The power of BI


Published on

In today's information age, organizations have loads of data but only a small percentage of that data is being converted into information that results in actionable knowledge for the decision makers. In the attached presentation I have tried to define the process of Data, Information, Knowledge, and Decision/Action. The process is defined based on an online marketing business.

Published in: Technology, Design
  • Be the first to comment

  • Be the first to like this

No Downloads
Total views
On SlideShare
From Embeds
Number of Embeds
Embeds 0
No embeds

No notes for slide

The power of BI

  1. 1. Data  Information  Knowledge  Action (empowering the business decision making) Shahid J. Butt
  2. 2. A web site as a Business Case • The example used here is for a web site (name has been changed) that provides users the ability to search and compare various products/services like loan refinancing, auto insurance, cell phones etc. • The organization owning the website runs various marketing campaigns to attract visitors to the site. Business would like to measure the effectiveness of the campaigns in terms of profitability. • Upon landing on the website page and after submitting an application the customer request is matched with multiple vendors providing the services. • The next slide shows how and where does a campaign start in order to bring a customer to the website, what are different stages that a customer goes through before committing to an action that results in revenue for the business.
  3. 3. Channel: Channel Medium through which web traffic is driven to the website. • Affiliates • Media • No Campaign • Retention Emails • Search Engine Affiliates Media (MSN, Yahoo etc.) Retention Emails Search Engine Impression/Text Link Click Impression/Text Link: • An impression is an image that is shown to a potential customer. It could be a pop up, ad banner etc. • A text link is simply a URL link on a site (for example an affiliate site) which redirects traffic to the website. Click: • Action of clicking on an impression or a text link. This action brings the potential customer to the website. • You may or may not have click for an impression Session: • Once a visitor lands on the site as a result of clicking on an impression or text link a session is started and a web log entry is created in the web logs. Cost Session Inquiry Inquiry: • Application submitted by a customer. An Application contains multiple pages. • The customer information provided through the application gets fed to the matching engine in the back end. Lead: • A completed inquiry is run through the matching engine and a lead is generated if a match is found. The lead results into revenue for the business. Lead $Rev Cost: • The amount of money paid to various publishers and affiliates for advertising and marketing the website • Cost is calculated differently for different cost structures. Example: CPA (Cost per application, CPC (Cost per click), CPT (Cost per thousand impressions)
  4. 4. Where is the data stored? • Every single action shown in the funnel on the previous slide generates variety of data starting from presenting an impression/link on a potential customer’s browser, visiting the targeted web site, and then finally taking an action to either abandoning or committing a revenue generating transaction. • The data is stored at various locations, for example the Impressions and the clicks on those impressions are stored on the Ad-Serving agency’s (like double-click) servers. Publishers and affiliates also have their own data stores tracking different activities. The targeted web site also has it’s own data in the web logs and the databases.
  5. 5. Sample of a web log after a potential customer visits Web Logs • Does this data make any sense? • Can any meaningful information be extracted out of this just by looking at these raw logs? • I’m sure the answer will be resounding no. • These raw logs constitute the Data stage which companies have in abundance in today’s information age. • Having just the data alone does not provide much business benefit. Data - [28/Jun/2012:01:54:03 -0700] "GET /fh/refinance/refinanceWizardOnePage.jsp?sourceid=sid3761-4914 HTTP/1.1" 200 22469 "-" "Mozilla/4.0 (compatible; MSIE 6.0; Windows NT 5.0)" 0 "" "SERVER_COOKIE=; path=/; expires=Thu, 28-Jun-07 08:54:03 GMT"
  6. 6. Extracting the Information Client IP Address Target website Address Date/Time Stamp of the Log - - [28/Jun/2012:01:54:03 -0700] "GET /fh/refinance/refinanceWizardOnePage.jsp?sourceid=sid-3761-4914 HTTP/1.1" 200 22469 "-" "Mozilla/4.0 (compatible; MSIE 6.0; Windows NT 5.0)" 0 "" "SERVER_COOKIE=; path=/; expires=Thu, 28-Jun-07 08:54:03 GMT" The name of the page visited Cookie to uniquely identify a visitor ID used for identifying a campaign. • The logs are starting to make some sense about the data. You can understand the individual data elements but the context is still lacking. Up to this point the Information is extracted from the available data. Data Information
  7. 7. Converting information to Knowledge • Web log is one of the data sources but there are more data sources that are required to complete the picture: Impression/Clicks data from the ad serving agency Cost data from publishers Cost data from affiliates Search engine cost data Other internal data sources • Data comes in all kinds of formats: Excel Files Text Files XML Files • The data is integrated with the web logs for different marketing campaigns that result in web traffic for • The integrated result provides the Knowledge or actionable intelligence around the marketing campaigns. Data Information Knowledge
  8. 8. Examples of knowledge gained after going through the three stages of Data, Information, and Knowledge Data Information Knowledge Month Target % Chg 90 day % Chg MTD MTD % Chg Daily % Chg 90 day Net Contribution % Chg MTD MTD % Chg Daily % Chg 90 day % Chg MTD MTD % Chg 4.3 -2.6 2.5 Daily % Chg 90 day % Chg MTD Net Revenue -3.1 MTD % Chg No Campaign Other Total Inquiries 2.1 -1.3 0.2 -1.3 0.2 3.0 -2.5 4.5 0.5 -1.5 26 1 208 Affliates 144 28 15 6 9 7 32 14 12 1 4 Marketing Source Media MSN yahoo bankrate weatherbug aol Search Engine Google Overture Retention Daily Sessions
  9. 9. Examples continued….. • This kind of knowledge would have not been possible without going through the Data, Information, and Knowledge stages. • Now marketing managers are in a much better position to monitor each campaign from different angles and make fact based decisions in the best interest of the business. Data Information Knowledge
  10. 10. Decision Making/Action • Business intelligence is ineffective if no decisions or actions are taken based on the information provided. • In the case of online marketing campaign, business might take one or more of the following actions: • Re-negotiate costs with publishers • Abandon campaigns resulting in lost revenue • Negotiate higher prices with vendors for products resulting in high quality leads. • Customize/Optimize campaigns for different regions Data Information Knowledge Action
  11. 11. Competing on Analytics • Hopefully the presentation was able to provide a basic understanding of the business intelligence process and it’s power. • Organizations who are quick to adopt and mature in the process certainly gain a competitive edge. • More information can be found about the BI maturity models and frameworks at the following links: Data Information Knowledge Action