Ask “how much for a system” and you get a guess. Wholesale inventory and install dispatch are both “business systems”; documents, states, and the phone are not. The ranges below put a checkable budget on the table — not the lowest subscription or a bloated all-in quote.
| Phase-one scope | Typical cycle | Where the money goes |
|---|---|---|
| One-warehouse inventory: purchase, sales, stock, basic reconcile | 8–14 weeks | Documents and permissions, stock locks, pilot on real orders |
| Work-order close: repair, dispatch, parts, done + follow-up | 8–12 weeks | State machine, mobile, write-back to order/stock |
| Customer contracts + cash nodes (project or terms) | 6–10 weeks | Contract model, apply, sales/finance views |
| The above plus finance / courier / legacy ERP | 3–6 months | APIs, match rules, old-data clean |
These are experience ranges, not a quote. Packaged SaaS can be cheaper and faster; when fields miss the floor you are back on two tracks. Compare When SaaS does not match how the floor actually works and How packaged software actually differs from a custom system.
Why can two “inventory” quotes differ by several times?
Warehouse count, price rules, batch/expiry, whether negatives are allowed, approval layers, mobile pick, how old sheets migrate, whether finance connects. The hidden item: how many exceptions — free goods, split shipments, lends, verbal rush. More exceptions means change the floor or write them in — both show up in person-days. Time the customer spends on rules and pilot orders is often longer than coding.
Can we launch in two weeks?
Two weeks usually means configure packaged software or one very narrow document. Walking intake, lock, outbound, and reconcile — plus training and a real-order pilot — will not be solid in two weeks. If a bid or a key account is pushing, launch the main path as phase one. Do not ship a shell that still runs on WeChat. Why groups and sheets fail is in Why a company cannot run operations on spreadsheets and WeChat alone.
Who cleans old data and sheets? Will it slip?
Yes — a common slip. Duplicate SKUs, duplicate customer names, mixed units become “legal” stock once imported. Decide in the plan: what must be cleaned, who signs, how much history. Full history does not have to enter phase one. Open in-flight documents must.
How do we control budget without building a fill-in ledger?
Pick the main job: inventory, work orders, or contract cash. Phase one is only the documents and permissions that hold that job — outbound writes the book, a work order can close. Boards and kitchen-sink modules wait. ERP first vs. custom modules is in Should a mid-size company buy ERP first, or start with custom modules. The growth step that requires an operations system is in Why growth requires an operations system of your own.
Count hidden cost or a low price doubles later
Outside the quote: on-site discovery, barcode gear, lining old sheets to codes, finance mapping, sales and warehouse on the pilot, two weeks on site after launch. If those are not in the plan they become “why isn’t it done.” Compare vendors on the same document list, and write whether migration, training, and pilot are included. A total-price fight rewards whoever left items out. Stage it: prove lock and outbound, then decide finance integration. That is less risk than signing everything once.
After work is clarified, DaXi Technology gives scope, stages, and an acceptance list — not an uncheckable total first. Spell out warehouse count, documents, and whether a legacy system exists. Go to Business systems service page and submit. Accept against visible efficiency, not a feature menu. See The efficiency a business system can make visible.