Tough times call for more strategic decision-making, which is why the market for business intelligence and data management tools is still growing. Research firm Gartner Inc
. expects BI to be one of the fastest-growing software segments in the coming years, and it’s remained among the top priorities of CIOs, according to its 2009 annual survey of CIO technology priorities.
The BI platform market is expected to grow 6.3 per cent through 2013, while the combined BI, analytics and performance management market is expected to grow 8.1 per cent in the same period.
In Gartner’s BI magic quadrant for 2010, market leaders are Oracle Corp., IBM Corp., Microsoft Corp., SAS Institute Inc., SAP AG, Information Builders Inc. and MicroStrategy Inc. Challengers include QlikTech International AG, Tibco Software Inc. and Tableau Software Inc., while niche players include Targit AS, Actuate Corp. and Panorama Software Ltd., among others.
Open source BI tools are becoming more mainstream, according to Gartner, though functionality is not yet on par with large commercial platforms and is rarely seen as an enterprise-wide BI standard. However, Gartner predicts open source BI production deployments will grow five-fold through 2012. SaaS BI is also becoming a more enticing option, and many pure-play vendors are offering this as an alternative business model, but it’s still a small percentage of overall deployments.
A typical purchase cycle
BI projects vary in size and complexity, and organizations typically approach them with varying levels of competence in planning and execution, said Gareth Doherty, senior research analyst with Info-Tech Research Group. Most BI projects take, on average, between six to 12 months to roll out, most being closer to the one-year timeframe.
“That doesn’t include the inevitable hurdles around encouraging end-users to actually adopt the tool,” he said. “A fully implemented and configured BI toolset does not mean the organization is getting any benefit from it. End-user adoption is critical to success.”
SMBs typically have simple IT environments and modest analytical and reporting needs, which make product selection much easier, and most SMBs use only a fraction of the functionality available in a BI tool. Most, if not all, BI toolsets will adequately address their needs, so they can focus more on selection factors such as cost and ensuring they have the right skills in-house to maintain the solution post-implementation.
For larger rollouts, there are more factors to consider. But many BI projects are primarily IT-focused, with too little involvement from the business, said Doherty. BI is a tool for business users and its main value lies in the insights it provides to end-users, who can make more effective strategic and operational decisions. Business users should be an integral component of a BI project, he said, from initial planning and business case development through to requirements elicitation and product selection.
“Line-of-business managers are typically at the forefront of a BI initiative, since they’re the ones who need to generate reports that enable them to make fact-based decisions,” said Mario Ianniciello, vice-president of sales with SAS Canada. A marketing person, for example, wants to understand the response rate of a given marketing campaign; BI can be used to better design future campaigns to achieve better results.
At its simplest level, BI projects using self-service tools that link to existing data sources such as Microsoft Excel or PowerPivot can take minutes, while more complex projects take months to roll out, said Daniel Shapiro, senior product manager of application platforms with Microsoft Canada. Small businesses can typically take advantage of out-of-the-box solutions right away, such as Microsoft Excel, SharePoint and SQL Server, which can act as standalone BI solutions or work together. With larger or more complex businesses, however, customization and integration are important project requirements.
What goes wrong
One common mistake is the lack of adequate business involvement and corresponding business strategy, such as metrics, to truly leverage the tool, said Doherty. Another issue is the failure to understand the complexities of changing business processes to include the use of BI, or adequately understand change management issues.
And they don’t always budget appropriately, he said, since licensing and support costs are only a fraction of the cost of deploying and maintaining a BI tool. Many IT managers don’t think in terms of total cost of ownership and end up with much more expensive projects than anticipated. Timelines also tend to be underestimated, which results in unanticipated costs, such as labour.
Organizations often struggle to integrate a BI product with existing products, said Ianniciello. Every BI application requires a data integration and data quality solution in order to feed the appropriate information back to the BI application. If those products are not fully integrated with the BI application, it will add cost and time to the deployment, as well as adding to the project’s complexity. Typically, 80 per cent of the effort required for a BI initiative comes down to the data preparation phase.
Some organizations fail to factor in their needs a year or two down the road, said Ianniciello. As their business grows, they may find they require business analytics, but the tool they purchased only provides reporting. Organizations should look for tools that can grow with the company, he said, from simple reporting tools to predictive analysis (the ability to create forward-looking predictions based on analytical and statistical analysis of historical trends).
Be a better buyer
“Organizations should spend more time up front planning the project and ensuring they’ve involved the business in the process,” said Doherty. “All of that will take more time, but ensure better product selection. They should not attempt to optimize the purchase cycle in terms of speeding up the process until they’re sure they understand their needs.”
Talk to other businesses that have experience in selecting a solution and have gone through a complete implementation, said Ianniciello. Look beyond BI and include the evaluation of other areas such as data mining, data integration and data quality in order to have a complete picture of an end-to-end BI offering.
The optimization of the purchase cycle for BI centres on defining deliverables with measurable business outcomes, said Shapiro. “BI means a lot of different things to different people, so it’s important that everyone clearly understands how value is being delivered from the proposed solution,” Shapiro said. Some organizations are looking for efficient delivery of simple reports; some need rich BI capabilities to optimize specific business processes, which require more advanced (and expensive) solutions.
“In my experience, if driven by IT, the decision is more focused on operational and architectural considerations than specific business requirements,” said Shapiro. “If the decision is driven by the business community, the decision is more focused on business requirements at the expense of operational and architectural considerations.” In a perfect world, both parties are at the table during the entire process.
End-user adoption is the most critical success factor, said Doherty. After all, a tool that isn’t used can’t deliver value. “Once end-users start demanding more reports and functionality, you know you have a successful BI project.”