How to implement EDS and not screw it up
In large and medium-sized companies, sooner or later there is a need for an electronic document management system, since keep track of the infinitely growing number of documents becomes completely impossible. And in 90% of cases, the first attempt at implantation becomes a big pain in the heart and somewhere lower, although it could have been avoided. And you can avoid both the pain itself and the introduction in general, if you realize that you are not ready for it yet and can argue it.
As an employee of a company implementing EDS, I would like to share a check list of what should be done before implementation so that you will not be fired. So:
How many projects ended in mutual dissatisfaction between the customer and the implementer, because the first wanted a bright future, and the second tried to make a brilliant present (well, or vice versa). Before you even start discussing the implementation project with anyone, understand what your goal is. What do you want to achieve with this project? In principle, for the first time any painful problem that you want to solve is suitable. Do your employees periodically lose documents? Make a goal control of the paper version. Does everyone forget to fulfill the instructions related to the document? Let's tighten control over execution. Etc.
What will happen if you do not:
First, you will get a starship that does not fly: a lot of buttons, a lot of windows, dark reports, but what for it all is needed, it’s not clear to anyone: it’s not better!
Secondly, a smart contractor will replace your real goals with easily achievable for yourself and most effective looking. If he is able to make beautiful reports, then the system will have 100 reports showing how many times a day the employees thought about a document that continues to prop up the cabinet leg instead of lying on the table at the CEO.
And more: reduce the number of goals to a minimum. Otherwise, you will get an SED for the price of a death star.
')
This, in principle, concerns any goals in any areas of life. You can not set a goal: "lose weight", because such a goal does not have "brakes" and points for assessing success. And you either stay fat, or turn into anorexic. So with the workflow, do not set the goal “to dress better, control the execution” or “coordinate the documents faster”. Formulate more precisely, for example: “Now we are responding to clients with their letters with a delay of a month, we must respond within two weeks. Or take the cookies from the artist, if not managed.
What will happen if you do not:
You will come to a future that seems like a bright implementer. And, as you know, someone and the grave in joy.
This is my favorite! If any departments and employees can make demands on any process, then you will not go further than the technical assignment task. Each attempt to create a single process will be rested against the genitals of the heads of departments to determine which problems are more critical. A clear impression will be created that all these people work in different competing companies and try to bury the rival deeper.
Ideally, each process should have one owner, who can end any disputes with the phrase: “Because I said so,” and no one can argue with him. In companies operating under ISO 9000, it is more logical to designate the owner of the QMS. In others, for each process, the service is most penalized in the case of fakaps: for contracts, lawyers, for the primary, accounting, etc.
What will happen if you do not:
You look at the fable “Swan, cancer and pike”, delivered at your company's expense, in which the main roles will be performed by the department directors. And, if the contractor eventually writes a technical assignment, then there will be 500 pages in it, and the cost of implementation will again approach the death star.
The point intersects with the previous one, but is perhaps the most important. One way or another, the SED must be implemented in a company, like Orthodoxy in Russia - with fire and sword! No matter how beautiful the system is, no matter how it makes life better, there will always be employees sabotaging its use.
Be sure to find a manager not lower than the second level of the hierarchy in the company (gen. Dir or his deputies) who:
The ideal option is when at the end of the project an order is issued that from that day no document in the paper has been accepted. And also this manager himself works in the system: in this case, subordinates have less chances to say something about the non-viability of the implemented solution: after all, Ivan Ivanovitch himself works with him!
What will happen if you do not:
You will receive an SED in which no one works, and which lies in "dead weight". And if you do not complete the paragraph about the "strength of the seat" of the head, then there will be bingo! The system is not used, the project sponsor was fired, you are amazing!
And now it's time to dive into the details. Which way will you go: quick and cheap implementation of a “box solution” or a long and expensive development for you. What is the difference? If you ask the forester how to get to the other side of the forest faster, he will tell you exactly the short path. But if at this moment you are moving to Ferrari, then it is more logical to choose a highway leading around, albeit a few hundred kilometers longer. I explain: the boxed solutions of any implementer are average statistical functionality that suits any company in theory, and allows you to quickly and cheaply deploy it in your territory. If there is not enough money, then let the contractor “comb” your processes under the box. But if your processes have historically developed in a certain way and help the company to exist as efficiently as possible, and there are no problems with the budget, then do not even try to pull the “box” on you.
What will happen if you do not:
If you have little money, do not rush into personalized solutions: the stress will be very heavy, and the result will disappoint you, even if you fork out for an expensive project. You will never understand why IT costs so much?
If you are the owner of unique processes, then the introduction of a boxed solution will expose the implementer as a fool and you, if you try to take his position: after all, what is this “castrated” functionality? How can there be a system without color differentiation of pants?
How many times I met the requirements in the TOR to the contract: 4000 simultaneous users, 12 servers in a cluster, the planned load of 1 billion documents per year, and in fact at the end of implementation 100 people work in the system, who receive 3 documents per day. If you plan to achieve a similar load in only three years, then it is better to do the implementation now and lay down the scaling in a couple of years.
What will happen if you do not:
First, the price tag will again approach galactic scales. Secondly, you and the contractor will spend a lot of time and effort on setting up something that you cannot even check. No synthetic tests and load tests will never show the behavior of the system under real load. Law Murphy has not been canceled. And where in three years there will be a contractor who introduced the system to you is unknown. And according to the laws of the Russian Federation, the warranty period on systems is 2 years.
The only point that directly affects me as a contractor. Listen to the advice on the timing of implementation. You do not trust me, ask a couple of other specialists (also, by the way, universal advice). If everyone tells you that it is impossible to do this amount of work in less than six months, then that means that it is. And you will not decide this question with any infusion of money.
What will happen if you do not:
By the end of the project, you will get something from crutches and bicycles, holding onto snot and allowing you to go through acceptance tests. And then during the remaining term, the contractor will smoothly finish the non-finished, finish the not completed and accelerate not accelerated.
If at least one of the seven points seems not feasible to you, think about whether you want to stand at the origins of a problematic, if not completely, failed project?
As an employee of a company implementing EDS, I would like to share a check list of what should be done before implementation so that you will not be fired. So:
1. Determine the goal already!
How many projects ended in mutual dissatisfaction between the customer and the implementer, because the first wanted a bright future, and the second tried to make a brilliant present (well, or vice versa). Before you even start discussing the implementation project with anyone, understand what your goal is. What do you want to achieve with this project? In principle, for the first time any painful problem that you want to solve is suitable. Do your employees periodically lose documents? Make a goal control of the paper version. Does everyone forget to fulfill the instructions related to the document? Let's tighten control over execution. Etc.
What will happen if you do not:
First, you will get a starship that does not fly: a lot of buttons, a lot of windows, dark reports, but what for it all is needed, it’s not clear to anyone: it’s not better!
Secondly, a smart contractor will replace your real goals with easily achievable for yourself and most effective looking. If he is able to make beautiful reports, then the system will have 100 reports showing how many times a day the employees thought about a document that continues to prop up the cabinet leg instead of lying on the table at the CEO.
And more: reduce the number of goals to a minimum. Otherwise, you will get an SED for the price of a death star.
')
2. Express the goal in numbers
This, in principle, concerns any goals in any areas of life. You can not set a goal: "lose weight", because such a goal does not have "brakes" and points for assessing success. And you either stay fat, or turn into anorexic. So with the workflow, do not set the goal “to dress better, control the execution” or “coordinate the documents faster”. Formulate more precisely, for example: “Now we are responding to clients with their letters with a delay of a month, we must respond within two weeks. Or take the cookies from the artist, if not managed.
What will happen if you do not:
You will come to a future that seems like a bright implementer. And, as you know, someone and the grave in joy.
3. Find a process owner.
This is my favorite! If any departments and employees can make demands on any process, then you will not go further than the technical assignment task. Each attempt to create a single process will be rested against the genitals of the heads of departments to determine which problems are more critical. A clear impression will be created that all these people work in different competing companies and try to bury the rival deeper.
Ideally, each process should have one owner, who can end any disputes with the phrase: “Because I said so,” and no one can argue with him. In companies operating under ISO 9000, it is more logical to designate the owner of the QMS. In others, for each process, the service is most penalized in the case of fakaps: for contracts, lawyers, for the primary, accounting, etc.
What will happen if you do not:
You look at the fable “Swan, cancer and pike”, delivered at your company's expense, in which the main roles will be performed by the department directors. And, if the contractor eventually writes a technical assignment, then there will be 500 pages in it, and the cost of implementation will again approach the death star.
4. Determine the “sponsor” of the project
The point intersects with the previous one, but is perhaps the most important. One way or another, the SED must be implemented in a company, like Orthodoxy in Russia - with fire and sword! No matter how beautiful the system is, no matter how it makes life better, there will always be employees sabotaging its use.
Be sure to find a manager not lower than the second level of the hierarchy in the company (gen. Dir or his deputies) who:
- Will be ready to protect the project with his name;
- It can make everyone work in the system;
- It sits firmly enough in its place.
The ideal option is when at the end of the project an order is issued that from that day no document in the paper has been accepted. And also this manager himself works in the system: in this case, subordinates have less chances to say something about the non-viability of the implemented solution: after all, Ivan Ivanovitch himself works with him!
What will happen if you do not:
You will receive an SED in which no one works, and which lies in "dead weight". And if you do not complete the paragraph about the "strength of the seat" of the head, then there will be bingo! The system is not used, the project sponsor was fired, you are amazing!
5. Decide: you or you?
And now it's time to dive into the details. Which way will you go: quick and cheap implementation of a “box solution” or a long and expensive development for you. What is the difference? If you ask the forester how to get to the other side of the forest faster, he will tell you exactly the short path. But if at this moment you are moving to Ferrari, then it is more logical to choose a highway leading around, albeit a few hundred kilometers longer. I explain: the boxed solutions of any implementer are average statistical functionality that suits any company in theory, and allows you to quickly and cheaply deploy it in your territory. If there is not enough money, then let the contractor “comb” your processes under the box. But if your processes have historically developed in a certain way and help the company to exist as efficiently as possible, and there are no problems with the budget, then do not even try to pull the “box” on you.
What will happen if you do not:
If you have little money, do not rush into personalized solutions: the stress will be very heavy, and the result will disappoint you, even if you fork out for an expensive project. You will never understand why IT costs so much?
If you are the owner of unique processes, then the introduction of a boxed solution will expose the implementer as a fool and you, if you try to take his position: after all, what is this “castrated” functionality? How can there be a system without color differentiation of pants?
6. Live the present
How many times I met the requirements in the TOR to the contract: 4000 simultaneous users, 12 servers in a cluster, the planned load of 1 billion documents per year, and in fact at the end of implementation 100 people work in the system, who receive 3 documents per day. If you plan to achieve a similar load in only three years, then it is better to do the implementation now and lay down the scaling in a couple of years.
What will happen if you do not:
First, the price tag will again approach galactic scales. Secondly, you and the contractor will spend a lot of time and effort on setting up something that you cannot even check. No synthetic tests and load tests will never show the behavior of the system under real load. Law Murphy has not been canceled. And where in three years there will be a contractor who introduced the system to you is unknown. And according to the laws of the Russian Federation, the warranty period on systems is 2 years.
7. Consider the timing
The only point that directly affects me as a contractor. Listen to the advice on the timing of implementation. You do not trust me, ask a couple of other specialists (also, by the way, universal advice). If everyone tells you that it is impossible to do this amount of work in less than six months, then that means that it is. And you will not decide this question with any infusion of money.
What will happen if you do not:
By the end of the project, you will get something from crutches and bicycles, holding onto snot and allowing you to go through acceptance tests. And then during the remaining term, the contractor will smoothly finish the non-finished, finish the not completed and accelerate not accelerated.
Afterword:
If at least one of the seven points seems not feasible to you, think about whether you want to stand at the origins of a problematic, if not completely, failed project?
Source: https://habr.com/ru/post/317198/
All Articles