Geeks With Blogs

News Brandy Favilla is currently the Community Engagement Manager at New Horizons of Minnesota (www.nhmn.com). In this position, she helps people define and build their careers through educational events. Brandy focuses on networking with members of the Information Technology community to form partnerships in education. She strives to maintain New Horizons’ reputation as a high-quality learning center throughout the Twin Cities and beyond. Some of her accomplishments include chairing the committee for both the TechFuse and MindSurf Conferences hosted by New Horizons, co-chairing Twin Cities Code Camp and Twin Cities SharePoint Camp, Co-host of the Twin Cities Developers Guild, VP of PR for TechMasters (www.techmasters-tc.com), Communications Chair of PACT (www.pactweb.org), Programs chair for IAMCP, President of the Business Intelligence User Group (www.ncbisig.org) and actively participating in the SharePoint User Group (www.sharepointmn.com). *the thoughts and opinions in this blog are my own and do not reflect the opinions of New Horizons of Minnesota
Twin Cities IT and Developer Events Postings by Brandy Favilla

Business Intelligence Requirements
How to Conduct Business ANalysis to Ensure a successful BI Initiative

Trent Seashore and Dean Larson from SafeNet Consulting

My biggest takeaway:
BI Terms and Acronyms for Dummies
Business Intelligence; data warehouse; database data mark; data; store; OLAP; ODS; DQMgt; Data Analyst; Data modeler; ITL; Data miner; Information management; Data Lice Cycle Mgt; dashboard; scorecard; Performance mgt; data stewardship; Larry English; Data Quality; Data definers; Data users; Data sponsors; data architecture standards; ad hoc; bi architect; bi systems engineer; bi system administrator; BI competency center; OLTP; SOX compliance; Enterprise DW; DSS; informatics; Analytical modeling; metadata; Master data repository...

Overview of the presentation:
Did an exercise as a BA analogy... Fantasy football and Gas stations today vs. 15 years ago.

Build a successful team in fantasy football analogy.  We wrote a requirements document with ideas from the audience.  We talked about why it needs requirements, where this information should be stored, how our audience is going to access this information... when would a team owner need this information?  daily, before the draft, after draft, right after Monday night game.

Fantasy football - 15 years ago vs. now
1993 - Manually score & tabulate NFL Football boxscores from the newspaper on Monday morning – Average of 10 hours per week minimum to tally, score, manage & mail results.
2008 - Fully customized, real time scoring, alerts, symbols (visual), completely automated and/or self serviced by the customer

Another example that he gave was gas stations - 15 years ago vs. now
1993 - Full service or at least everyone paid inside (SA’s would close at 11pm)
2008 - Pay (a lot) at the Pump: 24x7 not as many customers pay inside – self serviced by the customer. How do we get the customer into the store to buy higher margin products? (promotions at the pump – video displays at the pump, personalized dynamic promotional offers….)

Background and examples of BI Projects
Differences and similarities between BI Projects and other application projects
Tipcs and TEchniques to consider on you rnext BI project

BI Definition:  Evolving strategy, vision, and architecture aimed at aligning an organization's operations with its strategic goals through rapid and easy access to actionable information
Can also be a reference to technology solutions that help deliver the above definition

Pulling information vs. pushing information

It's not a product or a system... not a stop or an ending... BI Iniatiatives should not be focused on the tools used, but the business needs and the use of the data and when its needed


Current trends

past - support mangement queries and analysis on a static store of historical transaction informaiton
today - identifying and responding to real-time business events and making timely informaiton available to operational decision makers
past - information latency was an acceptable characteristic of DW batch reporting
Today - standard is real time information

Data vs info vs knowledge
Data - pieces of information that have little meaning by themselves. e.g., 600 (raw, finite, simple values)
Information - collections of data that have relevant meaning and business context/definition.  E.g. fico = "600", borrower name = "fred"
Knowledge - the qualitative and quantitive understanding and use of information by a human being to help make business decisions and increase profits

Vision of Future BI State
Timely access - information consumers (individuals and systems) ahve access to the data they need when they need it
Reliability - improved information integrity across papplications
flexibility - data is structured to supprt common and indifidual business needs
quality - users can compare information from across the enterprise with confidnece
control - clear stewardship of quality and content.  Data stored in the minimum number of locations necessary.  Data movement is controlled and managed.

The overall process has not changed but distribution of data has changed significantly - speed of informaiton requirements has changed dramatically

Characteristics of BI Projects
1) Broad Audience – all levels of business users
(e.g. C-level Execs, Customers, Finance, Marketing, Sales, IT & Ops staff).
2) Very Tangible Outputs – CEO’s will gladly pay for these projects if the delivered output can be demonstrated on their Executive Suite PC.
3) Real time delivery expectations – desktop solutions, handheld delivery, text messaging.
4) Proactive Alerts – for opportunities as well as concerns, different users want different types, personalization, opt-in vs. required.
5) Delivery and distribution is critical – the vision of BI demands flexible and robust delivery.
6) Very visual – audiences’ needs are for visual summaries of key information (e.g. pie charts,wizards, alerts, dashboard indicators)
7) Need to demonstrate value quickly – normal SDLC may be too long and cost too much.
8) Desktop BI is an attractive option – quick delivered results often satisfy stakeholder needs, hand grenade theory “close enough” may be appropriate.
9) Multi-source is norm – integrate, analyze and deliver information that’s sourced from varied businesses and systems.
10)Speed is key – data retrieval, summary, and communication is often a key to project success.
11)Unique skills and roles – BI BA, KPI expert, Balanced Scorecard expert, DW Arch w/BI logical & physical data models, cube developers,data personalization experts, Data PM, Data Quality SME, Data Stewards/Governance

BA Tips for BI Projects
1)  Continually clarify at what level of process decomposition the stakeholder needs are at.
2)  Investigate whether any process decision points can be automated based on the BI analytics that's now being captured.
3)  Consolidate stakeholders into groups of users. The fewer the better. Once these groups have been identified (and named), then personalization can be done for these groups.
4)  Push back (when appropriate) on stakeholders' stated need for information. Try to avoid “everything but the kitchen sink” requirements. Question stakeholders in a way that makes them explain how particular information will add value.
5)  Alerts: What business events should trigger alerts? Who should receive them? Make stakeholders justify alerts in order to avoid desensitizing overuse of alerts. Determine delivery methods (e.g. email, mobile device, dashboard, calling tree).
6)  Metrics: When identifying what should be measured, consider what's useful for "downstream" business process participants as well as "upstairs" process owners.
7)  Prepare for a BI project by learning about and understanding the systems and data repositories that contain the data that will feed the BI analytics engine.
8)  It’s crucial that the source data is heavily scrutinized for quality (junk in /junk out)
9)  Identify which data is important, then find out how clean it is. Where appropriate, datacleansing plan must be developed and used. BAs need to understand the meaning of source data so that it’s quality can be analyzed.
10)  Meta-Data is critical. Valid business data, unless tied to its meaning, is still meaningless.
11)  Assess the BI users capabilities (power users or casual viewers) and create outputs accordingly

Ask these simple questions & then listen for opportunities
Who is using data? Number & Types of users (low, medium, high)
What data are they using? How many attributes, tables, tools?
Why do they need it? Nice to have vs. Mission critical, operational metrics?
Where did they get the data? Source systems, tables, tools?
When do they need the data? Annually, Monthly vs.
(Weekly/Daily/Hourly/Minutes/Seconds......) Real time or not?
How do they receive/view data outputs? Push or Pull strategy? Web,
FTP, Paper, CDRom, tape, CSV to other application, Excel, Graphical Dashboards, etc…
What tools do they use today? Which Data Mgt & BI tools, or MS Access, Excel,
others?

 

 


  Posted on Monday, June 16, 2008 2:55 PM | Back to top


Comments on this post: BA Symposium - Session Blog on How to Conduct Business Analysis to Ensure a Successful BI Initiative

No comments posted yet.
Your comment:
 (will show your gravatar)


Copyright © TwinCitiesITandDevEvents | Powered by: GeeksWithBlogs.net