Recently released is a report by Gartner about all the things that can go wrong in your BI implementation. Aided by our teams’ experiences here at insightsoftware, we’ve summarized them into nine distinct threats to your BI implementation.
Of course, these threats can easily be overcome, but it’s a sure bet that they will try to creep their way into your business plans. In Gartner’s report, an analyst goes to great pains to say that there is “much more risk associated to non-technology issues than there is to deploying the infrastructure, tools, and apps.”
Reading this got us excited because if you’ve seen some of our presentations, you’ve probably heard someone here say that BI is more about people, process, and methodology than it is about technology. So, we feel validated. And since forewarned is forearmed, we’re going to walk you through these nine threats so that you can understand them and prepare to either avoid or to minimize them, at the very least.
1. If you build it, they will come.
Let’s start with this one. The “if you build it, they will come” mentality. Starting a BI project with this mindset implies that the initiative is sponsored by IT and is generally led from a technical or a data-centric perspective. The problem with this is that it is being led by IT, not by the business.
You should ask, “Who is going to use the system?” The answer is, “Never IT.” Unless IT has a business case or a user story of their own that they want to insert here, then they’re just not involved in the ultimate consumption of the BI deliverable and should play only a narrow role in its development.
The way to fix this is to make sure that the project team has significant representation from the business side of the house—from the very inception of the planning stage to the ongoing maintenance stage. For a complete list of roles needed for a BI project and their specific duties, check out our guide:
2. Your culture of Excel.
Threat number two is your culture of Excel. It’s Excel, we love it, you love it. It’s so hard to beat, but you know the drill. Because, admit it—you’ve done it, we’ve done it, too—you extract data from an internal system, and you load it into Excel, and then you manipulate the numbers.
The danger here is obvious: Each and every person who does this is going to have a different frame of reference. They are going to have different ways of combining numbers into metrics. They might even have different agendas. We can almost guarantee you different results from each, and you end up with no data integrity whatsoever. To us that equals one thing, and that’s risk. Risk to the business. It’s risk that comes from making expensive decisions on faulty information.
However, there is a solution here. The mechanical solution is to build a data warehouse. This will be the place where everybody has come together and agreed upon the data that you are going to use to manage your business, and how that data is going to be combined with which methods you’re going to use to calculate things like gross profit and net income.
The non-mechanical way to do it is to put a business sponsor on the team who believes in a transparent, fact-based approach to management. The sponsor should also have the clout to cut through cultural barriers and create change.
3. Data quality issues.
Here’s the ugly truth: Everybody has a data quality problem. The issues that cause it are ubiquitous and the impact on BI is significant. This is because people won’t use BI applications that are founded on irrelevant, incomplete, or questionable data.
The solution is to establish a process or a set of automated controls to identify bad data and block it from entering the data warehouse or BI environment. Interestingly enough, a BI environment in and of itself can solve this problem. That’s because nothing will highlight data quality issues like seeing them pop up in a dashboard. That will give you a solid clue as to how to chase them down. Learn how to prepare your data for BI.
4. There are way too many options out there.
If you type “BI solutions” into Google, or you go to any related tradeshow, or you read too many articles about BI, then you will get confused. There are a lot of answers out there. But how do you know which BI solution is right for you?
The way to fix this one is to not put the cart before the horse. First, evaluate your needs. Evaluate them from a business perspective, from a technology perspective, and then let that exercise inform the way that you go out and search for solutions and solution providers.
5. Let’s just use our current IT partner to do this.
BI is different from other projects. There really is a little alchemy involved because you’re dealing with human beings and their processes and the shifting sands of the business world all mixed together.
We’re convinced that you can’t rely on a company that dabbles in BI projects to make you successful. The best approach is to find a firm that is dedicated to building BI environments. You need a specialist for this.
Once you determine that they have the technical skills and the track record to make you successful, don’t be shy about evaluating them on fit, desire, and commitment. Those things are harder to assess, but BI projects tend to go on for a while so you will be working with these folks for a good period of time. Mistakes get costly in that scenario.
6. Finally, it’s perfect. We’re done.
Threat number six is really all about how long things should take in your BI implementation. This one is really about setting expectations, because another hard truth about BI is that no BI project is ever really done.
A BI project can go into a low-cost maintenance mode, but your business is never going to stop changing, and since BI is made to model your business, then it has to keep changing as well.
Even if your business didn’t change, the users will. According to Gartner, “Companies can expect their BI project to change by 35 percent to 50 percent in the first year after implementation, just due to user feedback.” That’s a good sign, by the way, because it means that your people are engaged and using it.
The way to prevent this from becoming a problem is expectations. You must expect that success in BI is a moving target over time. Let everybody know how it’s going to work. Plan for it, schedule deliverables against it, and budget for it.
7. Just give me a dashboard. Now!
We’re all guilty of this one sometimes. It just doesn’t seem like it should be that hard to produce the numbers that people need to do their jobs. From the outside, and even from the inside sometimes, it’s hard to understand all the moving parts that need to work together to surface your data.
The fix here is kind of like the last one: Don’t be afraid to set proper expectations on time frames. It’s also about being willing to spend the time and the money and to be able to tolerate the risk of a real BI project.
Also, be wary of vendors out there baying about their pre-built reports that are ready for you to use out-of-the-box. These days, it’s pretty easy to do that. Instead, the most important questions to ask about the reports are: Are those going to be worth anything to you? Do they reflect your business needs? Do they use data from all of your systems? Do they have the industry specific KPIs that you need? Do they model the way that your people need to see your data? That stuff is really unique to each company, and it all takes effort.
8. Why is my gross profit figure different from yours?
Everybody has been to that meeting where two finance people show up and . . . guess what? They have slightly different numbers. The first 15 minutes of the meeting is spent arguing about who is right. That’s a big time-waste and, more importantly, it destroys trust. It destroys trust in the numbers and it destroys trust in each other.
We’ve all heard about the concept of the one version of the truth and that’s what a data warehouse is for. At insightsoftware, we always start talking about this concept one step before the data warehouse. It is imperative that our company get together and agree on the data before we start building that data warehouse. For example:
- How do we want our data to be structured?
- What do we actually mean by sales?
- How do we segment our product lines for reporting?
The way to prevent this threat is to commit to doing this foundational work and to give it the energy it needs upfront. It will pay off.
9. Strategy? That sounds like overkill.
This could actually be the biggest threat of all, because failing to devise and then document the BI strategy will become a problem for you.
BI is something that can really turn the light on in your company. Imagine everybody drawing on the same pool of valid data and trusted numbers. Everybody focused on the same numerical targets. Working with the same KPIs to drive action and improve processes. It’s the power of data, it’s what it does. But it doesn’t happen by itself. And it certainly doesn’t happen because you buy and install a piece of software.
Strategy sounds hard and laborious, but it doesn’t have to be. As Jack Welch, the former CEO of GM, says about strategy, “Get into a room, work out your plan, build your spreadsheets, but don’t spend TOO much time on it.”
That’s because no plan survives contact with the enemy. Things are going to change. Think of your strategy document as a road map. It should be a plan, or a reference, or a guide. If you make some small adjustments on the way, it’s okay. That implies health and growth for the company.
Those are the nine threats that can hinder you progress on this BI journey. Hopefully, this has given you a guide for things to watch out for. Contact us with questions or browse some of the additional resources below: