Description
‘Tableau Desktop’ as Business Intelligence tool
BUSINESS INTELLIGENCE
Use of ‘Tableau Desktop’ as Business Intelligence tool
Tableau Desktop
Tableau Software is an American computer software company headquartered in Seattle, WA, USA. It produces a family of interactive data visualization products focused on business intelligence. Tableau Software offers three main products: Tableau Desktop, Tableau Server and Tableau Reader. Tableau's products have been incorporated into product suites of multiple independent software vendors, including Oracle for its Oracle Essbase Visual Explorer product. On February 11, 2010, Tableau released a fourth product line named Tableau Public. It is a free-to-use program that offers analytic capability, with the limitations that visualizations are limited to 100K rows of data and can only be saved to the Tableau Public servers. The BI tool which we worked on is Tableau Desktop. Tableau Desktop is based on breakthrough technology that lets you drag & drop to analyze data. You can connect to data in a few clicks, then visualize and create interactive dashboards with a few more. This is the system that supports people‘s natural ability to think visually. User is not stuck in wizards or bogged down writing scripts. One just creates beautiful, rich data visualizations. It‘s so easy to use that any Excel user can learn it. It is faster to use having good user interface and thus takes fewer efforts. Tableau Desktop comes in two levels – Personal Desktop and Professional Desktop Personal Desktop accesses Excel, MS Access or Text Files. Professional Desktop accesses MS SQL Server, MS Analysis Services, Oracle, IBM DB2, Netezza, Hyperion Essbase, Teradata, Vertica, MySQL, PostgreSQL, Firebird, Excel, MS Access or Text Files. Professional Desktop allows user to publish your Dashboards and Worksheets to Tableau Server for distribution via a Web browser
FEATURES
?
?
?
Tableau Desktop Pro is a business intelligence tool that allows user to easily visualise, analyse and share large amounts of data. When data is imported into Tableau it automatically extracts the dimensions and measures, ready for analysis. One useful feature is its support for generated elements, giving extra information about the data that can be used to extract additional insights. It'll immediately generate an initial chart or map (if one is using geographic information), which one can then work with to explore your information. Tableau can be used to produce a wide selection of different chart types, from familiar bar charts to complex linear geographic plots. Tableau Desktop provides a choice of palettes where colours have been chosen to help understand, rather than obfuscate, the
1
information — something that's hard to do using tools like Excel, where design is not so important. Things can be even quicker, as one can just import the data choose a couple of fields, then press 'Show Me' to automatically generate visualisations. Analyst can use the Show Me option to quickly shift between different types of graph, with inappropriate charts greyed out. Tableau aggregates information that would normally be scattered across different tools.
ADVANTAGES OF USING THIS SOFTWARE
? As it imports the data, Tableau also tries to identify and categorise it, for the most part successfully. Tableau doesn‘t limit user to working with a single data set at once. Like an Excel spreadsheet, a single project can encompass a number of different worksheets, each using a different data source. Or, user can use the same source in different sheets, each time highlighting different aspects of the data or the same aspects in different ways. Once data‘s in Tableau, analyst can illustrate it with graphs, diagrams and even maps, in ways that make relationships and trends instantly clear. The program divides the data into dimensions and measures. Roughly speaking, ?dimensions? are the things that are being measured and ?measures? are the actual measurements or figures.
?
?
LIMITATIONS
Tableau Desktop is only part of a suite of tools. If user wants to share results with a larger community of people in a way that lets them interact with the data, user need to use Tableau Server. Tableau Desktop is easier to use than many other desktop analysis tools, but it's still not a tool for the complete neophyte. One can't just throw data at Tableau and expect to get insights — he/she really do need to have an idea of what information one is looking for in the data
APPLICATIONS
1. Banking: With Tableau bank can give its customers the tools to monitor and manage their investments, including the ability to do what-if analyses to drive better decisions. 2. Real Estate: This dashboard lets user see changes in home prices in the context of foreclosures in the area, which is clearly an important influence on the price. 3. Healthcare: In a large hospital, thousands of patients come and go each week. How do you know if you are planning correctly to meet the needs of your population?
2
Healthcare providers and insurers need to know which diseases are the most prevalent in a given population, which ages are most affected, and what the associated costs are. 4. Telecommunications: Keeping a complex network running at full capacity requires a lot of data-- in real time. Finding new customers is a struggle for every business. And deploying new network infrastructure to access completely new markets can be enormously expensive. With Tableau, analyzing networks for expansion opportunities, whether existing or new, is as easy as dragging and dropping.
BUSINESS PROBLEM
Blog, ?www.nanoworm.blogspot.com‘ is a theme based blog containing reviews of already launched and to be launched cell phones. Blog was started in 2008 and it used to get traffic from all over the world with majority visitors from India. In 2009, it reached its peak when on an average there used to be 30-40 visitors daily. In 2010 and subsequently in 2011, the updating of blog slowed down thus resulting in less traffic on the blog. Surprisingly, the traffic in 2010 and 2011 predominantly came from outside India. To regain the traffic on the blog, there is a need to restructure the blog which suits the requirements and taste of the reader. This will require data analysis so that traffic behaviour can be studied on the blog.
DATA SOURCE
The data regarding the traffic on the blog is extracted from third party software, ?Statcounter‘ which was installed on the blog. Following information has been extracted from the software, 1. 2. 3. 4. 5. Day, date and time of visit. Location of the visitor. Browser and operating system used by the visitor. Whether the visit by the visitor is first time or is a repeat visit. Referring link
BI TOOL USED
Tableau Desktop, Professional Edition, Version 6.1
EXPECTED ANALYSIS
a) How browser information help achieve more traffic on blog/website?
3
Ans. If 20% of one‘s visitors use the Chrome browser, one would expect that about 20%
of the orders placed would be from Chrome users. However, if one sees, for instance, that 10% of visitors use Safari, but 0% of orders come from a Safari browser, there may be a bug on the site that prevents users of that browser to finish placing an order. It might be useful to further investigate any browser incompatibilities. Similarly in case of a blog using analytics it can be seen that what features are not compatible with blog features. By working on that, blog experience can be improved. Also if it can be identified that which browsers are used majorly by the visitors, then those components can be added/retained on the blog which does not take much time to load and are easily compatible with the blog. Extending this information to analyze traffic on commercial websites, HTML, CSS, or other programming errors can cause website to not load properly or to not work with older browsers. It is thus necessary to make sure that web page is going to work with all recent Internet browsers and that anyone who may have disabilities is going to be able to read the page. b) How can data of visitor‘s location be used to improve traffic on the blog? Ans. If website/blog is serving content that is readable to other countries, and traffic outside of one‘s country is supporting one‘s site‘s objective (buying, reading, registering, etc.), then geographic data becomes very important. If people in London is reading more content on my site, and supporting great outcomes, then maybe I need to act by writing in UK English style. Also if all of a sudden people from outside country are visiting on the blog, it means recent posts are more relevant and interesting to people outside the home country. c) How information regarding repeat/ new visit can help? Ans. A high number of new visitors suggest that you are successful at driving traffic to your site while a high number of return visitors suggests that the site content is engaging enough to keep visitors coming back. Loyal visitors are frequently highly engaged with your brand and a high number of multiple visits indicate good customer/visitor retention. A high number of new visitors (i.e. those on the left of the histogram) indicate strong visitor recruitment. d) How information regarding the referring link be used to improve blog performance? Ans. Any visit to a website/blog can be traced due to the following reasons, i) ii) iii) iv) Direct link of the blog/website Link of the blog posted on facebook/orkut/tweeter/LinkedIn/email signatures/other social networking sites. Link obtained from blog forums, blogging communities etc. Google/yahoo/other search engines suggests search which might include one‘s blog/website containing relevant content.
4
Of all the above possible sources of accessing the blog/website, the one where the link features on any of the search engines search result is the most important. To get more and more new visitors, blog has to contain relevant content and should be rated high by search engines crawlers. Thus, extracting information about the referring links, strategy can be devised to make blog popular.
PRE-PROCESSING THE DATA
Data extracted from Statcounter is into different files. It was compiled into one excel taking care that the data does not get polluted. Data is then sorted based on date (format of data is first corrected so that it can be used for sorting). Data related to repeat/new visit is then formatted to represent new visit as 1 and repeat visit as 0 so that it can be used as value measure, a requirement in tableau software.
DATA ANALYSIS
A.
Graph shows number of visitors each year along with number of visitors visiting for the first time. For example in 2009, there were total 129 visits of which 74 were first time visits. This shows that new visitors are being attracted to the blog and also there is retention of the already visited visitors.
.
5
This graph shows the number of visitors each year from different locations across the globe. Some of the issues that need to be looked at are that in 2009 there were 72 visitors from Mumbai compared to only 7 in 2010. On the other hand, Visitors from Russia increased from 1 in 2010 to 10 in 2011. It might be due to the reason that the reviews of cell phones posted during this period were high end or concept phones and thus it attracted more traffic from other countries as compared to traffic from Indian cities.
.
6
C.
The graph shows the browser used by visitors visiting the blog. Maximum number of visitors on the blog used Firefox (3.0.3.0 &3.5.3.5) followed by Internet Explorer 6.0. After studying this fact it can be reconfirmed that most of the addons and applications loaded on the blog were compatible with Firefox and some were specially built to be supported in Mozilla. Thus, other users might be having problem viewing content of the blog. This needs to be taken care of.
7
D.
The above graph is a part of the referring links through which people reached the blog. Submission of blog in the blog communities has helped blog achieve good amount of traffic. Also using Stumbleupon, Technorati, Digg and Delicious has helped blog get good amount of traffic. Apart from using these web discovery sites and social blog communities, lot of traffic has been routed through Google and other search engines. For any website/blog to be successful, it should get maximum amount of traffic through search engines. In our case, this is not the case. Though search engines are directing traffic to the site, still it is very less. Certain steps that can be taken to use this option more effectively are: a. Use of RSS feed which till now was ignored. By using RSS feed, more people are easily able to subscribe to the blog and it improves the ranking of the page. b. Blog needs to be hyperlinked to other popular blogs/websites. Search engine spiders use links to navigate the web and index the websites that they find. c. Frequent updating of blog is very necessary which was found missing in later half of 2010 thereby causing lesser and lesser traffic. d. There should be a good network of inbound links on your site. Make sure search engine crawlers or bots has access to every single page from more than one sources. e. Use of keywords intelligently is very necessary for more traffic. Instead of using "Click Here" or "Click This Link", use keywords or title of the page as anchor text of a link.
8
E.
9
Another and better way of displaying same data is shown below,
The table here reflects the behaviour of the traffic. It includes the location of the visitors along with the kind of browsers used by them. Table also includes whether the visitors from the particular location are repeat visitors or first time visitors. Looking at the table, it can be seen that Mumbai, Chennai and Moscow has provided maximum traffic to the blog using Firefox browser.
10
ACTIONABLE BUSINESS STRATEGIES
In the data analysis section, some of the charts analyzed have been shown. Based on the analysis done, following strategies are drawn upon which will help blog regain its traffic, a. First and foremost blog needs to be updated more frequently which is clearly reflected by lack of repeat visits. Visitors not finding new content on the blog are gradually losing interest in visiting the blog. More first time visits clearly show that updating of blog is must. b. The content that is posted also needs to be focussed upon. Only writing content on to be launched phones and concept phones are not attracting too many visitors from India. There are many websites writing reviews on cell phones. To compete with them and to extract traffic, content has to be relevant, crisp and good research will have to be done before posting it as review. c. Most of the features on the blog are configured for Firefox and are not supported in Chrome and other browsers. Also too many widgets cause blog to load slower. Thus restructuring of blog is required which should be supported by maximum number of browsers and unnecessary widgets can be removed. d. To get traffic through search engines, lot of recommendations have been made in the data analysis section, chart D. e. There is also need to include some visuals/audio content which is missing on the blog.
11
doc_560056208.docx
‘Tableau Desktop’ as Business Intelligence tool
BUSINESS INTELLIGENCE
Use of ‘Tableau Desktop’ as Business Intelligence tool
Tableau Desktop
Tableau Software is an American computer software company headquartered in Seattle, WA, USA. It produces a family of interactive data visualization products focused on business intelligence. Tableau Software offers three main products: Tableau Desktop, Tableau Server and Tableau Reader. Tableau's products have been incorporated into product suites of multiple independent software vendors, including Oracle for its Oracle Essbase Visual Explorer product. On February 11, 2010, Tableau released a fourth product line named Tableau Public. It is a free-to-use program that offers analytic capability, with the limitations that visualizations are limited to 100K rows of data and can only be saved to the Tableau Public servers. The BI tool which we worked on is Tableau Desktop. Tableau Desktop is based on breakthrough technology that lets you drag & drop to analyze data. You can connect to data in a few clicks, then visualize and create interactive dashboards with a few more. This is the system that supports people‘s natural ability to think visually. User is not stuck in wizards or bogged down writing scripts. One just creates beautiful, rich data visualizations. It‘s so easy to use that any Excel user can learn it. It is faster to use having good user interface and thus takes fewer efforts. Tableau Desktop comes in two levels – Personal Desktop and Professional Desktop Personal Desktop accesses Excel, MS Access or Text Files. Professional Desktop accesses MS SQL Server, MS Analysis Services, Oracle, IBM DB2, Netezza, Hyperion Essbase, Teradata, Vertica, MySQL, PostgreSQL, Firebird, Excel, MS Access or Text Files. Professional Desktop allows user to publish your Dashboards and Worksheets to Tableau Server for distribution via a Web browser
FEATURES
?
?
?
Tableau Desktop Pro is a business intelligence tool that allows user to easily visualise, analyse and share large amounts of data. When data is imported into Tableau it automatically extracts the dimensions and measures, ready for analysis. One useful feature is its support for generated elements, giving extra information about the data that can be used to extract additional insights. It'll immediately generate an initial chart or map (if one is using geographic information), which one can then work with to explore your information. Tableau can be used to produce a wide selection of different chart types, from familiar bar charts to complex linear geographic plots. Tableau Desktop provides a choice of palettes where colours have been chosen to help understand, rather than obfuscate, the
1
information — something that's hard to do using tools like Excel, where design is not so important. Things can be even quicker, as one can just import the data choose a couple of fields, then press 'Show Me' to automatically generate visualisations. Analyst can use the Show Me option to quickly shift between different types of graph, with inappropriate charts greyed out. Tableau aggregates information that would normally be scattered across different tools.
ADVANTAGES OF USING THIS SOFTWARE
? As it imports the data, Tableau also tries to identify and categorise it, for the most part successfully. Tableau doesn‘t limit user to working with a single data set at once. Like an Excel spreadsheet, a single project can encompass a number of different worksheets, each using a different data source. Or, user can use the same source in different sheets, each time highlighting different aspects of the data or the same aspects in different ways. Once data‘s in Tableau, analyst can illustrate it with graphs, diagrams and even maps, in ways that make relationships and trends instantly clear. The program divides the data into dimensions and measures. Roughly speaking, ?dimensions? are the things that are being measured and ?measures? are the actual measurements or figures.
?
?
LIMITATIONS
Tableau Desktop is only part of a suite of tools. If user wants to share results with a larger community of people in a way that lets them interact with the data, user need to use Tableau Server. Tableau Desktop is easier to use than many other desktop analysis tools, but it's still not a tool for the complete neophyte. One can't just throw data at Tableau and expect to get insights — he/she really do need to have an idea of what information one is looking for in the data
APPLICATIONS
1. Banking: With Tableau bank can give its customers the tools to monitor and manage their investments, including the ability to do what-if analyses to drive better decisions. 2. Real Estate: This dashboard lets user see changes in home prices in the context of foreclosures in the area, which is clearly an important influence on the price. 3. Healthcare: In a large hospital, thousands of patients come and go each week. How do you know if you are planning correctly to meet the needs of your population?
2
Healthcare providers and insurers need to know which diseases are the most prevalent in a given population, which ages are most affected, and what the associated costs are. 4. Telecommunications: Keeping a complex network running at full capacity requires a lot of data-- in real time. Finding new customers is a struggle for every business. And deploying new network infrastructure to access completely new markets can be enormously expensive. With Tableau, analyzing networks for expansion opportunities, whether existing or new, is as easy as dragging and dropping.
BUSINESS PROBLEM
Blog, ?www.nanoworm.blogspot.com‘ is a theme based blog containing reviews of already launched and to be launched cell phones. Blog was started in 2008 and it used to get traffic from all over the world with majority visitors from India. In 2009, it reached its peak when on an average there used to be 30-40 visitors daily. In 2010 and subsequently in 2011, the updating of blog slowed down thus resulting in less traffic on the blog. Surprisingly, the traffic in 2010 and 2011 predominantly came from outside India. To regain the traffic on the blog, there is a need to restructure the blog which suits the requirements and taste of the reader. This will require data analysis so that traffic behaviour can be studied on the blog.
DATA SOURCE
The data regarding the traffic on the blog is extracted from third party software, ?Statcounter‘ which was installed on the blog. Following information has been extracted from the software, 1. 2. 3. 4. 5. Day, date and time of visit. Location of the visitor. Browser and operating system used by the visitor. Whether the visit by the visitor is first time or is a repeat visit. Referring link
BI TOOL USED
Tableau Desktop, Professional Edition, Version 6.1
EXPECTED ANALYSIS
a) How browser information help achieve more traffic on blog/website?
3
Ans. If 20% of one‘s visitors use the Chrome browser, one would expect that about 20%
of the orders placed would be from Chrome users. However, if one sees, for instance, that 10% of visitors use Safari, but 0% of orders come from a Safari browser, there may be a bug on the site that prevents users of that browser to finish placing an order. It might be useful to further investigate any browser incompatibilities. Similarly in case of a blog using analytics it can be seen that what features are not compatible with blog features. By working on that, blog experience can be improved. Also if it can be identified that which browsers are used majorly by the visitors, then those components can be added/retained on the blog which does not take much time to load and are easily compatible with the blog. Extending this information to analyze traffic on commercial websites, HTML, CSS, or other programming errors can cause website to not load properly or to not work with older browsers. It is thus necessary to make sure that web page is going to work with all recent Internet browsers and that anyone who may have disabilities is going to be able to read the page. b) How can data of visitor‘s location be used to improve traffic on the blog? Ans. If website/blog is serving content that is readable to other countries, and traffic outside of one‘s country is supporting one‘s site‘s objective (buying, reading, registering, etc.), then geographic data becomes very important. If people in London is reading more content on my site, and supporting great outcomes, then maybe I need to act by writing in UK English style. Also if all of a sudden people from outside country are visiting on the blog, it means recent posts are more relevant and interesting to people outside the home country. c) How information regarding repeat/ new visit can help? Ans. A high number of new visitors suggest that you are successful at driving traffic to your site while a high number of return visitors suggests that the site content is engaging enough to keep visitors coming back. Loyal visitors are frequently highly engaged with your brand and a high number of multiple visits indicate good customer/visitor retention. A high number of new visitors (i.e. those on the left of the histogram) indicate strong visitor recruitment. d) How information regarding the referring link be used to improve blog performance? Ans. Any visit to a website/blog can be traced due to the following reasons, i) ii) iii) iv) Direct link of the blog/website Link of the blog posted on facebook/orkut/tweeter/LinkedIn/email signatures/other social networking sites. Link obtained from blog forums, blogging communities etc. Google/yahoo/other search engines suggests search which might include one‘s blog/website containing relevant content.
4
Of all the above possible sources of accessing the blog/website, the one where the link features on any of the search engines search result is the most important. To get more and more new visitors, blog has to contain relevant content and should be rated high by search engines crawlers. Thus, extracting information about the referring links, strategy can be devised to make blog popular.
PRE-PROCESSING THE DATA
Data extracted from Statcounter is into different files. It was compiled into one excel taking care that the data does not get polluted. Data is then sorted based on date (format of data is first corrected so that it can be used for sorting). Data related to repeat/new visit is then formatted to represent new visit as 1 and repeat visit as 0 so that it can be used as value measure, a requirement in tableau software.
DATA ANALYSIS
A.
Graph shows number of visitors each year along with number of visitors visiting for the first time. For example in 2009, there were total 129 visits of which 74 were first time visits. This shows that new visitors are being attracted to the blog and also there is retention of the already visited visitors.
.
5
This graph shows the number of visitors each year from different locations across the globe. Some of the issues that need to be looked at are that in 2009 there were 72 visitors from Mumbai compared to only 7 in 2010. On the other hand, Visitors from Russia increased from 1 in 2010 to 10 in 2011. It might be due to the reason that the reviews of cell phones posted during this period were high end or concept phones and thus it attracted more traffic from other countries as compared to traffic from Indian cities.
.
6
C.
The graph shows the browser used by visitors visiting the blog. Maximum number of visitors on the blog used Firefox (3.0.3.0 &3.5.3.5) followed by Internet Explorer 6.0. After studying this fact it can be reconfirmed that most of the addons and applications loaded on the blog were compatible with Firefox and some were specially built to be supported in Mozilla. Thus, other users might be having problem viewing content of the blog. This needs to be taken care of.
7
D.
The above graph is a part of the referring links through which people reached the blog. Submission of blog in the blog communities has helped blog achieve good amount of traffic. Also using Stumbleupon, Technorati, Digg and Delicious has helped blog get good amount of traffic. Apart from using these web discovery sites and social blog communities, lot of traffic has been routed through Google and other search engines. For any website/blog to be successful, it should get maximum amount of traffic through search engines. In our case, this is not the case. Though search engines are directing traffic to the site, still it is very less. Certain steps that can be taken to use this option more effectively are: a. Use of RSS feed which till now was ignored. By using RSS feed, more people are easily able to subscribe to the blog and it improves the ranking of the page. b. Blog needs to be hyperlinked to other popular blogs/websites. Search engine spiders use links to navigate the web and index the websites that they find. c. Frequent updating of blog is very necessary which was found missing in later half of 2010 thereby causing lesser and lesser traffic. d. There should be a good network of inbound links on your site. Make sure search engine crawlers or bots has access to every single page from more than one sources. e. Use of keywords intelligently is very necessary for more traffic. Instead of using "Click Here" or "Click This Link", use keywords or title of the page as anchor text of a link.
8
E.
9
Another and better way of displaying same data is shown below,
The table here reflects the behaviour of the traffic. It includes the location of the visitors along with the kind of browsers used by them. Table also includes whether the visitors from the particular location are repeat visitors or first time visitors. Looking at the table, it can be seen that Mumbai, Chennai and Moscow has provided maximum traffic to the blog using Firefox browser.
10
ACTIONABLE BUSINESS STRATEGIES
In the data analysis section, some of the charts analyzed have been shown. Based on the analysis done, following strategies are drawn upon which will help blog regain its traffic, a. First and foremost blog needs to be updated more frequently which is clearly reflected by lack of repeat visits. Visitors not finding new content on the blog are gradually losing interest in visiting the blog. More first time visits clearly show that updating of blog is must. b. The content that is posted also needs to be focussed upon. Only writing content on to be launched phones and concept phones are not attracting too many visitors from India. There are many websites writing reviews on cell phones. To compete with them and to extract traffic, content has to be relevant, crisp and good research will have to be done before posting it as review. c. Most of the features on the blog are configured for Firefox and are not supported in Chrome and other browsers. Also too many widgets cause blog to load slower. Thus restructuring of blog is required which should be supported by maximum number of browsers and unnecessary widgets can be removed. d. To get traffic through search engines, lot of recommendations have been made in the data analysis section, chart D. e. There is also need to include some visuals/audio content which is missing on the blog.
11
doc_560056208.docx