand.

Model verification: all three cases recalculated; contribution and cash/rent reconciliations checked; input stress changes, selected/unselected blanks and zero demand tested; exported formula-error scan returned none. All workbook sheets were rendered for visual inspection. Excel desktop recalculation and native What-If behavior were not tested in the desktop application. The sensitivity table was tested in the authoring engine and exported as a native two-input table.