How does the fixed assets management system requirements fit? Elements for preparing the demand list
Before you do the fixed asset management system, demand combing this most common card. Some people come up and copy the list of functionality that others have in their systems, and they find out that business parties don't even accept a bill; some stick in business interviews, take dozens of pages of minutes, and then turn into scattered points. And there is also a more common situation: business says "I want a lot," developers say "I want to do it not finish", and no conclusion for an afternoon. The key to demand combing is not to write the function, but to find out what the business side really wants to solve. There are two criteria for determining whether needs are in place: one is to answer "To solve what specifics under which scene" and the other is to be identified or modified by each side. If you can't do both, the list is a waste of paper.
Need to comb the entry point: from business scenes or from data fields?
When many teams get a demand assignment, the first reaction is to find a functional list template, copy these generic features of "assets card, depreciation, inventory counting, statements." After they're done, they either think they have all the names signed or they can't figure out the name. The problem with this approach is that it comes from someone else's system, not from your own management scene.
Another approach is to sit down and talk about first-hand and financial, asset managers, warehouse custodians: buy in assets to end of life, go through several departments, who fills out what forms at each point, and where the common problem is. Like a company that does warehousing logistics, with assets scattered over more than a dozen warehouses, the worst thing to do is "no find on the books, no find on the books," so the core need is to synchronize the swipe count with offline data, not the fancy board.
The test is simple: if you can list at least three specific business pain points (e.g. "Depreciation charge takes three days per month to manually check" "Can't find a signature" Zinventory counting and the financial statements) then push it from the scene; if you just say "Systems to manage assets", don't rush back, talk about two rounds of business.
Structure of the demand list: by module or by role?
The list of requirements is common in two types of structure: unbundling by system module - asset management, depreciation management, inventory management, report management, each module with several functional points; debundling by role - what the asset manager does, what the finance does, what the warehouse keeps, and what the department head does.
The modular approach is clear, it is appropriate for the development team to see the technology come through; role styles are easy to match business perceptions and identify with business at the evaluation sessions. The trade-off between them depends on who is primarily on the list. If one is involved in multisectoral review within one side, we usually recommend a role-based review, a module is a back-up line, and each character needs a section of the action by its business. For example, under the role of "assets manager" you write first what fields are needed for "new buy-in library", then what approval streams are required for "transfer," and what annexes and confirmation links are required for "end-of-life disposal".
The blending has one advantage: business people look at what they're going to do every day, developers see who calls every function and avoid "the functionality is developed but nobody will use it."
The control of the particle size: to the field level or only to the process level?
Requirements documents most commonly go wrong in two ways. If you specify every input validation rule and button location, you end up with a 200-page list that locks the development schedule in place. If you write only “supports asset depreciation” and “supports inventory counts,” developers interpret the requirements themselves and may deliver something far from what the business needs.
The process level (also called the business level) must be clear: trigger conditions, participate in role, input output, abnormal branch, data relations with external systems. This layer is a basis for acceptance and not save. Field level requires "key fields" and "business rules" to be written, without all page elements being counted. For example, the depreciation rule, which reads: "Support average years, sum of years, workload method, depreciation parameters are configured by asset class, residual value is default 5%" is sufficient, and it is not necessary to specify which fields are displayed in the columns in the depreciation schedule.
There's a simple way to determine whether the particle size is right: ask the operator, if he can say "yes, that's what I want" or "that's not right," and it should be "XX", and the particle scale is in place. The business side looks at it all too vaguely, it means it's too abstract; the developer looks at the details, it means it's too rough.
The rhythm of collaboration: business or technology?
The two approaches are the same: first, business-to-business talk, then final work to technical teams to assess feasibility, with the benefit of high demand, not being carried away by technology, risk that there may be "no" for business, and "no" for business. And the other is: business-to-business, business-to-demand, technology-to-syllable, or not, whether or not to do it, change options, fit for a team with strong in-house skills, less bending, risk that the technology will intervene prematurely to suppress real claims from business, leading to "can do" filtering out, rather than "shouldn't".
Shanghai-based Skill Co. Ltd. usually moves at a three-wheel pace in the project: first round, where only business parties talk about the status quo and pain points, where technology does not participate, avoid pre-emptive constraints; second round, where the list of first drafts is reviewed by business parties, one by one, and then go down; third round, where technical aspects are valid, how long will it take to "do" and "do it?" Each round clearly indicates that the business side is free from technology, and the technology side is not troubled by imaginary demand.
The most easy details on the list.
The most frowned-down feature, but the easiest to ignore is offline synchronization. Fixed asset inventory sites often do not have stable networks -- warehouses, workshops, field project offices. If you only mention "support mobile end-points", without detached collection and network recovery conflict resolution, there will be problems at delivery. The eligible requirement should state: the time of local storage of the inventory data, the rules for covering the conflict fields, whether or not it supports multiple people taking the same area at the same time.
- The depreciation rules are underestimated. The financial requirements for depreciation are much more complex than the "selection method": whether to charge on a monthly basis, how to allocate over a month, whether historical depreciation is recalculated after a change in asset class, and whether the last month of the month of end-of-life is counted. These rules are not included in the list, and the cost of later adjustments is two to three times the amount of development work.
- The list of rights is easily cut across. Many lists only say "administer handles all the functions, and ordinary users can only view them", but in the actual scene the head of department looks at his department's assets without doing it, stores are recorded without review, and finances are only recorded without recording, but they guide the accounts. What each character sees, what it changes, what it approves, and it is best to have a power matrix that is ten times clearer than a text description.
- The demand for a report is easily skewed by "wanted." Business says "want to want" and sounds reasonable, but "all" cannot be accepted. Each report must write: "Use, calibration, data source, refresh frequency. Like "Assets of monthly increases or decreases, number 3, count new in the month, decrease, allocate original values and quantities of assets, group up by asset class," develop what can be used to deliver.
Self-check before delivery
Before the list is finalized, do you have a business scene for each functional point? Does every business scenario have a clear responsibility role? Do each character's permission boundaries are clear? Do depreciation, inventory, redeployment, and all four core processes are covered? Do you have statistical calibrations for each report? Where does data come from, where do you write it?
Another easy to ignore is the asset code rule. It is only halfway through many projects that the code code is not set, which makes it impossible to migrate historical data or to match new system coding and old systems. The asset code rules are determined at the demand stage, including the coding structure (a few paragraphs, several paragraphs, meanings), whether an extension is reserved, and whether it is compatible with the old code.
The process of combing the demand is to turn vague management claims into a list of acceptables. The more solid this thing is, the more it is developed later, the more it is tested, the less it is. If you find that a piece is always unwritten in the combo, either the business itself is not right or there are multiple departments involved in an evaluation, and then the hard work will just push the problem backwards.

Ms. Lei
Mr. Pang