Treating Your BI Project Like an Entrepreneurial Startup
Startups know that what customers say they want isn’t necessarily what they want. The same principle is critical to the success of your BI project.
- By Ken Collier
- May 15, 2012
As a proponent of agile data warehousing and business intelligence, I am constantly looking for new techniques for delivering value to customers faster, adapting to their feedback, and evolving toward the right business solutions (regardless of initial requirements). I recently read the new book, Lean Startup by Eric Ries (Ries, 2011) and it has rocked my world. In the short time since this book hit the shelves in September, 2011, it has exploded in popularity. Be forewarned, this book is about entrepreneurship and high-tech startups. It isn’t about data warehousing, BI, or analytics ... or is it?
Most startup companies fail. Startups typically rely on a good plan, a solid strategy, and thorough market research. The problem is that startups operate with a high degree of uncertainty. They don’t yet know who their customers are or what their product should be. They can’t predict the future. “Planning and forecasting are only accurate when based on a long, stable operating history and a relatively static environment. Startups have neither,” Ries points out.
Corporate DW/BI/analytics initiatives have much of this same uncertainty. Our customers are highly diverse, from across the enterprise, acting in varying roles, and having rapidly changing goals. For many reasons they can’t accurately tell us what “products” they want, and they don’t fully understand what our BI systems can do. Moreover, our BI strategies and “market research” are based on yesterday’s business needs, which are in a constant state of change.
My clients routinely tell me stories about building dozens of BI reports based on customer spreadsheets or requests, only to discover that most of these are rarely or never used. Like startups, enterprise BI initiatives don’t have a long, stable operating history or a static environment.
Ries points out that “The fundamental activity of a startup is to turn ideas into products, measure how customers respond, and then learn whether to pivot or persevere.” Lean Startup techniques follow a Build-Measure-Learn feedback cycle. This cycle begins with an idea or hypothesis immediately followed by building a minimal viable product (MVP). Customer response to this MVP is carefully measured and the resulting data provides the basis for learning and adjustment. The goal is to move through this cycle as fast as possible, and as many times as necessary to converge on the product that customers want.
This cycle is aimed at quickly building something, getting it in the hands of customers, and measuring their behaviors. “Instead of making complex plans that are based on a lot of assumptions, you can make constant adjustments with a steering wheel called the Build-Measure-Learn feedback loop.,” Ries explains. In this cycle, we start with assumptions, build the smallest/simplest product, test our assumptions, and decide whether to pivot or persevere.
Two critical elements of this cycle are the MVP and the validated learning that is based on scientific testing of customer acceptance. The MVP is the very smallest, fastest thing you can introduce to your customers to gauge their response. For a business intelligence “product,” this might be a disposable prototype report or dashboard mockup populated with snapshot data. For analytics, it might be a mockup of a scoring algorithm based on a rudimentary predictive model. It is the simplest version of what we think customers want, so that we can find out if our assumptions are correct.
It is essential to quickly get the MVP in front of your customers and scientifically measure their behaviors. Is there a growth in the number of business users who demand access to the new “product”? Is there a deepening in the usage of the new “product” by existing users? Designing the right metrics will provide you with the evidence you need to either persevere (continuing to invest in the maturity of the “product”) or pivot in favor of a different hypothesis. Designing the right metrics for validated learning is difficult but essential. Ries’ book provides a great set of guidelines for developing these measures that we can apply in our DW/BI/analytics environments.
Giving the Customer What They Want
Sometimes we accidentally build something that nobody wants, in which case it doesn’t matter if we do it on time and on budget. The goal of a startup is to figure out the right thing to build as quickly as possible. Agile BI development helps accomplish this by delivering a few working BI features every few weeks. Agile BI calls for a prioritized backlog of user stories, which are the smallest aspects of business value that we can get our customers to articulate. Agile BI calls for building production-quality features during every iteration and deploying them into production as often as it makes business sense to do so.
This is highly effective way of evolving toward the right solution, assuming you can get your customers to accurately articulate and prioritize their stories. Lean Startup techniques make no such assumptions. The question in Lean Startup is: How fast can we decide to pivot, and how many times must we pivot, so we can ensure that we are building what customers want?
Once we have correctly discovered what customers want, then we can use agile BI techniques to build, refine, and mature the solution. Ries describes this approach as “…killing things that don’t make sense fast and doubling down on the ones that do.” This theme makes as much sense for BI directors as for startup entrepreneurs.
We must learn what customers really want, not what they say they want or what we think they should want. We must discover whether we are on a path that will lead to growing a sustainable data warehousing, business intelligence, or analytics program.
Data warehouse and BI program leaders are entrepreneurs within the enterprise. It is the job of these entrepreneurs to quickly determine which efforts are value-creating and which are wasteful. By fostering Lean Startup thinking within your BI department, you can effectively establish a pattern of learning and adapting, which is the essential measure of progress for startups. It’s time to start thinking of your DW/BI/Analytics initiative as a startup. Will it be a successful one or not?
Eric Ries, The Lean Startup: How Today's Entrepreneurs Use Continuous Innovation to Create Radically Successful Businesses. New York: Random House, 2011.