Friday, May 26, 2006

I am on the *right* track !

There was an encouraging note from the IT Head of India’s one of the most respected business groups. He liked my views on the issue in the current BI industry and their solutions. My sincere thanks to him once again.

At the outset let me appreciate the articles written on this subject - the key point is highlighting the real issues which an user organisation must take into cognizance before embarking on such an initiative. I am fully in line with these articles hence I had contacted you.

I felt rewarded for writing that article and also feel more charged up to show the true value of BI to the decision makers.

Embracing Business Intelligence - Part 1

Purchasing BI tools, creating data warehouse & reports is not end of the BI implementation. To me its a begining of your journey to reap the benefits from this amazing science. Its important to have a comprehensive plan for adoption of this practice within the roadmap itself.

I would like to talk about an approach that can be taken to embrace business intelligence within an organization.

Take Small but Quick Steps If you are just beginning, dont start with a big bang approach with multi-million budget. In most of these cases, the individual decision makers dont get the required attention from the implementation partner (BI vendor). Prioritize the work, start with the function where the need of quick visibility is inevitable. With this approach, you will have more confidence on the effectiveness of this practice.

Simplicity is the KEY According to a survey, only 5% of the users of this system are advance analysts who care about flexibility in analyzing the data. Rest of the 95% care about the information in easiest possible way. For them the relevant information must be few clicks away. In the race of competition, the commercial BI solutions companies dumped a lot of features which are irrelevant to a big % of the user base. Infact with my own experience, a quick and relevant report in excel format is worth million than having the same from a fancy but complex to use tool.

Need to push off for a meeting, more in Part 2 :)

Saturday, May 13, 2006

Stepping into Decision Maker's Shoes

For the BI implementation partner, the first and the foremost thing is to get into the decision maker’s shoes and understand his pain points. The focus of the BI solution should be more about the user and less about the technology. Realization of true potential of the BI is all about transforming the data into information, passing it to the right hands and the right time.

There has been disconnects within the perspective of the IT side and the business decision maker. The IT side of it will have more focus on the tools, data transformation, etc. On the other side for business top questions will be: (a) Who are my most profitable customers? What factors make them so? What can we do to raise the profitability of other customer below this level? (b) Which are the top 5 efficient sales channel? Why? (c) etc

At core of it the decision makers needs the answers to What is the problem area, Why it is a problem & How can be solve / prevent it? This can be a big challenge for the implimentation vendor to map these requirements into tangible solutions. Enough care should be taken who will be user of the given report. This fact was well summarized by Gartner:

“The selection of BI tools and applications must be based on a segmentation of the BI user community. Different user segments will require different modes of delivery of BI, ranging from basic static reporting to sophisticated analytic applications. Deploying the proper mix of tools and applications is critical to achieving optimal benefits from BI.” — Gartner

Marketers and analysts—bridging the communications divide

Check out this article from senior members at Fair Issac. It highlights the common disconnets that Analytics and Marketing teams have along with their possible solutions. his is on the same line as the importance of having a strong linking between Business and Business Intelligence. I have personally observed this within my career. This article is must read for people in Marketing and using (or planning to use) Analytics for decision making.

Read full article.

Friday, May 05, 2006

Industry focused Centers of Excellence

This might sound weird and someone can ask that what is the need to having centers of excellence (CoE) within an analytics services company. Let me justify and describe the role they can play. In my last post on bridging the gap between Business and Business Intelligence, I mentioned that that understanding of the domain (industry) is the key to the success of a BI implementation. This skill set is missing within most of the companies who claimed to be an expert in BI. Knowing the domain better equips a solution provider to come up with a useful framework. Some of the roles these CoE can play are: Key Performance Indicators for Business After thorough research the CoE can come up with the KPIs for a given industry. The decision makers in the companies need to be on top of these numbers always. They can be used to measure the overall health of the business. The CoE should have an understanding about what all need to done in case of a large variation in these KPIs. This will help the people using BI taking the timely actions. Implementing best practices using analytics These CoE can collaborate with or follow the industry leaders and management consultancy companies for the best practices they use within a given vertical. For example, McKinsey coming up with their recommendations on how a company in retail can optimize its store operations, these CoE can understand from an analytics perspective. After going through proper research, they can come up with a framework, talking about what-it-takes and how to deploy the same using BI. Cross Domain frameworks Analytics world has evolved many domain specific terminologies. For example, revenue maximization is a hot area within telecom domain and this concept is not being used in any other vertical. I strongly feel that CoE have a role to play here in leveraging the successful models from industry in another similar domain. They can understand the core concepts and apply these principals for a new problem. The time is running out, need to catch up my team at 7 am. Will talk about some more charters that can be allocated to these centers of excellence. Their existence will definitely add a tremendous value in the BI value chain. Convinced? :) Happy Decisions !!

Thursday, May 04, 2006

Bridging the gap between Business and Business Intelligence

During several years of my career within BI space, I have always felt that there has been a big gap between the business and business intelligence. If you try to analyze the reasons for failure of the BI implementations, you will notice that the vendor who did a whole lot of hard(?) work in creating the nice dash-boards, charts and alerts lacked in understanding of the customer's business. Lets drill down on some of the reasons. Non-Actionable Analytics Lack of understanding of business leads to many fancy charts which show some values but this information is mere good to have. However a decision maker needs actionable insight which means, show what is not as per plan, what is causing it and how to tackle that. So it is important to define the information delivery that can take care of this issue. Lack of sense for Number This is another area where people skills plays a major role. Not many people in the organization have this capability to understand the trends being shown by the BI tools. Again its important for the BI service provider to design the report in such a way that the information can be assimilated by a layman. Complex Tools Simplicity kills !! I am firm believer of this. You can take examples of Google & i-Pod, they got hugely popular because of their ease of use. Most of the commercial BI tools available in the market are very complex. This leads to poor adoption (usage) of the framework. I do understand they try to provide maximum features they can offer to their customers, but 95% of the time a basic user uses 5% of the features. So simplicity plays a significant role in better adoption. I remember a self service reporting tool I created during my career at Andale, it was an instant hit just because of its very simple two clicks operation to get the required information. Adhoc Analysis A typical BI implementation will give you a visibility on the pre-defined metrics. However to use the real potential of the application, one needs to have people with understanding the data level to do the adhoc what-if analysis. These analysis plays a significant role in taking strategic decisions. In other words, creating standard reports is not BI is all about, one needs to to exploit it in best possible way. So an involvement of people having domain expertise as well as analytics perspective can be a great value addition to the organization. Apart from that most of the companies don't have the people who can understand the reason for drop or jump in a particular trend so a typical implementation can give you a signal that something is wrong but can't tell the exact area to focus. Identifying an internal (or external) resource will be a good idea. As per my understanding, BI implementation doesn't ends after creating data warehouse, reporting system and a couple of models but its just beginning. The BI framework will provide you the visibility and fun is to keep asking further questions (using ad-hoc analysis) to understand the area of focus completely. This is the point I feel one truly realizes the actual power of business intelligence and that's what I call empowering business decisions. Happy Decisions :)

Wednesday, May 03, 2006

The Birth of Analyticsworks !

Its official now !! Analyticsworks (http://www.analyticsworks.com) is here to help companies realize the true power of business intelligence without shelling out their fortunes on the astronomical implementation cost. Analyticsworks is going to use the Open Source technologies to minimize the total cost of ownership of the Business Intelligence implementation. These products are very stable now and can be used even for enterprise deployments. Much more details to come by end of this week !! Happy Decisions !