“How much is IoT” is not a useful question. Gates into HR, sensors into tickets, and scans into stock are different jobs. The table below puts budget and cycle in a checkable range so a cheapest box or a luxury platform does not steer you.
| Type | Typical cycle | Where the money goes |
|---|---|---|
| Single-scene access (e.g. access events into HR, protocol open) | 4–8 weeks | Integration, rights sync, failure alerts, pilot |
| Alert-to-ticket loop (several sensors + existing maintenance) | 8–14 weeks | Rules, dispatch, suppress/escalate, close checks, training |
| Multi-system (access + energy MES + scan WMS) | 3–6 months | Site inventory, protocol, offline catch-up, reconcile and staged go-live |
These are experience bands, not a quote. Hardware is extra; you can name the brand, but software must reach the business: Must we supply the hardware, or can you name the brand. A closed protocol stretches integration: If the device protocol is closed, can we still integrate.
Why do quotes for “doing IoT” differ by several times?
The gap is point count, protocol openness, offline cache, whether exceptions enter tickets, how many systems you join, and who gives a live integration window. Hidden items: a box with no API, no site network, messy HR master data. Client prep often outlasts coding. Steps from brief to launch: From brief to launch: how a hardware-software project runs.
Can we buy the box first and do software later?
You can buy, but write data ownership and the interface in the contract. Otherwise software only orbits a closed cloud—not integration: When a closed IoT box will not release your data. Safer: define the loop, then decide to buy or change devices.
Can we squeeze the cycle to two weeks?
Two weeks usually buys a lab green light or a one-point demo. Weak network, shifts, rights, false-alert convergence and business reconcile will not harden in two weeks. Better one high-consequence exception in phase one than an unmaintainable wall in fourteen days.
How do we hold budget without buying a decorative wall?
Name one acceptably closed loop: access dies on the exit day, or overtemp must open a ticket. Shrink “we want a smart park” to three lines: devices, exception, which system. The gap with and without a loop: With devices in the business vs. without: what exception handling looks like.
Put the integration window and spares on the plan
Beyond the quote: a downtime window, vendor license fees, SIMs and data, certificates, night integration hours. Leave them off the plan and someone asks “why isn’t it done.” Rebuild vs. extend: Rebuild the old IoT platform, or extend it. DaXi Technology returns scope, stages and an acceptance list after the brief—tickets and rights, not a demo animation. To name devices on site and the exception you want closed, go to the Hardware-software service page.
When you compare two vendors, quote the same closed-loop list
A lump-sum contest rewards whoever dares to omit: weak-network catch-up, close checks, HR sync-failure alerts, training and dropping paper. Ask them to quote the lines you actually need among “events into tickets, rights into HR, scans into stock,” and say whether live integration is included. Why integration is not another box: Why companies should connect devices to business systems. Gains mapped to roles pass budget reviews more easily: Hardware-software gains you can actually see.