Purchasing Agent setup and negotiation

Role of a Purchasing Agent in build 3674

A Purchasing Agent is the Headquarters employee who owns an importer contract and converts an approved order plan into inbound stock for a selected warehouse or factory.

The agent is not a shop buyer, a delivery driver, or the manager who distributes warehouse stock. Their responsibility ends at the receiving industrial location. The downstream Logistics Manager, vehicle, and driver move goods from a warehouse to stores or from a factory to a warehouse/export destination. This division is essential in 1.0 because importers operate on a weekly cycle while internal logistics operates daily. A valid retail supply chain therefore needs both purchasing and logistics management unless the player continues to buy manually from a wholesaler.

This guide is scoped to 1.0 build 3674. EA 0.9 changed importer orders from boxes to individual units, removed the general minimum-order floor, added Smart Delivery, imposed per-item weekly limits per receiving location, and moved standard receipts to Monday at 08:00. EA 0.10 refreshed importer product assignments and factory inputs. Build 3674 fixed an employee quit-state problem that could block imports globally, so a save affected on build 3672 should be allowed to advance into the next day after updating before its plans are dismantled.

Complete purchasing dependency chain

The green order button is only one link; every management and receiving dependency must be present at the lock and receipt events.

The chain is an active Headquarters business to a working computer workstation to an assigned and scheduled Purchasing Agent to a Purchasing Agent plan to one importer contract to a selected receiving warehouse or factory to one or more item quantities to an order mode to sufficient cash and compatible free pallet capacity. The developer's July 2026 diagnosis demonstrated that merely returning an employee to the HQ schedule is insufficient: the agent also had to be reassigned to the Purchasing Agent plan before Sunday 20:00.

Downstream readiness should be audited separately. A successful Monday import can still produce empty stores if the warehouse has no Logistics Manager route, assigned vehicle, qualified driver, source stock target, or store capacity. Likewise, factory ingredients can arrive correctly while production remains stopped because a line lacks a recipe, input material, worker, schedule, or output space. Prove the inbound inventory delta first, then move to the next controller rather than calling every shortage a Purchasing Agent failure.

Hiring, workstation, assignment, and schedule

Create the operational seat before negotiating a contract so that the employee is available when the weekly plan locks.

Recruit a Purchasing Agent through the current office/HQ recruitment channel, assign the employee to a Headquarters, and schedule them at a complete computer workstation. A desk or monitor alone is not a complete station; verify that BizMan exposes the workstation and allows the employee to occupy it. Keep employee demands and weekly hours satisfied because an avoidable resignation can remove the plan owner immediately before the ordering event.

Training creates a timing hazard. Sending the agent to training, moving the HQ, or changing their business assignment can remove the employee from the plan even if they later reappear on the office schedule. The documented 2026 case failed because the employee was scheduled but not reinserted into the Purchasing Agent plan before Sunday 20:00. Treat plan membership as an independent field and audit it after every training, move, replacement, resignation, or business conversion.

One contract per agent and importer selection

An agent manages one importer relationship, so product-source coverage determines the minimum management footprint.

The current management model assigns one importer contract to a Purchasing Agent. Product importers occupy the earlier piers, while raw-material importers occupy the later piers after the EA 0.10 expansion, including the opening of Pier 9. Do not rely on a pre-0.10 pier list because cinema products, tea, slushies, and expanded factory recipes changed catalogs. Open F1 Help for the product, inspect the current importer contacts, or visit the pier before committing the employee.

One contract does not necessarily mean one warehouse: the live plan determines eligible destination choices and items, but weekly limits are per receiving location. Scaling often requires additional agents because different importers own required product families, because a second location is needed for per-location limit capacity, or because management risk should be separated. Count distinct importer relationships and receiving-location requirements before renting HQ seats; a large office without enough agents cannot execute extra contracts.

Negotiated price, agent skill, and Import Index

Purchasing economics is set at contract formation, not continuously improved simply because the employee gains skill.

A developer reply identifies two price inputs: Purchasing Agent skill and the Import Index. The index changes daily, and the negotiated result is locked when the contract is made. In the documented behavior, improving the agent later did not retroactively improve the existing contract; realizing the new skill required renegotiation, while the current index could make the replacement contract better or worse. This creates a real lock-in decision rather than a guaranteed benefit from frequent renegotiation.

Let p0 be the displayed base or comparison price, s represent the unknown skill effect, and I represent the current Import Index effect. The game has not published a reliable universal formula, so model the offered unit price as p = F(p0, s, I) and compare the actual offer, not an invented percentage equation. For N ordered units, expected merchandise cost is N times p before any urgent premium. Record old contract price, new preview, index, agent skill, and the break-even volume before replacing a favorable contract.

Weekly order clock and Sunday 20:00 lock

Standard importer planning is a deadline system: configuration must be valid before the order locks, then goods arrive Monday morning.

Official EA 0.9 notes establish Monday 08:00 as the standard delivery event. A July 2026 developer investigation adds the operational cutoff: the employee must be assigned to the Purchasing Agent plan before Sunday 20:00. The order is locked the night before. Do not interpret a Monday morning employee seated at a desk as proof that the Sunday lock had a buyer. Keep the plan stable through the cutoff and preserve enough cash for the charge.

The downstream warehouse already performed its Monday internal delivery at 02:00, six hours before the import. Consequently, Monday's new goods normally cannot reach stores through automated warehouse logistics until Tuesday 02:00. This lag belongs in purchasing quantities: preserve a bridge quantity for Monday's route and do not plan the warehouse to reach zero just before the Monday receipt. An urgent order uses a different path and should be logged with its requested and expected arrival rather than assumed to share the weekly clock.

Smart, recurring, one-time, and urgent orders

Smart and recurring answer different questions: how much to buy and whether the plan repeats.

The developer confirmed that Smart Delivery adjusts quantity toward the selected target; recurrence is separate. Smart plus recurring recalculates and orders the shortfall each weekly cycle. Smart without recurring performs a target-fill only for the manually confirmed order. Recurring without Smart orders the entered fixed quantity again even if old inventory remains. Neither toggle changes standard delivery timing. This distinction prevents the common error of enabling Smart alone and expecting autonomous weekly purchases.

Let T be the smart target and H be inventory counted at the lock. Ignoring constraints, smart quantity Qs = max(0, T - H). Actual receipt is bounded by the item's weekly location limit, cash, and receiving capacity. In a confirmed example, T = 10,000 and H = 9,000 produces 1,000 ordered, while a non-smart recurring quantity of 10,000 would add another 10,000. Urgent delivery is the recovery tool for a midweek mistake; official preview notes introduced it at double cost, but later difficulty/custom settings can alter the penalty, so read the live quoted fee rather than publishing two times as universal for every 1.0 save.

Forecasting a weekly purchasing target

Base the target on complete, in-stock unit consumption and explicit safety stock, not on revenue or an arbitrary round number.

For one item feeding n stores, let a_i be average complete-day sales at store i, H = 7 days, g be expected weekly growth, f be reliable factory or alternate inbound units, and B be safety stock including the Monday bridge. A planning target is T = ceil(H times sum(a_i) times (1 + g) - f + B). If three stores consume 200, 300, and 450 units per day, g = 0.10, f = 1,500, and B = 1,000, then T = ceil(7 times 950 times 1.10 - 1,500 + 1,000) = 6,815 units.

This is an external planning formula, not a game algorithm. Reject data from days when the item was unavailable, the store opened late, price or promotion changed, or capacity prevented sales. Use the game's last-seven-days information as evidence, but remember that sales during shortage are censored. For new products, begin with a conservative affordable target, watch daily store consumption and warehouse outflow, then revise after a complete week. Preserve the old target and reason for each change so a transient spike does not become permanent overstock.

Units, boxes, limits, and storage

The ordering interface uses units, while shelves physically hold boxed inventory with product-specific packing ratios.

EA 0.9 officially changed wholesaler and importer ordering to individual units because box sizes differ by product. Entering 10 means ten items, not ten boxes. Converting a weekly unit target to storage requires the current units-per-box value for that SKU and the current capacity of the chosen pallet furniture. If U is units ordered and k is units per box, required boxes are ceil(U / k). Because k varies and balance changes have altered box sizes, this guide deliberately gives no universal pallet-shelf unit capacity.

Each item also has a weekly ordering limit per receiving location. The per-location qualification matters: a second eligible location may create another independent limit, but it also requires storage, purchasing-plan coverage, logistics, and cash. A nominal smart target above the limit cannot force an oversized receipt. Track target, computed shortfall, allowed weekly amount, and actual receipt separately; otherwise a capped order looks like a Smart Delivery error.

Cash-flow and landed-cost control

A profitable retail margin does not protect an order that lacks liquid cash at the weekly event.

For items j, expected merchandise invoice M = sum(Q_j times p_j). Let u be the live urgent multiplier or surcharge representation, and let O include wages, rent, logistics, and other costs due before the next sales recovery. A safe liquidity requirement is L = M times u + O + R, where R is a deliberate reserve. For a standard order use the displayed standard terms rather than assuming u = 1 if the interface adds fees. For urgent orders read the quote; difficulty settings may change the premium from the original two-times design.

Evaluate unit economics with landed cost rather than importer price alone. Approximate landed unit cost c = negotiated unit price + allocated agent/HQ cost + allocated warehouse/logistics cost + expected urgent and spoilage/overstock cost where relevant. The game does not publish one allocation rule, so state yours. A cheaper contract can still be operationally worse if it forces another warehouse, manager, and vehicle for a small volume. Conversely, a skilled agent with a well-timed index may produce savings large enough to recover the extra HQ seat quickly.

Scaling agents, contracts, and destinations

Add Purchasing Agents when a documented contract or location constraint binds, not simply when store count increases.

The primary gate is importer diversity: one agent cannot own every importer contract. The second is the per-item weekly limit at one receiving location. The third is risk: an agent resignation, plan error, or HQ move can interrupt a large share of supply. The fourth is economics: a separate high-skill agent may lock a better contract for a high-volume category. The fifth is destination topology: factory raw inputs may need to arrive at factories while retail goods go to distribution warehouses.

Create an ownership matrix with one row per contract and columns for agent, HQ, importer, receiving site, items, smart state, recurrence, Sunday status, weekly value, and backup procedure. Keep names descriptive enough that an empty legacy plan is obvious after moving HQ. The developer specifically warned that empty Purchasing Agent plans in another HQ are often missed and can disrupt ordering. Before adding an agent for capacity, confirm whether the existing limit is truly reached and whether factory production or a different source is the intended 1.0 progression gate.

Contract renegotiation procedure

Renegotiate as a controlled investment decision because the current index can erase the benefit of higher employee skill.

Before terminating anything, export or record every item, destination, target, Smart toggle, recurrence setting, and current contract price. Record agent skill and today's Import Index. Estimate expected volume over a chosen horizon and calculate gross price benefit G = sum(volume_j times (old_price_j - offered_price_j)). Subtract any transition cost, duplicate overhead, and expected shortage loss. Accept only when the result and operational timing justify the change; a small unit saving immediately before Sunday lock may not compensate for configuration risk.

Perform the change early enough to rebuild and audit the plan before Sunday 20:00. Use one low-risk item to verify the next receipt and retain enough warehouse stock to survive a missed cycle. If the interface does not provide a reversible preview, make a save before termination. The 2023 developer reply confirms the strategic intent of locking a favorable index, but any current UI behavior that differs should override the historical procedure and be documented as a build 3674 retest.

Diagnostic procedure for a missed import

Diagnose the scheduled event backward from Monday 08:00 and preserve timestamps.

First determine whether the order was standard or urgent and whether the expected time has actually passed. For a standard order, verify that the plan was valid before Sunday 20:00: active HQ, agent assigned to HQ, scheduled computer, agent inserted in plan, contract active, destination selected, positive item quantities, correct Smart/recurring state, and the order actually confirmed where required. Then check cash, per-location item limit, and receiving pallet capacity. Read the exact notification rather than inferring a cause from an empty shelf.

If the warehouse received goods at 08:00 but a store is empty, the purchase succeeded. Check the daily 02:00 logistics route and remember that Monday's route occurred before the import. If every import and delivery failed on a build 3672 save, update to 3674 and advance to the next day before retesting because the hotfix repaired a shared employee-state blocker. If only one item is absent, inspect its target, source eligibility, limit, and counted on-hand inventory.

Reproducible smart-order test

A one-SKU controlled week can establish what the live build counts and prevent argument from impression.

Choose a low-cost item, one warehouse, and no factory input for that SKU. Before Sunday 20:00 record build, contract, destination, target T, counted warehouse on-hand H, Smart state, recurrence, current limit, free storage, unit price, and liquid cash. Predict Q = max(0,T-H), then apply limit, cash, and capacity bounds. Keep the agent assigned through lock. At Monday 07:59 record inventory again; at 08:01 record receipt and transaction. Do not let a 02:00 route confuse the baseline without recording its outbound quantity.

Repeat next week after creating a known change, such as reducing H by a measured amount while holding T fixed. With Smart plus recurring, the new receipt should follow the new shortfall. For a comparison, use another low-cost SKU with recurrence on and Smart off; it should repeat its fixed quantity rather than target-fill. If results differ, save screenshots and submit an F2 report containing exact expected and observed values. Do not generalize from a test where stock was sold, moved, or produced during the measurement window.

Tracking fields and weekly control routine

A Purchasing Agent ledger should reconcile contract economics, order calculation, physical receipt, and downstream use.

Maintain one row per contract-item-destination combination with build, week, HQ, agent, skill, importer, destination, Import Index at contract, locked unit price, smart target, recurrence, Sunday pre-lock inventory, predicted quantity, current weekly limit, free boxes, forecast invoice, actual receipt, actual charge, and warning. Add complete-day aggregate consumption, stockout hours, target rationale, and next review date. This turns a missed shipment into a specific variance instead of a vague claim that purchasing did not work.

Review Thursday or Friday for demand changes, Saturday for cash and capacity, and Sunday before 20:00 for plan ownership. On Monday distinguish 02:00 warehouse outflow from 08:00 importer inflow, reconcile the invoice, and update days of cover. Change targets only when the reason is recorded: new store, changed hours, pricing, promotion, competitor shift, shortage correction, factory substitution, or deliberate buffer change. Avoid weekly oscillation caused by reacting to one abnormal day.

Minimum viable purchasing pilot

A deliberately small first contract reveals the clock, cash charge, target calculation, and storage conversion before the empire depends on the result.

Choose one inexpensive SKU whose source is unambiguous in the current F1 Help. Use one active agent, one contract, one receiving warehouse with obvious free capacity, and no competing inbound source. Set a target that creates a visible but affordable shortfall, enable the intended Smart and recurrence combination, and preserve the agent-plan binding through Sunday 20:00. Record the quoted unit price, index, target, pre-lock inventory, expected quantity, and expected total charge. Capture any 02:00 outbound movement separately so Monday's receiving baseline is not contaminated.

At Monday 08:00 reconcile three independent values: inventory added, transaction charged, and the plan's expected quantity after limits. Do not declare success solely because a notification appeared. Continue to Tuesday 02:00 and verify that the downstream warehouse route can use the new stock. A successful pilot proves purchasing and retail replenishment as separate stages. A failed pilot remains cheap and produces a compact evidence set for diagnosis or F2 reporting.

Promote the contract gradually: add products whose box ratios and limits have been checked, then add destinations or additional contracts. Preserve at least one weekly cycle between major changes so each variance has a plausible cause. The pilot is complete only when the Sunday lock, Monday receipt, storage, invoice, and Tuesday distribution reconcile. This discipline is especially valuable after an HQ move, agent replacement, contract renegotiation, difficulty change, or patch that mentions imports, employees, or management plans.

Version caveats and excluded claims

Historical mechanics remain useful only where later official changes have not superseded them.

Reject guides that describe daily standard importer deliveries, quantities in boxes, a universal minimum order, or one global weekly cap. EA 0.9 explicitly changed those mechanics to Monday 08:00, individual units, no general minimum, and per-item limits per location. Reject pre-0.10 product and pier maps because importer catalogs expanded. Treat the 2023 contract-price lock reply as developer evidence but verify the current contract UI; the underlying statement predates several economy revisions even though no accepted source in this audit announces its removal.

Do not publish exact importer prices, product caps, urgent multipliers, units per box, or agent wage as universal build 3674 facts. They can depend on difficulty, custom settings, Import Index, negotiation, product, and balance. Do not claim that Smart chooses a safe target, predicts demand, or schedules more frequent deliveries: it calculates quantity toward the player's target. Do not claim recurrence is implied by Smart. Finally, do not diagnose a Monday store shortage from an 08:00 warehouse screenshot without accounting for the earlier 02:00 outbound event.