Tuesday, July 6, 2010

Beyond Self Service BI

In the past few years there have been significant changes in the BI Industry: acquisitions, consolidation and evolution of graphical capabilities leading to a paradigm change on how the BI tools are being leveraged in the market today. This paradigm change has provided many companies with the opportunity to introduce a Self Service business intelligence strategy, albeit for different reasons. Some companies wanted to try the Self Service model to give a fresh approach to BI, as previous attempts of building an information driven culture had failed, and the new BI manager was looking for a way to make their mark and re-engage the business stakeholders: The Do-Overs. Other companies were trying to get rid of the “problem”, it was very expensive for IT to keep responding to all the changes the business folks needed to make; it would be a lot easier (and cheaper for IT) if they were able to do those reports by themselves: The Cost Cutters. There were some national and global companies which wanted to appease complaints that they were not nimble enough and that IT was not satisfying the local reporting requirements, as they were concentrated only on corporate reporting: The Peace Keepers. It was really a very small group of organizations, usually lead by a visionary business leader, which were looking at BI Self Service as a way to empower the organization to make the right decisions across all levels: The Next Gen. This article will explore what kind of results each of these groups of companies achieved in their journey to Self Service BI and shed some light on the different levers Business Intelligence professionals can pull in their own organizations to influence their journey towards Self Service BI. These four companies groups are depicted in the diagram below.

The Do-Overs
Companies in this group want to re-do their BI implementation because something did not go well in the past; they are trying to run from their mistakes and use the excuse of implementing a new tool or technology to do over their failed implementations. They want to implement Self Service BI because they think it would give the new toolset they are planning to implement a fresh approach and make it feel differently to the business stakeholders. You might be wondering, does this work? The answer is absolutely Yes, Self Service BI can actually be used to position a whole new different paradigm for BI. The problem is not Self Service BI, or the new toolset, but rather the people in these companies are so worried about not repeating the same mistakes from they predecessors that they do not realize that Self Service BI can only be implemented under the proper BI ecosystem: the technology platform is stable, the data quality is good and the business users want to learn how to pull the information themselves. Starting from the top outlining roles, responsibilities and tool permissions or capabilities will get the business people excited at first, but invariably when the business users realize that the metrics they need are not well defined, the data is incorrect and the system is very slow, the new BI program looses all credibility and it is time to Do-Over again.

The Cost Cutters
It is fair for any IT department to look for ways to decrease cost and optimize resources. Given the economy today, this is something that many CIOs will be expected to do this year. However, like when doing pruning, you need to know where to cut and how much to cut. Positioning Self Service BI might help lower IT costs in the long run, but it most likely will demand a heavy investment in time & resources upfront that these companies fail or refuse to plan for. If the business users are habituated to going to IT for any single report change, no matter how small it is, it becomes very difficult to break this dependency. In most companies with this situation, business sees IT as a black box of tremendous complexity with an army of people that work behind the scenes to make things work. They see IT as being at a different plane of existence where they are not worthy of understanding what happens there. So imagine their surprise when they get to work one day and find an email in the magic box (aka computer) telling them that from now on they will be responsible for creating their own reports and managing their own changes. It is like telling the mortals that the gods no longer consider them worthy of fire and now they will have to live in the dark. Following this analogy, if the users have never seen how IT creates and modifies reports, they do not know of SQL or MDX, the users most likely will never be able to figure out how to leverage the BI Self Service capabilities on their own. Many companies have spent millions in infrastructure, licenses and training just to find out six months later that nobody is using the new Self Service system and the business has hired a bunch of data wizards that are using a combination of Excel, Access and the old BI system to run the company. Needless to say that in these scenarios the savings never materialized and the company who was trying to save money, ended-up spending even more for a not so cheap and even less Self Service BI solution.

The Peace Keepers
The people in charge of the BI strategy strongly believe that they are implementing Self Service BI for all the good reasons: they want the business users in each of the local markets to be able to produce their own reporting that better allows each market to respond to unique local challenges. They are also aware of the pressures that the local markets are putting into corporate for defining a centralized view of the world, which does not enable local reporting. These BI managers blindly trust that if everyone is given the capabilities to customize their own view of the world they will not challenge the corporate definitions as they can use the Self Service capabilities to define unique reporting hierarchies and structures for their particular market’s use. In the long run, what ends-up happening in these situations is that each market creates a parallel structure to the official, corporate one and gradually the local reporting needs keeps growing more and more apart from corporate. A few years pass and the “guy” who knew how the local reporting systems were set-up leaves the company and IT is asked to take over these “applications”. However, given that the corporate structures and view of the data is now radically different, IT has no clue on how to support these local reports. Then people suddenly realize that letting each of the markets implement their local reporting needs on their own is not a good idea, and they should centralize the generation of reports. A business case is built with the potential of millions of dollars in savings, but only if the markets are willing to relinquish control back to a centralized entity. At this point the same battle that they were trying to prevent in the beginning is about to break loose with no clear winner in sight.

The Next Gen
No, these are not the companies of the future, but they could as well be. These are the companies that see the implementation of BI Self Service as the next generation of Business Intelligence, not because the previous generation failed, but because their predecessors established a solid foundation with the right metric definitions, excellent data quality and a solid technical platform. The Self Service BI culture is driven by an enthusiastic business sponsor who is pushing the use of information to take decisions directly from the top. These organizations continuously organize trainings and information sessions to share best practices and let the users from different business areas connect among themselves building communities around the reporting applications. A company in this position has probably failed implementing BI before and they have realized that the problem was not on the tools but rather on key foundation blocks (Metric definitions, data quality and proper infrastructure) that were not present or were incomplete.

Never underestimate the commitment that BI Self Service requires. Most of companies that have implemented Self Service BI successfully experienced drawbacks early in their journeys that provided critical insights and learnings to course correct. Let us take the example of a retailer that started this journey in the early 2000’s. This particular retailer already had a solid BI program enabled by an in-house developed tool, and had established solid metric definition and data quality principles. Business users had limited Self Service exposure as they could choose any combination of attributes from a limited product, location and time selectors. While the capabilities enabled some degree of self service, most of the users had predefined reports which they looked at every day/week/month and seldom ventured in ad-hoc capabilities like “adding new stores” or “categories” to their reports.
Given the changing competitive landscape, this retailer decided to revamp their business intelligence infrastructure and bring market leader tools, the investment was in the millions of dollars and many stakeholders in the organization saw it a as a waste. “We have all the reports that we need” they used to say. The first release was not successful, in a high profile meeting the new set of tools performed slower than the in-house developed application and the information provided was incorrect. That meeting almost killed the program before someone realized that the problem was in a piece of equipment that had been wired incorrectly. This setback instead of demoralizing the BI team, pushed it try even harder. The BI organization was divided into two teams, one that would focus on the technical issues and one that would focus on how to get alignment and buy-in from the business users. This strong teaming, coupled with a new approach on Self Service BI helped the BI organization to generate enough buzz with the business stakeholders to give it another try.
After only six months, the business users were impressed with the capabilities of the new tool, not only could they select any product, location or time period they wanted but they had access to new metrics, and functionality that never before any of them had seen. For example, the new platform provided Market Basket reports On-demand, where the users could specify a driver product and the system would identify all the items that had some affinity. The business users now had the capability to run full ad-hoc reports, selecting the KPIs they wanted to see in real time while they were negotiating with vendors and suppliers. The business users could compare the performance of a particular product against a competing product, or look at its performance during different time period when the item used to be on sale, promotion or temporary price reduction.
The level of excitement grew daily as the BI business team had highly talented individuals walking the halls of the merchandising, pricing, marketing and supply chain organizations daily, helping the business users to leverage the new Self Service capabilities. A key element for success was the establishment of a monthly breakfast where one of the business users presented the results that they had achieved leveraging Self Service BI to their colleagues and got recognized by one of the Senior Vice-presidents of the company for their contributions.
In just a few months, a project which very few people believed-in, had suddenly jumped to the spot light, providing the company with a tremendous competitive advantage. Years went by and the popularity of the system kept growing, store users started asking for having access to the Self Service capabilities and they were willing and able to take all the necessary training to properly use the system. The company enjoyed a period of incredible growth fueled by business innovation initiatives that would not have been possible without the implementation of Self Service BI.

No comments: