X

Why your business intelligence implementation will fail

Star IT projects fail; it’s sad, but they do. For a project lead tasked with a business intelligence implementation project, failure may not be an option going in, but it may occur anyway. To make sure your business intelligence implementation actually delivers its tremendous potential benefits to your enterprise, watch out for the risks that can cause it to fail. Here’ a guide.

Letting end users drive your business intelligence implementation

Of course, eventually, the BI system has to be used by your company’s employees. It makes sense for them to be involved in helping the enterprise decide on what sort of functionalities its BI systems should have. However, too many times, excessive focus on letting end-user stakeholders drive the business intelligence implementation causes companies to overshoot budget and timelines, eventually ending with a very complex system that’s just too difficult for most people to use. Understand that end users, particularly those with primarily customer facing roles, might not appreciate or understand technologies. If they try to dictate the technical specifics of the business intelligence implementation, you’re headed for trouble. To do better, make sure you thoughtfully select end-user representatives and invest a lot of time in understanding their expectations and concerns, translating them into technical descriptions yourself. Many enterprises use the 5 whys technique that entails asking the "why" question five times to understand the deep-rooted motivations, fears, and expectations of people.

Taking it too easy with manual testing

Because a business intelligence implementation means new processes, applications, and workflows for enterprises, there obviously needs to be a lot of testing done before such an implementation can go live. However, it’s pretty natural for project teams to not have specific testing capabilities, which causes them to slack off on manual testing. Restriction of testing to mere preliminary manual testing invariably leaves a lot of defects in the BI system, which gets highlighted during user acceptance testing, causing a lot of rework and delays. Because these defects can also entail logic and process errors, the time and dollar cost of corrections can cause the project budget to surge. Instead, business intelligence implementation teams would do well to plan on testing right from the time of blueprint completion.

Dated technology

Choosing the BI technology for your enterprise is, clearly, a crucial decision. Yes, CRM and enterprise resource planning (ERP) software were the first aggressive marketers of BI solutions. However, for obvious reasons, their BI systems are married to their core solutions such as ERPs, and hence, need careful analysis before being shortlisted as a BI tool for your company. There are vendors in the market whose BI systems are extensions to their ERP systems, and hence rooted in obsolete business models and tech standards. These solutions, apart from their core flaws, are also not developed with the idea of cloud-compatibility in mind. Business practices are evolving rapidly, and BI tools need to be updated and flexible enough to make enterprises comfortable in replicating their changing processes in their BI systems. Choose a BI technology that brings data visualizations, drag-and-drop self-service dashboards, and mobile accessibility to the table.

Ignoring broad data integrity issues

BI tools are meant to deliver tremendous data processing, presentation, and analytics capabilities to enterprises. However, if you feed garbage data to the system, all you will get out is neat-sounding garbage. This is where enterprises need to deliberately divide their focus, paying as much attention to data practices as to the business intelligence implementation. Remember, BI will drive enterprise-wide decisions, so you simply can’t afford to be myopic about the BI implementation and lose sight of the broader data integrity requirements playing out at the horizon. Any errors in financial reporting and regulatory compliance can be catastrophic for the reputation of a business. Done rightly, BI can bring such anomalies to light, helping enterprises make course corrections. So, businesses need to have a cohesive, deep-rooted, deliberate, and proactive approach to bolster BI data processes to identify potential data integrity issues at the earliest.

Absence of business support

Too many IT projects falter because business buy-in takes too long, costs too much, or just never happens at all! Lack of business support can quickly take the fizz out of your BI systems. This, in turn, could result from a lack of proper communication, and lack of training for users. Unless your BI project team creates clear actionable response mechanisms, it wouldn’t matter how much or what kind of data analyses your BI system does. Under no circumstances should you proceed with a business intelligence implementation if you sense that it’s on track to becoming a typical pet-project for a core group of managers — without the target end users truly coming on board. Selection of core user groups, planning on detailed training plans, and delivery of the training are key to getting business support after BI implementation.

Too much measurement

Vimeo

For a project that has measurements at its very core, can there be a thing such as too much measurement? Seems the answer is yes. Key performance indicators (KPIs), as much as they’re crucial for progress and success tracking, can open the floodgates for the waters of “complexity” to suffocate your project. Overuse of performance indicators, according to experts, is one of the most commonly committed mistakes of scorecard and dashboard implementations. Not only do complex KPIs need more maintenance and analytics resources, but they can also make routine BI tasks turn into veritable headaches both for IT as well as for the business. If you must begin with your large list of KPIs and performance scores, at least have a clear plan in place to review them quarterly and consider elimination and consolidation to reduce the count with time.

Managers believe that ROI measurements of BI projects are difficult to calculate. They’d also do well to understand how laxity on certain aspects can cause project costs and timelines to multiply manifolds, killing your notional ROIs. Avoid these risks, and avoid the failure.

Photo credit: Shutterstock