Over the years we have observed and been involved in banking business intelligence (BI) projects across the world. When it comes to planning for successful projects, five key considerations will help ensure the project will go as smoothly as possible.
To give some context to why you should consider business intelligence and analytics in banking here’s a reminder of what they enable you to do when implemented successfully:
- Business Intelligence enables you to discover WHAT happened.
- Analytics enables you to discover WHY it happened.
- Combining the two provides insight into how to ACT on any arising problems or opportunities.
With that in mind, here are five key steps to follow in order to deliver successful business intelligence in banking.
Start with strategy
What are the goals for the BI project? Business intelligence goes hand in hand with the strategic objectives of the financial institution. When implemented effectively, it enables the financial institution to monitor the dozen or so key performance indicators (KPIs) it needs in order to track its performance in line with strategy.
BI can help with strategic initiatives such as:
- Expanding the branch network and analysing efficiency and profitability of new branches compared with existing branches.
- Introducing a new line of business or new product lines and analysing their fees and yields in comparison to existing products.
- Assessing the viability of a merger or acquisition by consolidating the balance sheets of the entities before and after the event.
- Consult business users and identify key deliverables
This step should include identifying your users and what they need in order to perform their roles effectively. What information do they need that they don’t get today? Can the frequency of information delivery be increased – for example, moving from monthly to daily delivery of performance information.
When all stakeholders or representatives of stakeholders are consulted pre-implementation, it helps gain valuable buy-in early in the process, which in turn helps with user adoption, user empowerment and ultimately motivation and accountability in achieving strategic objectives.
- Plan phases using a simple timeline
Successful business intelligence projects can be broken down into two or three phases. The phases won’t necessarily be highly complex; they can be incredibly simple. Doing things in phases allows the core users to adopt and embed the BI applications into processes with as little disruption as possible before embarking on a wider roll out to the rest of the company.
When looking at the overall rollout, it is a good idea to simply break sets of tasks or deliverables into 3 or 6 month chunks. Identify what needs to be delivered in the first phase, then execute and allow to bed-in, ensuring user acceptance and expected value is being derived. The second phase could then start in, say, six months from start date, the third phase in 12 months etc.
The key here is ensuring that realistic deliverables are established, prioritised and attached to a timeline. When these key deliverables have been identified, it becomes much easier to plan how you will meet them and to communicate that plan with interested stakeholders.
Evaluate potential solutions
It’s important to evaluate all of your options to establish which will give you the greatest chance of a successful BI project. Your options could include:
- Building in house by acquiring hardware with licenses for databases, software and visualisation
- Hiring consultants to do some of the work for you using a combination of the above.
- Using a pre-built application to plug your data into.
Each of the above has its own pros and cons. The first two points above could mean that you get BI built to your requirements, so long as everyone can understand those requirements and implement them as specified. In order to be viable, forward planning and clear requirement definitions are crucial, and time taken to deliver value is likely to be much longer than the third option.
The third option is likely to be the most time and cost effective, though it might mean compromising in phase one to get 80-90% of what you need quickly, with the benefit being that it has already been built. The remaining 10-20% could be implemented in later phases. The third option is likely to be the most beneficial way to implement business intelligence if you can live with potential short term compromises.
Top tip: When evaluating your options, you should look for case study examples of previous success and speak to previous customers to learn about their experiences and relate your findings to your own situation.
- Establish ownership
It is essential to establish ownership for steps 1-4 so it is understood who is responsible for each step of the pre-implementation planning and execution processes. This will ensure that communication channels are open, tasks and phases are progressing as expected and that alerts are raised in the event of unforeseen challenges.
There are a few important processes where ownership should be considered as part of the pre-implementation planning:
- Definition of requirements and who is responsible for user acceptance testing (UAT) these, data reconciliation and ad-hoc requests once implemented.
- Definition of the extract, transformation and load (ETL) process to enable the flow of data from core source systems to business intelligence applications.
- Administration of users and security standards for the new BI applications.
Hopefully these steps do not seem daunting. Banking business intelligence projects don't need to be complicated or confusing, especially if you spend some time planning upfront by consulting users, putting timelines in place and finding the right solution partner as part of your evaluation process.