Addressing the challenges faced in telecom billing migration projects
In my previous blog post, I listed some of the key challenges faced by Communication service providers while migrating their billing systems. Now let us see how these challenges can be addressed:
1. Development methodology: The Methodology adopted for migration plays a critical role in the success of billing migration. Traditionally a high risk approach - big bang approach (waterfall model) is followed for a migration program. Instead of going for a big-bang approach, a progressive/incremental approach (agile methodology) would be a better approach. The later approach will help track the migration solution closely, provide continuous feedback for the subsequent phases of migration, keep the minimum down time for the billing cycle and will help the development team to cope up with the changes due to the new requirements raised by business in parallel to the migration.
2. Usage of real time data for design and testing:
a. Design and analysis teams must look at the real time data to get good understanding of the data for better mapping/design. Mapping not only needs to address the best case scenario but also needs to address the deviated customer data.
b. Testing teams should use real-time data (old copy of production data) which will help discover the defects at an early stage. This minimizes the number of issues in production as handcrafted data may not work in production.
c. In case of a CSP having large customer base on legacy stack, to ensure that new functionality works for the migrated customer – use migrated data for performing functional testing of new stack.
3. Streamlining the business processes: In addition to the data cleansing activity, CSP can also use the migration program as an opportunity to stream line their business processes, keeping package customizations to a minimum level and gaining the ability to provide out of box features – reducing the time to market.
4. Parallel Runs during UAT :
a. To minimize issues post migration, it is beneficial to have a parallel run approach during testing of the migration solution. Hence bills are produced from legacy and new stacks before and after migration. This can be limited to acceptance testing phase.
b. In case of highly complex environments and huge data on the legacy stack, this approach can be extended to production for initial two bills.
5. Educating the customer: We can expect some bill format differences post the migration due to many reasons, eg. Proration of charges or presentation of usage which spans across off-peak and peak time period. It is essential to identify the known differences upfront and educate the customer indicating the expected changes in the presentation of the bill.
2. Usage of real time data for design and testing:
a. Design and analysis teams must look at the real time data to get good understanding of the data for better mapping/design. Mapping not only needs to address the best case scenario but also needs to address the deviated customer data.
b. Testing teams should use real-time data (old copy of production data) which will help discover the defects at an early stage. This minimizes the number of issues in production as handcrafted data may not work in production.
c. In case of a CSP having large customer base on legacy stack, to ensure that new functionality works for the migrated customer – use migrated data for performing functional testing of new stack.
3. Streamlining the business processes: In addition to the data cleansing activity, CSP can also use the migration program as an opportunity to stream line their business processes, keeping package customizations to a minimum level and gaining the ability to provide out of box features – reducing the time to market.
4. Parallel Runs during UAT :
a. To minimize issues post migration, it is beneficial to have a parallel run approach during testing of the migration solution. Hence bills are produced from legacy and new stacks before and after migration. This can be limited to acceptance testing phase.
b. In case of highly complex environments and huge data on the legacy stack, this approach can be extended to production for initial two bills.
5. Educating the customer: We can expect some bill format differences post the migration due to many reasons, eg. Proration of charges or presentation of usage which spans across off-peak and peak time period. It is essential to identify the known differences upfront and educate the customer indicating the expected changes in the presentation of the bill.
Comments
Ashok,
In addition to the above, one of the most aspects for any successful Billing migration is the 'People' aspect. By 'People' I mean the internal customes, the Business users, the Billing advisors etc.Do the Business users/billing advisors understand how the new system is supposed to function. The first tendency is to expect the new system to have the same functionalities/ processes as the old system which is never really the case. Different systems will implement the same process in a different way and it is important that the business users understand this upfront. There is an inherent tendency to compare the old with the new as the users are accustomed to using the old system.It is important that the business users get trained on the use of the new system to enable fast adoption in a positive way.
Commented by: Kanwar Virender | February 12, 2010 12:44 PM
Thanks Virender for the comment. I agree to your point. Educating the business users along with the customers is very much important.
Commented by: Ashok Reddy | February 16, 2010 8:48 AM
A big bang (Waterfall) Vs an Incremental Approach (Agile) always remains a point to debate with the Incremental approach trying to push more by taking risks, which are under many circumstances, un necessary as the fear of failure and the impact of failure remain less. Also the business involvement is less due to the above reasons. A Big Bang / waterfall type of approach always delves on quality and accuracy of the As Is and To Be analysis and since the stakes and risk of failure is high, it also goes through with more due diligence and business involvement, hence causing fewer issues long term.
Commented by: Vignesh | March 15, 2010 2:58 PM