Set opening hours from customer traffic
Turn hourly customer history into lean schedules that cover demand without empty payroll.
Decision first
Keep an hour open only when its expected gross profit reliably exceeds the labor and operating cost.
This guide is calibrated to Big Ambitions 1.0 with hotfix build 3674, released after the August 28 full launch. Opening hours are treated as a measurable operating decision, not a fixed tier list.
Confirm the executable reports 1.0/build 3674.
Complete operating playbook
- ARTICLE LENGTH
- 5,283 words
- DEEP-DIVE SECTIONS
- 14
- RESEARCH LEDGER
- 7
Scope, build baseline, and evidence rules
This guide is calibrated to Big Ambitions 1.0 with hotfix build 3674, released after the August 28 full launch. Opening hours are treated as a measurable operating decision, not a fixed tier list.
Build 3674 matters because its official notes fix a failure state in which employees trying to quit could interrupt several scheduled systems. If a pre-hotfix save produced inexplicable missed work, deliveries, imports, or training, advance to the next day after updating before using the affected day as evidence. A single broken day is not a demand curve.
Developer replies establish the durable rule: business traffic varies by business type, weekday, weekend, morning, and evening. Some concepts can trade profitably around the clock, while others concentrate demand in narrow periods. That statement supports testing by hour; it does not support copying one universal schedule.
The primary instrument is BizMan, then Insights, then the customer graph with Yesterday selected. Record the hourly customer bars alongside the staff schedule, price, promotion, capacity warnings, stockouts, and cash result. A customer bar without costs cannot prove that an hour should remain open.
F1 help and the live business requirements screen outrank an old forum screenshot. If a capacity label, service time, or scheduling rule in build 3674 differs from an Early Access post, use the live value and note the discrepancy in the test log.
Community charts are useful as hypotheses only when they name a version and explain the test method. A versioned 1.0 hour-by-hour dataset can tell you which hours to test first, but district, competition, pricing, promotion, capacity, staffing, and inventory still make the save-specific answer different.
Action checklist
- Confirm the executable reports 1.0/build 3674.
- Open F1 and reread the help entry for the chosen business.
- Discard days affected by a quit-related automation failure.
- Label every imported chart with its tested version.
- Keep claims local to the tested business and neighborhood.
| Topic | Observation | Operational meaning |
|---|---|---|
| Evidence | Use | Do not infer |
| Official patch note | Establish current fixes and release context | An optimal schedule |
| Developer reply | Establish intended hourly variation and Insights workflow | Exact profit in this save |
| Versioned test | Choose candidate hours for experiments | Permanent universal hours |
| Live F1/UI | Resolve current capacities and requirements | Future patch behavior |
The decision unit is one staffed hour
An opening-hour decision is marginal: keep an hour only when the sales or service contribution attributable to that hour justifies the labor and other costs caused by opening it.
Use the working formula hourly contribution = hourly revenue minus cost of goods sold minus wages scheduled for that hour minus incremental cleaning, security, and marketing allocation minus an allowance for losses caused by queues or stockouts. Rent is usually committed for the lease, so it is useful for judging the whole business but not for deciding whether one extra late hour should be staffed.
For retail, estimate cost of goods sold from units sold by SKU multiplied by landed unit cost. For office services, use direct labor and any service-specific consumable or workstation cost that genuinely changes with activity. Do not call revenue profit, and do not compare a high-margin jewelry hour with a low-margin supermarket hour using customer count alone.
Treat boundary hours carefully. The first open hour may include startup effects and the last may suffer because the customer curve is already declining. Test moving the boundary by one hour while keeping prices, promotion, fixtures, and staffing stable. Compare like weekdays rather than a Monday after a price change with a Saturday under a campaign.
Example: a shop records $620 revenue from 21:00–22:00, $260 estimated goods cost, $210 direct wages, and $45 of incremental guard and cleaner expense. Its measured contribution is $105 before uncertainty. That hour is promising, but the owner should repeat it across several comparable days and retain a safety margin before extending later.
A zero-customer hour is easy to remove only after verifying that the shop was actually operational. An unstaffed checkout, missing required furniture, empty display, closed access path, or unhappy absent worker can imitate zero demand. Diagnose operation first and demand second.
Action checklist
- Compute contribution rather than reading revenue alone.
- Separate committed daily costs from costs caused by the extra hour.
- Compare the same weekday across repeated weeks.
- Hold price, marketing, and layout steady during a boundary test.
- Investigate operational faults before declaring demand absent.
| Topic | Observation | Operational meaning |
|---|---|---|
| Field | Record | Reason |
| Customers | Per hour from Yesterday graph | Shows realized traffic |
| Revenue and units | Hour if observable; otherwise daily with notes | Supports margin estimate |
| Direct wages | Every role required in that hour | Captures extension cost |
| COGS | Units times landed cost | Separates sales from contribution |
| Fault flags | Queue, stockout, absence, capacity | Prevents false demand conclusions |
Build a seven-day baseline before trimming
The safest discovery method is to operate a complete candidate schedule for one full Monday-through-Sunday cycle, then remove weak blocks only after the business has stabilized.
Choose a deliberately broad but affordable schedule. A mature supermarket or fast-food operation may justify a 24-hour discovery week; a fragile startup should use a shorter span around expected demand because an experiment that causes insolvency teaches little. Every open hour must have a complete minimum crew and all required equipment.
Freeze controllable variables for the baseline week. Do not change price, marketing, layout, employee skill mix, or logistics targets unless a failure would otherwise close the business. Write every unavoidable change into the log so the affected before-and-after days are not treated as a clean comparison.
At the end of each game day, copy the Yesterday customer graph into a ledger by weekday and hour. Add total revenue, COGS estimate, wages, shortage warnings, queue complaints, satisfaction, promotion, and capacity. The hour graph explains timing; the daily financial view explains whether that timing made money.
Mark each hour green, amber, or red. Green means repeated positive contribution with operating buffer; amber means small, volatile, or confounded results; red means repeated negative contribution or no customers despite full operation. Keep amber hours for another cycle rather than optimizing from one noisy day.
After seven days, trim only the outer red block, never isolated middle hours that would create awkward split shifts without checking employee availability. Run another complete week and compare the same weekdays. The goal is a stable schedule, not the narrowest possible opening window.
Action checklist
- Start with a financially survivable broad schedule.
- Maintain a complete crew for every tested hour.
- Freeze price, promotion, layout, and logistics settings.
- Transcribe Yesterday results before they roll forward.
- Retest amber hours and every changed boundary.
| Topic | Observation | Operational meaning |
|---|---|---|
| Status | Definition | Action |
| Green | Positive contribution on repeated comparable days | Keep and staff to observed load |
| Amber | Near break-even, volatile, or confounded | Repeat without other changes |
| Red | Negative or empty while fully operational | Trim from an outer boundary |
| Invalid | Stockout, absence, queue, patch bug, or closure | Repair and rerun |
Read the customer graph without overfitting
BizMan’s hourly graph reports realized customers, which combine demand with every bottleneck in the business. It is evidence, but it is not an isolated demand meter.
A flat line at a capacity ceiling can mean demand exceeds throughput. In that case, extending hours may spread sales, but adding the missing capacity or staff may be more profitable. Read the building limit and required-item capacities before interpreting a saturated graph as the natural market curve.
A sudden trough surrounded by strong hours often signals an operational event rather than genuine preference. Check employee schedules, break boundaries, register coverage, workstation validity, product stock, cleaning, access, and security staffing. If the trough repeats on the same weekday with no fault, it becomes stronger timing evidence.
Graphs from days when the player stood inside the store deserve a note. Developer explanations distinguish simulated operation from physical customer movement while the player is present. Congestion, blocked paths, and queue geometry can suppress the on-site result even when off-site simulation would use formulaic throughput.
Use smoothing conservatively. A three-week median for each weekday-hour is more robust than the best day or the arithmetic mean distorted by an outage. Preserve raw values so a patch, competitor opening, or price change can be identified rather than hidden inside an average.
Do not convert a graph percentage from a community database directly into customers or dollars. Percentages may describe a normalized curve, not your building cap, demand share, price acceptance, or promotion. Use them to choose likely windows and validate them in the current save.
Action checklist
- Check whether bars are capped by capacity.
- Annotate absences, stockouts, and schedule gaps.
- Note whether the player was physically present.
- Retain raw weekday-hour observations.
- Use external curves as test order, not final orders.
| Topic | Observation | Operational meaning |
|---|---|---|
| Graph shape | Likely explanations | Next check |
| Flat at ceiling | Capacity or staffing bottleneck | F1 requirements and staffed fixtures |
| Single-hour hole | Shift gap, stockout, or obstruction | Schedules and operational alerts |
| Late taper | Natural demand decline or price sensitivity | Repeat boundary test |
| Whole-day collapse | Closure, missing requirement, patch fault, competition change | Business status and build |
Translate demand into a minimum viable crew
A schedule is valid only when the business can deliver its full customer journey. One omitted role can make a cheap-looking hour economically meaningless.
List the roles that must be present for the business to transact: checkout or service workers, production staff where needed, and any required supporting role. Then add guards where the security model calls for them and cleaning coverage sufficient to prevent dirt from depressing satisfaction. A scheduled body is useful only if assigned to a valid workstation.
Build the minimum crew from the operational bottleneck. For a retail hour, match checkout coverage to observed customers and basket behavior, then verify every sold category has stock. For an office hour, match staffed service desks to realized clients rather than the lease’s maximum capacity. For food, include every required production and service station.
Calculate labor burden as the sum of each employee’s hourly wage multiplied by hours scheduled, not a generic average wage. Include expensive specialists individually. If extending one hour forces a worker into overtime-like scheduling conflicts, violates desired weekly hours, or requires hiring a whole additional employee, use the actual incremental cost.
Use staggered shifts around peaks. An opening shift can prepare the business, a core shift covers the middle, and a closing shift handles late demand and cleanliness. Verify handoffs do not leave a minute in which a required station is unstaffed; customer graphs aggregate by hour and may hide a brief but damaging gap.
Employee satisfaction constraints are design inputs, not nuisances to ignore. Full-time and part-time wishes, preferred days, and demanded items can make a theoretically optimal schedule unstable. A slightly shorter schedule with reliable satisfied staff may outperform a fragile maximum-hours plan.
Action checklist
- List every transaction-critical role.
- Verify each scheduled employee has a valid station.
- Cost the actual workers used in the extension.
- Inspect handoffs for uncovered minutes.
- Reconcile the plan with employee hour and day demands.
| Topic | Observation | Operational meaning |
|---|---|---|
| Business pattern | Minimum-hour question | Common hidden cost |
| Retail | Can stocked customers check out? | Second cashier and guard coverage |
| Office | Are valid service desks staffed? | Highly paid idle specialist |
| Food | Can production and sale both occur? | Parallel stations and cleaning |
| Night venue | Is demand worth a late full crew? | Late security and shift fragmentation |
Retail schedules: stock and checkout qualify the result
Retail opening tests are invalid when shelves, displays, baskets, or registers constrain customers. Hours cannot compensate for a broken inventory path.
Before each test day, verify total store stock and display capacity for every sold SKU. In 1.0, stock is distributed more evenly among appropriate fixtures, but that does not create inventory. A display that reaches zero may refill from store storage under the game’s refill rule; insufficient reserve can therefore appear as a late-day demand collapse.
Set logistics targets to cover display capacity plus sales expected before the next successful delivery. If a late hour repeatedly lacks goods, increase the relevant target or delivery cadence before closing early. The diagnostic sequence is stock first, staffed checkout second, capacity third, and natural hourly demand last.
Paper bags and other support stock can affect retail operation. The 1.0 release specifically changed distribution across multiple fixtures, so inspect all registers rather than assuming one box belongs to one station. Record support-item alerts with the same seriousness as a missing product.
Use units and gross contribution by product where practical. One late customer buying a high-margin item may be worth more than several early customers buying low-margin goods. Conversely, a busy supermarket hour can look healthy while shrink, low margins, and a large checkout crew make contribution weak.
Scenario: a clothing store shows strong customers until 18:00 and nearly none afterward. The closing shift is complete, but one clothing category is empty and its capacity indicator is limiting the business. Restocking restores evening traffic; trimming the schedule would have treated an inventory fault as a demand fact.
Action checklist
- Check every product and support item before opening.
- Set targets above combined display capacity when needed.
- Confirm all required checkout stations are staffed.
- Log stockout and low-inventory alerts by hour.
- Compare contribution by product mix, not customers alone.
| Topic | Observation | Operational meaning |
|---|---|---|
| Symptom | Retail check | Decision |
| Late sales fall | Store and display stock by SKU | Restock before changing hours |
| Customers cap | Building and item capacity | Add justified capacity |
| Queue complaints | Staffed register throughput | Add or stagger cashier |
| Traffic but weak profit | Product margins and shrink | Adjust assortment, price, or security |
Office and service schedules: sell utilized seats
Office businesses serve customers digitally, yet a current developer reply confirms that traffic index still matters. Their schedule should be sized around billable use of staffed workstations.
For each hour, compare customers served with valid staffed desks. Seat utilization = customers served divided by staffed service seats for that hour. If the live help states a different service cadence for the role, adjust the denominator. The metric is diagnostic, not a promise that every seat can always serve one customer.
Estimate contribution per staffed seat-hour = service revenue attributable to the hour minus the direct wage of that seat minus incremental support allocation. A prestigious large office filled from day one can lose money simply because demand cannot occupy the payroll.
Start with a small complete team and broad enough hours to reveal the curve. Add a workstation and worker only when repeated peaks approach current staffed throughput and price satisfaction, promotion, traffic, and demand are not the limiting factors. Remove idle boundary shifts before abandoning the district.
Travel Agencies and Event Planning Agencies joined the existing Law Firm, Graphic Designer, and Web Development Agency categories before 1.0. Their service curves and employee economics are not interchangeable. Test each category separately even if a blueprint can place similar desks.
Scenario: a 50-capacity office has twelve specialists working all day but serves three to five clients in most hours. The building is not the problem. Reduce simultaneous seats, concentrate shifts in proven periods, check price and promotion, then expand only after utilization and lost-demand evidence justify it.
Action checklist
- Count staffed valid service seats by hour.
- Calculate seat utilization and contribution.
- Keep each office category in a separate dataset.
- Scale workers after measured saturation, not lease size.
- Check traffic, demand, price, and promotion before extending hours.
| Topic | Observation | Operational meaning |
|---|---|---|
| Metric | Formula or source | Interpretation |
| Seat utilization | Customers divided by staffed seats | Low values flag idle payroll |
| Seat contribution | Hourly service revenue minus direct labor and allocation | Tests whether the hour pays |
| Building capacity | Live premises/F1 value | Upper bound, not hiring target |
| Realized customers | BizMan Yesterday graph | Combined result of market and operation |
Nightlife and other strongly time-shaped concepts
Developer guidance uses nightlife as the clearest example of a business whose useful hours lean toward night and early morning. That is a direction for testing, not permission to ignore the ledger.
A nightclub should be given a candidate window covering evening, midnight, and the early-morning tail. Record days separately because Friday and Saturday behavior may differ materially from Monday. Do not average all seven days before deciding whether the weekend deserves later closing.
Midnight-crossing schedules require special attention. Verify that shifts actually cover both sides of midnight, that the displayed opening period matches the intended day, and that cleaners and guards remain assigned while customers are present. A schedule that visually looks continuous can contain a day-boundary gap.
Late hours often require a full operational stack even when customer count is lower: service staff, production or bar staff, security, and cleaning. This creates a steeper break-even threshold than a shop where one cashier can safely cover the tail. Calculate the actual stack rather than applying daytime average labor.
A venue can have a strong hourly curve and still be a poor business if pricing, capacity, or operating costs are wrong. Hours optimize an otherwise coherent operation; they do not rescue a concept with insufficient demand, expensive idle capacity, or chronic dissatisfaction.
Scenario: Friday remains profitable through 03:00, while Tuesday turns negative after midnight. Use day-specific closing times and employee groups rather than forcing one weekly schedule. Retest after meaningful market or price changes because the curve can shift.
Action checklist
- Test evening through early morning across all weekdays.
- Audit both sides of midnight for coverage.
- Include the entire late-hour crew in cost.
- Keep weekday schedules distinct when evidence supports it.
- Reopen the test after a major market change.
| Topic | Observation | Operational meaning |
|---|---|---|
| Question | Evidence | Action |
| Which nights run late? | Weekday-hour contribution | Set day-specific closing |
| Is midnight covered? | Schedule blocks across day boundary | Repair gaps |
| Why is a busy hour weak? | Full crew and security cost | Trim or raise contribution |
| Did demand move? | New repeated week after market change | Recalibrate |
Security, cleaning, and support coverage
Support roles are part of the opening-hour decision because they may be required only while the business is trading, and their absence can corrupt both satisfaction and loss data.
Developer guidance says guards protect during open hours. For a high-value store that needs simultaneous guards, extending one hour adds both wages, not a fractional daily average. Confirm camera and detector coverage as well; a nominal employee schedule cannot repair an incomplete security setup.
Read theft expense in the individual store’s EconoView and compare equivalent open-hour blocks. Security’s economic value is avoided loss minus incremental guard and equipment cost, with a separate note for margin lost when stolen stock cannot be sold. Do not assume security increases demand directly.
Cleaning can be scheduled before, during, or after peaks depending on the business, but a dirty store during a test changes satisfaction and realized traffic. Record dirt or cleaning warnings. If a cleaner works after closing, allocate that time to the day that created the cleaning load.
Support workers may have different hour preferences from sales staff. Build their shifts first around mandatory coverage, then test the customer-facing boundary. Extending customer hours without matching support coverage creates an invalid comparison and can trigger staff dissatisfaction.
Scenario: a jewelry shop’s final hour produces acceptable gross margin before security but becomes negative after two guards are included. The correct choice may be an earlier close, not reduced protection during trading. Compare avoided theft and contribution before deciding.
Action checklist
- Cover every open minute with required security.
- Read store-specific theft expense in EconoView.
- Log dirt and cleaning warnings.
- Allocate after-close support cost to the operating day.
- Never test an extension with deliberately incomplete protection.
| Topic | Observation | Operational meaning |
|---|---|---|
| Support cost | When to attach it | Evidence screen |
| Guard wages | Every protected open hour | Schedule and employee wage |
| Camera/panel cost | Amortized across test horizon | Installed equipment and coverage |
| Cleaning wages | Hours caused by operation | Schedule and cleanliness |
| Theft | Observed per business and period | Store EconoView |
Change one boundary at a time
Controlled schedule changes reveal causality. Simultaneously changing opening, price, advertising, staff, and inventory makes any result impossible to attribute.
Choose one boundary—opening one hour earlier or closing one hour later—and leave the opposite boundary unchanged. Run at least the same weekdays used in the baseline. A complete week is preferable when weekend behavior differs, and multiple weeks are warranted for small marginal values.
Use a paired comparison: baseline Tuesday 08:00–09:00 versus test Tuesday 08:00–09:00 under the same price, promotion, capacity, and staffing standard. Report the difference in customers, revenue, COGS, wages, faults, and contribution. Do not compare total daily profit alone because other hours can mask the boundary.
Set an uncertainty buffer. If measured contribution is barely positive and normal variation could erase it, classify the hour amber. The buffer can be a player-selected percentage of wages or expected margin; document the rule before testing so it is not changed to justify a desired answer.
After a successful one-hour extension, test the next hour separately. Demand often declines nonlinearly, while the staffing requirement can jump when the extension no longer fits existing employees. The first profitable late hour does not prove the entire night is profitable.
If two changes cannot be separated—for example, an additional worker is required to extend at all—treat them as one intervention and price the whole package. State that the result measures schedule-plus-hire, not opening time in isolation.
Action checklist
- Alter one opening or closing boundary.
- Pair identical weekdays.
- Record hourly deltas, not only daily totals.
- Define an uncertainty buffer in advance.
- Price mandatory bundled changes together.
| Topic | Observation | Operational meaning |
|---|---|---|
| Test | Control | Pass condition |
| Open one hour earlier | Prior same weekday | Positive repeated contribution after buffer |
| Close one hour later | Prior same weekday | Positive repeated contribution after full crew cost |
| Add a worker for extension | Existing schedule | Package contribution exceeds package cost |
| Second extension hour | First extension retained | Independent positive result |
Common early mistakes
Most bad schedules come from reading one visible number as the whole system. The following errors are especially expensive because they create confident but false conclusions.
Copying an online ‘best hours’ table ignores version, district, competition, and operation. Even a rigorous 1.0 dataset is a starting prior. Enter its suggested window as a candidate, then verify the current save with BizMan.
Running 24/7 by default mistakes revenue for profit. Some business types can support continuous trade, as developers have said, but every overnight hour must carry its own labor, support, and inventory burden. Empty staffed hours are not marketing.
Closing immediately after one weak day overfits noise. A stockout, absence, queue, price move, or build-3674 transition issue can depress the result. Require repeated valid observations and separate weekdays.
Scheduling to the building’s customer capacity mistakes an upper limit for forecast demand. Hire and open around observed throughput. Capacity is useful when bars saturate, not as a command to employ enough people for the theoretical maximum.
Ignoring employee wishes creates churn and missing shifts that then look like bad demand. A stable roster with lawful, desired hours is part of the tested configuration. Recheck assignments after training or moving employees between businesses.
Action checklist
- Reject unversioned universal schedules.
- Cost every overnight support role.
- Require repeated valid days.
- Treat capacity as a ceiling.
- Audit satisfaction and assignments after schedule edits.
| Topic | Observation | Operational meaning |
|---|---|---|
| Mistake | Misleading observation | Correction |
| Universal hours | Another save reports profit | Run local controlled week |
| Revenue-only decision | Busy hour appears good | Subtract COGS and incremental labor |
| One-day trimming | Random dip appears structural | Repeat same weekday |
| Capacity hiring | Large lease triggers large payroll | Scale from realized use |
| Ignored wishes | Absence looks like no demand | Stabilize roster |
Troubleshooting decision tree
When an hour underperforms, move through operation, capacity, market, and economics in that order. This prevents expensive strategic changes in response to a simple fault.
First ask whether the business was open and complete. Confirm the opening schedule, employees physically assigned to the right business, valid workstations, all required furniture, checkout or service coverage, and no inaccessible entrance. If any answer is no, repair it and discard the observation.
Second ask whether customers could be served. Inspect queues, building capacity, product or service capacity, stock by SKU, support items, dirt, and security. A saturated bar or warning identifies throughput, not weak demand. Correct the bottleneck and rerun.
Third ask whether the market proposition was stable. Compare district demand and providers in MarketInsider, traffic index, promotion, price satisfaction, and recent competitor changes. Build 3674 fixed MarketInsider opening the correct current neighborhood, so verify the displayed district before recording.
Fourth calculate contribution. If operation and market are healthy but the hour loses money after direct costs, trim it or redesign the staffing block. If contribution is positive but volatile, extend the sample. If positive and robust, retain it.
Finally inspect patch and save integrity. Update to build 3674, advance a day after the employee-quit automation fault if relevant, save, reload, and reproduce. Submit an F2 report for a persistent state that contradicts F1 and repeated tests.
Action checklist
- Open and complete? Repair or continue.
- Throughput constrained? Remove bottleneck.
- Market stable? Rebaseline if changed.
- Contribution positive after buffer? Keep or trim.
- Persistent contradiction? Reproduce and report.
| Topic | Observation | Operational meaning |
|---|---|---|
| Node | Yes | No |
| Fully operational? | Check throughput | Repair; invalidate hour |
| Capacity or stock constrained? | Fix and rerun | Check market |
| Market conditions stable? | Calculate contribution | Start new baseline |
| Contribution robustly positive? | Keep hour | Trim or retest amber |
| Behavior reproducible on 3674? | Document result | Treat as transient |
Management-screen record sheet
A compact ledger turns the schedule from intuition into an auditable operating policy. Capture values before they roll out of Yesterday or are overwritten by later changes.
Identify the row by save date, game day, weekday, business, neighborhood, build, test label, and whether the player was on site. Record scheduled opening and actual operational window separately, because an absent critical employee can shorten the latter.
For every hour, record customers, staffed transaction positions, capacity or queue flag, stockout flag, and unusual event. Attach daily revenue, sold units or services, estimated COGS, wages by role, marketing, cleaning, security, theft, and net result.
Snapshot market controls: current prices, price satisfaction, promotion, traffic index, district demand, and provider count. These fields explain why an old schedule may cease to fit. Mark any Pricing Manager change with the effective day.
Record employee controls: names or role groups, skill, hourly wage, requested weekly hours, full-time or part-time status, satisfaction, and absence. When a schedule changes, note whose hours moved and whether another hire became necessary.
Store the decision statement in plain language: ‘Close Tuesday at 22:00 because 22:00–23:00 lost money on three valid Tuesdays after full security cost.’ A reversible, evidence-linked rule is easier to audit than a color-coded cell with no rationale.
Action checklist
- Record build, day, business, and neighborhood.
- Separate scheduled from actually operational hours.
- Capture hourly customers and fault flags.
- Snapshot prices, market, and staffing.
- Write a testable decision sentence.
| Topic | Observation | Operational meaning |
|---|---|---|
| Ledger group | Minimum fields | Refresh |
| Identity | Build, date, weekday, business, district, test | Every observation |
| Hour | Customers, seats/registers, faults | Daily |
| Finance | Revenue, COGS, wages, support, theft, net | Daily/weekly |
| Market | Price, satisfaction, demand, providers, promotion, traffic | Before each test |
| People | Skill, wage, wishes, absence, assignment | After roster change |
Final opening-hours audit
The final audit is a release gate for the schedule. Pass every item before calling the plan optimized for the present save.
Confirm the evidence window contains at least one complete valid week and repeated observations for every marginal boundary. Remove days affected by stockouts, missing roles, severe queues, blocked layouts, unintended price changes, or the pre-hotfix quit bug unless the purpose was to test that fault.
Verify each open period has the required staff, workstations, inventory, support items, cleaning, and security. Check midnight crossings day by day. Compare scheduled headcount with the customer graph so peaks receive enough service without paying idle seats through weak tails.
Recalculate boundary contribution with current landed costs and wages. Prices, insurance, worker skill, and supplier economics can change; an hour that once cleared the buffer may not continue to do so. Keep committed rent separate from the marginal open-hour decision but include it in the whole-business review.
Review MarketInsider and the current neighborhood after competitors open, your own chain expands, or a major campaign changes. Rebaseline after a material price, promotion, layout, product, or capacity alteration. A schedule is a versioned policy, not permanent truth.
Archive the exact opening grid, rationale, last test date, build, and next review trigger. Useful triggers include two consecutive weeks below the margin buffer, a patch affecting demand or staff, a new competitor, repeated queues, recurring stockouts, or a roster change that makes the old boundary expensive.
Action checklist
- One valid full week plus repeated boundary tests.
- No unresolved operational confounders.
- Every open hour fully staffed and protected.
- Current costs used in contribution calculation.
- Schedule, rationale, build, and review trigger archived.
| Topic | Observation | Operational meaning |
|---|---|---|
| Audit item | Pass | Fail response |
| Evidence quality | Comparable valid days | Repeat baseline |
| Operational coverage | No missing requirement | Repair schedule/layout |
| Economics | Boundary clears chosen buffer | Trim or redesign |
| Market currency | Current district and prices recorded | Rebaseline |
| Version control | Build 3674 and date recorded | Label or retest |
Research ledger
These links establish mechanics or provide a reproducible lead. Any balance-sensitive number still has to be checked in the current save.
Questions answered
Is 24/7 opening always best for supermarkets or fast food?
No. Developer replies say some examples can operate profitably around the clock, but that is capability, not a universal optimum. Test every staffed hour in the current district and subtract COGS, wages, cleaning, security, and fault losses.
Where is the hourly customer view?
Open BizMan, select the business, open Insights, select the customer graph, and choose Yesterday. Transcribe it daily because later days replace the immediate evidence. Pair it with schedules and finances.
How many days should I test?
Use at least a complete Monday-through-Sunday baseline, then repeat changed boundaries on matching weekdays. Extend the sample when contribution is close to break-even or the result is volatile.
Why does an hour show zero customers despite high demand?
Check actual opening, required staff and stations, stock, support items, queues, access, capacity, cleanliness, and patch state first. Zero realized customers can be an operational failure rather than zero demand.
Should I use the same hours every weekday?
Only if the recorded curves support it. Developer guidance explicitly distinguishes weekday, weekend, morning, and evening behavior. Nightlife in particular often benefits from day-specific closing times.
Does traffic index stop mattering for office businesses?
No. A December 2025 developer reply states that traffic index still affects office businesses even though customers are handled digitally. Treat office seats as paid capacity that needs measured utilization.
Does build 3674 change demand?
Its published fix is not described as a demand rebalance. It fixes employee-quit states that could stop scheduled systems. Affected pre-fix days should be discarded because operational failure can distort the graph.
When should the schedule be reviewed again?
Rebaseline after a relevant patch or material change to price, promotion, competitors, layout, capacity, product mix, logistics, security, or roster. Also review after two weak weeks or repeated queues and stockouts.