As a Business Analyst/Testing Consultant who has worked for several Implementations, I often receive queries from my clients asking which is the best Software Implementation Technique available. Is it Big Bang or Phased Implementation. As far as Software Implementation and testing are concerned it is a very risky area of Software Development. Because there are so many internal and external risk factors involved in it.
Software Implementation or Software Adoption deals with the transfer (conversion) between an old system to a target system in an organization. So if a Bank/company works with an old software system, it may want to use a new system which is more efficient, has more work capacity etc. So then a new system needs to be adopted, where after it can be used.
There are several types of adoption that can be used to implement a system. The types ‘big bang’, ‘parallel adoption’ and ‘phased adoption’ form the main types that are used to adopt a system. The big bang relates to the cosmological theory (Big bang) where the start of the cosmos happened at one moment in time. This is also the case with the big bang adoption type where the new system is adopted on one date. In case of parallel adoption the old and the new system are running parallel so all the users can get used to the new system, but still can do their work using the old system. Phased adoption means that the adoption will happen in several phases, so after each phase the system is a little closer to be fully adopted by the organization.
When it comes to the world's most complex Software Implementations like Core Banking Software Implementation which has been linked to several upstream and downstream interfaces, it is really critical to do a meticulous planning from all the sides like stakeholders, developers, testers and other consultants. The other aspect to be considered here is, whether it is a New Implementation or it is a Replacement of a legacy system. If it is going to be new implementation everything will be from the scratch having a base model of the application. But if it is going to be the replacement there are several important aspects to be considered like Gap Analysis for the legacy and replacement system, right from mapping of appropriate fields from the legacy to replaced system, and Data Migration from legacy to replacement, accounting, Local and Global Compliance Reporting and the Replacement System's Integrity with the various existing interfaces has to be analysed in breadth and depth.
The other critical area to be concentrated is whether the implementation has to be done in a single country along with all the branches or it is for multiple countries. I prefer go in for big bang implementation, if it is done in a single country only. If the Implementation has to be done in several countries it has to be done phase-wise as the base model of a country can be re-modified to an other country's country specific requirement and the best practiced approach learning in one country can also be used for another country.
The adoption type or software implementation type has to be defined in front of the adoption phase and will be chosen based on the goals to be achieved and on the type of system to be implemented. The three types of adoption, Big Bang, parallel adoption and phased adoption, are ranging from an instant switch to a strategy where users progressively start using the new system over a certain period of time (can be weeks, months or even years).
The actual selection is done by prioritizing the goals to be achieved and then matching a strategy against it (Eason, 1988). Eason defines the following goals:
Possible existence of needed “critical mass” to make the system work.
If a large critical mass is needed for the system to work a big bang strategy might seem the answer. (Rogers, 1995)
Need for risk control, if risk is involved.
Minimizing risk to operation of the organization is key. Parallel and phased introduction might control this risk.
Need for facilitation of the change.
The organization has to be ready for the changeover. Socio-technical preparations such as training sessions and ready-made scenarios must be clear.
Pace of change over
The speed of in which the organization is changing over to the new system.
Local design needs
The system might be adjusted to the users needs. In this the chosen strategy must provide the opportunity to do so.
The actual selection of adoption type depends on far more factors then these goals, but they create a window to choose one of the types. Other criteria are called variables (Gallivan, 1996). Gallivan suggests that the appropriate adoption types depends on:
Innovativeness of the individuals
Attributes of the ones that are to adopt the innovation/system
The type of innovation
Is it a process or product innovation?
Attributes of the innovation itself
Preparedness, communicability and divisibility
The implementation complexity.
How complex is the implementation or what is its extent?
These variables are of a higher level then the criteria of Eason and should be handled as such. Based on table 1 and on the mentioned higher level variables by Gallivan, one can make a selection of an appropriate strategy to choose.
Prepare an organization for adoption or change management:
In order to prepare the organization for the adoption of the new system, the changes that will take place need to be determined. This is necessary to be able to have a plan or an overview of the changeover, and can be done by creating requirements for the system. Once the management has determined the requirements in a report of determined changes, they need to agree upon them to be able to move on with the change-process. If there is no agreement, the management needs to discuss the requirements again and again until they do agree. If agreement is achieved and the agreement contract is signed, the organization can take further steps. So now the test-phase can be prepared, in which the validity of the data that will be used will be checked and in which trials will be held (Eason, 1988).
Even though both the implementation techniques are unique and best in a way it depends on the goals and structure of the Bank/Company which is involved in the Implementation. In a nutshell if the Bank/Company is large with huge number of branches, wants to minimise the risk and has to be implemented in several countries Phased Implementation is the best suited one and if the Bank/Company is small with less branches and it is working only in a single country big bang or parallel execution is the best suited method.
Note: The views expressed in this blog are my personal views and a collection of references from the books listed below. These views does not reflects any other aspect of Software Implementation.
References:
Eason, K. (1988) Information technology and organizational change, New York: Taylor and Francis
Gallivan, M.J., (1996) Strategies for implementing new software processes: An evaluation of a contingency framework, SIGCPR/SIGMIS ’96, Denver Colorado
Rogers, E.M. (1995), Diffusion of innovations, New York: Free Press
Software Implementation or Software Adoption deals with the transfer (conversion) between an old system to a target system in an organization. So if a Bank/company works with an old software system, it may want to use a new system which is more efficient, has more work capacity etc. So then a new system needs to be adopted, where after it can be used.
There are several types of adoption that can be used to implement a system. The types ‘big bang’, ‘parallel adoption’ and ‘phased adoption’ form the main types that are used to adopt a system. The big bang relates to the cosmological theory (Big bang) where the start of the cosmos happened at one moment in time. This is also the case with the big bang adoption type where the new system is adopted on one date. In case of parallel adoption the old and the new system are running parallel so all the users can get used to the new system, but still can do their work using the old system. Phased adoption means that the adoption will happen in several phases, so after each phase the system is a little closer to be fully adopted by the organization.
When it comes to the world's most complex Software Implementations like Core Banking Software Implementation which has been linked to several upstream and downstream interfaces, it is really critical to do a meticulous planning from all the sides like stakeholders, developers, testers and other consultants. The other aspect to be considered here is, whether it is a New Implementation or it is a Replacement of a legacy system. If it is going to be new implementation everything will be from the scratch having a base model of the application. But if it is going to be the replacement there are several important aspects to be considered like Gap Analysis for the legacy and replacement system, right from mapping of appropriate fields from the legacy to replaced system, and Data Migration from legacy to replacement, accounting, Local and Global Compliance Reporting and the Replacement System's Integrity with the various existing interfaces has to be analysed in breadth and depth.
The other critical area to be concentrated is whether the implementation has to be done in a single country along with all the branches or it is for multiple countries. I prefer go in for big bang implementation, if it is done in a single country only. If the Implementation has to be done in several countries it has to be done phase-wise as the base model of a country can be re-modified to an other country's country specific requirement and the best practiced approach learning in one country can also be used for another country.
The adoption type or software implementation type has to be defined in front of the adoption phase and will be chosen based on the goals to be achieved and on the type of system to be implemented. The three types of adoption, Big Bang, parallel adoption and phased adoption, are ranging from an instant switch to a strategy where users progressively start using the new system over a certain period of time (can be weeks, months or even years).
The actual selection is done by prioritizing the goals to be achieved and then matching a strategy against it (Eason, 1988). Eason defines the following goals:
Possible existence of needed “critical mass” to make the system work.
If a large critical mass is needed for the system to work a big bang strategy might seem the answer. (Rogers, 1995)
Need for risk control, if risk is involved.
Minimizing risk to operation of the organization is key. Parallel and phased introduction might control this risk.
Need for facilitation of the change.
The organization has to be ready for the changeover. Socio-technical preparations such as training sessions and ready-made scenarios must be clear.
Pace of change over
The speed of in which the organization is changing over to the new system.
Local design needs
The system might be adjusted to the users needs. In this the chosen strategy must provide the opportunity to do so.
The actual selection of adoption type depends on far more factors then these goals, but they create a window to choose one of the types. Other criteria are called variables (Gallivan, 1996). Gallivan suggests that the appropriate adoption types depends on:
Innovativeness of the individuals
Attributes of the ones that are to adopt the innovation/system
The type of innovation
Is it a process or product innovation?
Attributes of the innovation itself
Preparedness, communicability and divisibility
The implementation complexity.
How complex is the implementation or what is its extent?
These variables are of a higher level then the criteria of Eason and should be handled as such. Based on table 1 and on the mentioned higher level variables by Gallivan, one can make a selection of an appropriate strategy to choose.
Prepare an organization for adoption or change management:
In order to prepare the organization for the adoption of the new system, the changes that will take place need to be determined. This is necessary to be able to have a plan or an overview of the changeover, and can be done by creating requirements for the system. Once the management has determined the requirements in a report of determined changes, they need to agree upon them to be able to move on with the change-process. If there is no agreement, the management needs to discuss the requirements again and again until they do agree. If agreement is achieved and the agreement contract is signed, the organization can take further steps. So now the test-phase can be prepared, in which the validity of the data that will be used will be checked and in which trials will be held (Eason, 1988).
Even though both the implementation techniques are unique and best in a way it depends on the goals and structure of the Bank/Company which is involved in the Implementation. In a nutshell if the Bank/Company is large with huge number of branches, wants to minimise the risk and has to be implemented in several countries Phased Implementation is the best suited one and if the Bank/Company is small with less branches and it is working only in a single country big bang or parallel execution is the best suited method.
Note: The views expressed in this blog are my personal views and a collection of references from the books listed below. These views does not reflects any other aspect of Software Implementation.
References:
Eason, K. (1988) Information technology and organizational change, New York: Taylor and Francis
Gallivan, M.J., (1996) Strategies for implementing new software processes: An evaluation of a contingency framework, SIGCPR/SIGMIS ’96, Denver Colorado
Rogers, E.M. (1995), Diffusion of innovations, New York: Free Press
No comments:
Post a Comment