Learn from the growing pains others have overcome
As project leader, you’re cradling an ambitious BI project, and you’re looking for theoretical and (especially) practical frameworks. You know that you can achieve better results using the right management information. Googling “Business Intelligence” returns 527 million results. “BI Tools” narrows it down a bit, but still returns 855,000 hits. Discussions are spreading out in all directions and it’s hard not to get dizzy. If you want to get serious about working with BI, it’s easy to end up in a roller coaster. You’ll want to get both feet on the ground again as soon as possible. But BI is dreaming, daring, and especially doing. But you don’t have to start from scratch; you can learn from the growing pains that others have already experienced. The road to hell is paved with good intentions. after all. Learn to recognize these symptoms and avoid them, or deal with them. This stories are based on the goals and ambitions of the hundreds who have taken our Business Intelligence training course.
Growing pain 1: Doing more with less
“The board of directors has very lofty ambitions, but has unrealistic expectations about the output of our department and the quality of our information supply. In reality, we still don’t have one version of the truth. The CEO heard through the grapevine that our direct competitor does have their business in order, thanks to a promising new tool. ‘Why don’t we have this,’ they wonder aloud. The ivory tower then issues forth global directives, without freeing up the budget for new people, systems, and (visualization) tools. To put it in stronger terms, I get the sense that directors in the boardroom are just making it up as they go. They’ve barely done any research and haven’t given the staffing and budgetary consequences any thought. At the same time, I have to admit that I don’t exactly know what kind of investment (inspiration, money, and people) is required to achieve real success with BI. The management team has to believe in the strategic value of BI and free up time and resources, at the cost of other projects. The directors responsible, however, are negligent in making any real choices. The innovation budget has long been used up, and the company has instituted a hiring freeze. In short: we’re missing the people, resources, and support. The classic pitfall “doing more with less” is a daily frustration.”
Growing pain 2: I can’t make the business case
“During my studies I’ve learned that the added value of BI is at its biggest when there’s an ample supply of good data, there’s a lot of competition, and the will and necessity to continuously improve and innovate (based on good data) are present. But that’s in theory. In practice, things aren’t as cut-and-dry. Directors are easily swayed by a polished PowerPoint presentation stuffed with platitudes (“we’re living in a dynamic, turbulent world, and we have to proactively work on the digital transformation”) and are rarely interested in a thorough, well-argued business case. How can I make the added value of BI explicit and financially measurable? How can I discern the difference between what’s relevant for us and what belongs to the category “hype and buzz words.” Drafting a structured document with a clear cost/benefit analysis, which makes a careful business consideration on whether or not to start a project – that, I can do. The direct, indirect, predictable, and unpredictable advantages of BI are shining on the horizon. But how do I present the business case convincingly in the boardroom? How can I get them to give BI the green light? How can I increase my persuasive power and force a decision? I often get the impression that I’m not taken seriously, so that they don’t make a real decision and I’m left empty-handed.”
Growing pain 3: Getting lost in the “jungle” of BI tools
“When it comes to topics like big data, business intelligence, data warehouses, and data lakes, there are dozens, if not hundreds, of tools on the market. The amount of tools, platforms, and architectures is overwhelming, and it becomes hard to see the forest for the trees. I just don’t have the man power to judge the functionalities of all the tools in detail. At the same time, switching between the various tools and apps in the workplace leads to a loss in productivity. If I had my way I’d have a set of tools with a low barrier to entry and one clear interface that connects all the tools and apps. Which tools are useful in quickly translating data into information, and information into insights? To put it concretely, I’m looking for objective and clear selection criteria that I can use to rank the available BI tools from “need to have” to “nice to have,” so that good tools aren’t discounted just because they happen to score poorly on one unimportant criterion. But training users also has priority, because everyone knows that a fool with a tool is still a fool.”
Growing pain 4: Are we using the right management KPIs?
“In our organization, we use dozens of (legacy) KPIs as management information. We’re all so comfortable with these KPIs that nobody dares to question them anymore. They’re taken for granted. They reflect our truth. We’re caught in our own filter bubble. I’m starting to feel less and less comfortable with this. How are all these KPIs still related to the strategic goals of our organization? How do I know that we aren’t headed straight for a cliff? Isn’t the dashboard cluttered with all those KPIs, lights, and gauges, while the real insight is missing? Have we ended up in a reports factory, where users can request reports all the live-long day, without actually reading them? Don’t be led by dashboards and reports. To what extent are we dealing with duplication: aren’t we delivering some reports, information, and insights twice?
In short: I’m driven to start looking for genuine KPIs. And what, really, is the difference between key performance indicators and key result indicators? In asking these kinds of questions, I’m finding out that the theoretical foundation of our BI architecture is paper thin. We should start building an entirely new BI framework.”
Growing pain 5: I feel like a toy of demanding users
“Users of BI systems can be found on the executive level and in the workplace. What they have in common is that they’re generally critical, impatient, demanding, and sometimes downright difficult. They always want custom reports. To what extent can scrum teams provide solutions to getting BI projects started? The user often plays the role of a tester and has an important voice, and rightly so, in the acceptance of the BI system. But I don’t want to put myself in a place of “you say jump, I say how high.” I do my best daily to translate the technical BI jargon into layman’s terms for all stakeholders. I strive for one version of the truth and try to prevent a BI trauma by not over-feeding users with data. Despite all that, it’s tough to put (and keep) BI on the agenda with management. The declining interest of users is also worrisome. I often wonder how I can keep the BI system as clear as possible for all users. On top of that I try to prevent a jungle of reports. I also want to serve multiple audiences with our BI strategy, both department managers, operations, the board of directors, external quality organizations, but also the research department, for example. I also want to create more added value from my function. Despite all these good intentions and hard work, as project leader I’m often the whipping boy. When the system is unexpectedly down for a few hours, for example, all the users blame me. In short: Would a self-service BI system be useful for us? With a solution like that, users can request the information that’s relevant to them. If needed, they can subscribe to new versions of reports.”
Growing pain 6: Shared responsibility is no responsibility
“Effective teams are characterized by their multi-disciplinary nature. They comprise project managers, BI consultants, BI architects, technical project leaders, (internal) clients, sponsors, ETL developers, business analysts, users, data scientists, database administrators, and system administrators. There are base roles and additional roles. Trainers, metadata administrators, data stewards, testers, and usability experts are examples of functions that can be deployed additionally. Base roles are required in all BI projects, no matter how big. Additional roles are mostly relevant to larger BI projects. Per role, the most essential skills that are important to the success of the BI project are displayed. In that way, we’ve mapped all the functions, tasks, and responsibilities using function descriptions and competency profiles. But paper is patient. In practice, it often proves difficult to separate the various roles. The question “who does what, when, why, and from which role” becomes very relevant in this context. I know: shared responsibility is no responsibility, but I haven’t found a real solution yet. The question of what position BI should have in the organization hasn’t been fully answered yet either.”
Growing pain 7: Data quality is an issue that keeps haunting us
“Data is correct or incorrect, there is no compromise. Data can’t be half correct. Meaningful data should be unique, consistent, current, logical, and complete. Every BI manager knows this. The reality is that data often appears fragmented in every corner of the organization. I would like to spot incorrect data sooner and present information faster and clearer to users. Add to that that data can seem correct in a separate source system, but when combined with other sources, mistakes and omissions can crop up. To make a long story short: I wouldn’t vouch for the quality of data in our organization. This begs the question if it’s such a good idea to start using BI if the data quality is so low. While this discussion is on-going, we’re looking for ways to improve the data quality. But what measures can you take? Should we go for a traditional data warehouse, or for a Data Vault that’s suitable to the demands of the external supervisors? And what does this choice mean for our data quality (now and in the future), the complexity, and the lead time of the project? One thing is for sure: our management team won’t have to just steer on KPIs, but also on data quality. “Garbage in, garbage out” means that software and rule systems can only provide useful information when they’ve been fed correct information. A computer system can be the most intelligent thing in the world, when it’s fed incorrect data, it will certainly generate a useless result.”
Growing pain 8: The data gap is widening
“The available time we have as an organization to make well-founded decisions is greatly decreasing, while the complexity of decision-making is only increasing. On top of that, the amount of data needed to make good decisions is growing exponentially. We simply need much more data from multiple sources. The volatile nature of markets, as well as consumers’ fickleness, requires more precise decision-making and larger volumes of data. These two variables, decision-making time and data volume, are opposites. This creates a BI gap and these two opposites will only keep drifting further apart if our organization doesn’t take action. As a result of shrinking margins of error and shorter time intervals, we need finer and more sensitive precision tools to cross this BI gap. The workplace is struggling with dated architectures and a great diversity of source systems. These legacy systems are hindering us from truly working with BI. For example, how can we ensure that the information provided by the legacy systems remains available to BI and Research departments? What’s the right angle of approach?”
Growing pain 9: Data governance is causing us problems
“Data governance is a buzzword that we can’t get a grip on. Data governance isn’t just about data; that much, we’ve understood. The concept requires a combination of people, rules and procedures, and supporting technologies to be successful. And that combination is a careful balancing act. Unfortunately, there’s no singular, workable definition of data governance that we can work with. The Data Governance Institute makes a noble effort to define data governance as “a system of decision rights and accountabilities for information-related processes, executed according to agreed-upon models which describe who can take what actions with what information, and when, under what circumstances, using what methods.” But that’s not the end of the story. That definition is still too general and broad. We’re primarily looking for practical handholds and useful tips to really support concepts like data governance, data lineage, and data classification in the workplace. I’m mostly wrestling with the question of how we can demonstrably remain in control when it comes to data.”
Growing pain 10: How can we escape the Excel hell?
“Everyone knows by now that Excel is a very useful tool, but not the sharpest one in the shed. We also know that it’s very labor-intensive to generate the right insights from all kinds of data sources. We know intuitively that Excel isn’t the best tool for this, because spreadsheets are very prone to errors. 95% of spreadsheets contain at least one mistake. Despite all that, we just can’t manage to definitively axe spreadsheets.
When putting together management information, we keep running into the same questions. How can we arrive at a small set of meaningful indicators and genuinely relevant insights? How can we get rid of Excel, and which affordable alternatives are out there? How can we consistently use the information for analysis, action, and process improvement? How can we employ the numbers in the planning and control cycles? And how can we integrate feedback in the process? How can we automate the process of management information as effectively and efficiently as possible? In short: switching from ad-hoc Excel reports to real-time BI software comes with a lot of baggage. Meanwhile, we’re ruled by the issues of the day as I consider the question: how can we make BI part of the Marketing and Sales disciplines? Finance is assertive enough and traditionally focuses on fast closing of financial figures and reports while demanding that the information is current.”
Growing pain 11: Lack of data-driven company culture
“Most of my colleagues aren’t steering based on facts and data, but are still following their gut feeling. Data-driven working using BI tools and applications has consequences for our employees, teams, and probably also for our other stakeholders. Does this mean we have to retrain everyone? Do we have a budget for that? What are we actually working with? For instance, it’s tough for me to judge how ‘data literate’ our organization actually is. Is our company culture suited to BI? Data literacy is getting more media attention, but I notice that we haven’t developed any policy around this. Can everyone work with BI software? Is everyone capable of reading graphs and drawing the right conclusions? Not everyone can formulate a sound hypothesis and test it using data. I struggle with questions like: how many data-illiterate people are we actually dealing with in the organization? To what extent can people be tested on this and categorized in levels of maturity based on their test scores? I’d like to help my colleagues get more out of data and increase their data literacy based on these insights. But no one wants to be labeled as computer or data illiterate. I don’t want to disrespect or cause offense, but I also want to make strides with BI and show quick wins.”
Growing pain 12: Keeping everyone on the same page
“My biggest challenge is keeping everyone on the same page. How do I create commitment with the (internal) clients who will have to use the BI tool and report functionalities? I also have to make sure management stays interested in this topic. How can you get management to the point of fully committing and staying involved? It feels like playing chess against myself. At the same time, I realize that these are sensitive areas. Should I design a separate process and communication plan? How can I approach implementation project-based? When do I involve the various stakeholders in the project? How do I control potential risks? In short: I don’t know where to begin, and what the complete track looks like. I’m also very curious about how we score compared to similar organizations. What BI maturity level have we reached, and which milestones are still on the horizon? How do you draft a BI mission, vision, and roadmap? I’m looking for practical approaches to this.”
Intelligent organizations think about the question of how they can achieve a higher BI maturity level. They realize, however, that just like in real life, growing up is hard and comes with a lot falling down and getting back up. Research shows that there’s a greater chance of success when an organization deliberately strives for a high BI maturity level and makes good on it in practice. After all, there’s little point to applying BI just to understand and coordinate the organization better. They want to improve and innovate successfully based on valuable insights, KPIs, analytical models, and material expertise. That means that the focus should be on the long term. That’s the only way to guarantee a return on BI investments.