Start small. That’s one piece of advice Lyndsay Wise, president and founder of WiseAnalytics and a monthly columnist for the BeyeNetwork, has for companies implementing business intelligence (BI) projects for the first time.
Last summer, Wise wrote a two-part series on how to limit BI project failures after noticing companies she was working with didn’t have a proper understanding of the undertaking or a way to measure their progress. She recommends that businesses start small by asking questions and creating “a formal scope of work,” and build big after that.
“It’s more of an iterative approach,” she said.
SearchBusinessAnalytics.com recently sat down with Wise to talk about her take on why BI is important and how companies can successfully implement BI projects.
Last summer, you penned a two-part series called “Limiting Business Intelligence Project Failures.” What drove you to write this series?
Lyndsay Wise: Different statistics quote that 80% of projects fail, but failure ends up being really broad. What I find with a lot of companies looking at business intelligence for the first time, what ends up missing is [this] advice. It’s not like large organizations. These companies are [implementing BI] for the first time and they don’t have anywhere to go to benchmark. A lot of companies I talked to need the advice … which helps them plan better.
In that series, you listed five areas that would allow businesses to implement successful BI projects. If a business was to heed only one piece of advice, what would you say is the most important?
Wise: I tend to break things down into two areas: the business/project-related side of things and the technical side of things. Realistically, a lot of times people really forget the business aspect of things, like what problems are we trying to address; what are we trying to achieve; who is going to be using it; what are we trying to get out of it? Companies sometimes look for the quick win. I’d recommend starting by doing something small but accurately. Many companies overlook this. Instead, they’ll start a big BI project with good intentions and then get bogged down by the details.
What are some common pitfalls companies fall into that lead to business intelligence project failure? And how can they be avoided?
Wise: My first example even goes back to the selection process, and that can affect implementation. Vendors do a great job of marketing, and companies don’t always understand what the differences are or how those differences will affect their organization. I see a lack of research or general knowledge to be a stumbling block. Companies will choose something that they don’t realize how much effort it takes until it’s too late.
On the other side, even though companies try to implement something successfully, they don’t take into account who will be using it. They’ll look at superusers and how they interact with BI. They’ll develop something that’s more high level and they’ll want to roll it out to others. If an organization is not taking into account the different levels of end users, then they’ve paid [for] and developed something that a majority of the people probably aren’t using.
You say that businesses should create realistic expectations, but you also advise understanding “what options exist and how each would affect the overall environment.” That sounds like a lot of legwork. For businesses, how realistic is this?
Wise: I think that’s why a lot of companies choose outside consultants or use professional services provided by vendors. It is a lot of legwork. I don’t think organizations -- if they don’t have BI -- realize the amount of legwork it takes. It’s not just doing and building from a technical perspective, but organizations also need to figure out what they want from a business perspective.
How common is failure when implementing a BI project?
Wise: That’s based on how an organization defines project failure. I don’t see -- or very rarely see -- projects totally blowing up and nothing working. But what I do see is projects going over time or budget or running into unforeseen circumstances.
Organizations will deploy solutions, and business units will come back and say, “I want X, Y, Z compared to what you provided us.” It’s more of an iterative approach.
In terms of outright failures or a project that doesn’t work, I rarely see that happening, but, at the same time, I’m not working with large deployments that have more to lose, let’s say.
So, if that’s the case, why should businesses bother? What’s the benefit of implementing a business intelligence project?
Wise: It totally depends on the organization. Different companies have different goals. What BI provides for a company on the back end is a way to consolidate information and get a broader view of what’s going inside the organization instead of having disparate systems running on different platforms. BI can tie in, for example, product sales to different stores and different suppliers, set goals, help determine what kind of planning is needed or evaluate performance in general.
A lot of the organizations I’ve worked with want more visibility into their company and identify opportunities they didn’t see because the information was in disparate data systems. They’ll all export their information to Excel, which will provide different views and different ideas of what’s going on, and then they’ll try to interact with that information. BI creates an effective framework and a set of standards.