Forget lengthy analyses, forget lengthy planning, and forget going too big all at once. Rapid prototyping in small chunks is the key to succeeding with business intelligence (BI).
This is the opinion of Paul Morgan, MD of Asyst Intelligence, who says: “The ideal strategy is to tackle BI in bite-sized chunks. Rather than spend six months in lengthy, complicated analysis and planning, companies should build rapid prototypes to provide fast results and answers."
He explains: “User acceptance is pivotal to the sustainability of BI in any organisation. However, people are driven by results. The longer you spend in the analysis and planning phases, the longer it’s going to take until you can give them tangible, real-life examples of what BI can do for them. Up until then, it’s all pie in the sky stuff. You want to be able to deliver dashboards, reports and queries as soon as possible to make it real and keep the momentum going. If not, interest will wane and acceptance of BI is at risk.
“The trick is to garner user acceptance in the early stages by involving them in testing and getting their feedback early on. So, instead of drowning in complicated analyses of the last three months sales figures, create and deliver a simple dashboard of the sales figures of the last week and show it to them. That way, users have a tangible example to refer to, and they can decide whether it meets their expectations before you go any further. Ideally, start with those areas where the most value and quick wins can be derived, whether it be sales figures or financials.
“Aside from winning their interest and keeping the momentum going, this strategy assists in defining the data architecture by offering a meaningful point of departure for identifying the main priorities and working out the drivers that will help users to live up to the organisation’s grand strategy. BI is an important and useful tool for driving performance management because users can monitor and measure performance against company objectives.
“For instance if the goal of a hotel group is 100% room occupancy across all it hotels for the 2010 Soccer World Cup, the dashboard should give a snapshot of where the group stands in respect of that goal and how many room nights still need to be sold for that to be achieved.”
There are other advantages of the rapid prototyping approach. For instance, instead of creating a lofty, long-term strategy and then only discovering after months of planning that the organisation does not have the systems in place to support the strategy, companies can identify early on what type of data they need and what systems they must have in place to support that data. For example, if the goal of the business is to achieve greater market-share, it will need systems in place to deliver the relevant market information.
“Simply put, this approach helps to inform what sort of information and systems are needed to deliver the output that is required. It also allows companies to adapt their strategy to meet their changing requirements as they go along,” comments Morgan.
He stresses however, that this approach does not advocate that no planning is required.
“There are a number of elements that go into formulating a BI strategy, even if the strategy is to tackle BI piece meal. As a starting point the strategy must deal with where the data is going to come from. It must also deal with how that data is going to be integrated and how data quality is to be ensured. So, there are, of course, the systems and infrastructure requirements to consider.
“Education and awareness must also form part of the strategy. People at all levels of the business must be onboard and support the effort to create an intelligent enterprise. User focus groups and workshops must be planned for to get user input, test outputs and gage feedback to ensure that the tool is relevant to their needs.
“In addition, the implementation of BI often means prescribing a new structure of doing things. Change management must therefore be built into the BI strategy.
“Of course, these elements must be planned for,” he says.
Morgan concludes: “Companies often focus intensely on the technical side of an implementation, when they should be focusing on the business needs first. Once the business needs are established, the necessary structures can be put in place to support those needs in line with organisational strategy.
“Most companies are not experts in BI so it is advisable for any organisation looking to implement BI to consult with BI professionals, even if it is only at an advisory level. It is much easier to look at a business from the outside than it is to look at in from the inside. Consultants will also be able to guide the process from tool selection to staff training on the new system. It increases the chances of long term BI success.”