Sunday, July 17, 2011
Web Analytics
Clickmap
Page view
Unique visitor
Visitor loyalty vs. visitor recency
Length of visit
Avg time on site
Bounce rate
Organic vs paid search
Goad Conversion rate:
A conversion occurs when a visitor reaches a goal. Goal Conversion Rate is the percentage of visits which resulted in a conversion to at least one of your goals.
There are three kinds of goals: URL Destination goals, Time on Site goals, and Pages per Visit goals.
Conversion rate, in percentage, equals Outcomes divided by Unique Visitors during a particular time period.
Packet Sniffer
To overcome the difficulty of tracking dynamic pages (pages with Ajax, Video, Flash) we can use event tracking
Voice of the People
A/B and multivariate test.
Objective is to change the behavior of customer on web site by making some changes to a specific page. Results are compared vs. a control page which is un-altered.
A/B test: one variable change (but can be different version to be compared like different size of a button (size is the variable)
Multivariate: change more than one variable at the time
Example of elements that are tested in A/B tests:
PPC = pay per cl
CTR = Click Through Rate
Page view
Unique visitor
Visitor loyalty vs. visitor recency
Length of visit
Avg time on site
Bounce rate
Organic vs paid search
Goad Conversion rate:
A conversion occurs when a visitor reaches a goal. Goal Conversion Rate is the percentage of visits which resulted in a conversion to at least one of your goals.
There are three kinds of goals: URL Destination goals, Time on Site goals, and Pages per Visit goals.
Conversion rate, in percentage, equals Outcomes divided by Unique Visitors during a particular time period.
Packet Sniffer
To overcome the difficulty of tracking dynamic pages (pages with Ajax, Video, Flash) we can use event tracking
Voice of the People
A/B and multivariate test.
Objective is to change the behavior of customer on web site by making some changes to a specific page. Results are compared vs. a control page which is un-altered.
A/B test: one variable change (but can be different version to be compared like different size of a button (size is the variable)
Multivariate: change more than one variable at the time
Example of elements that are tested in A/B tests:
- Call to action wording, size, color and placement
- Headline of product description
- Form length and types of fields
- Layout and style of website
- Product pricing and promotional offers
- Images on landing and product pages
- Amount of text on the page (short vs. long)
Ether for A/B or multivariate the goal needs to be defined in advanced: what is the desirable action of the customer that the change is trying to induce.
To be declared as a success we also need to make sure that the result is statistically significant it didn't occurred by accident but rather reflect a repeatable pattern - http://20bits.com/article/statistical-analysis-and-ab-testing
PPC = pay per cl
CTR = Click Through Rate
Hadoop
Hadapt
The company’s product approach is a full integration of the open source parallel data processing framework, Apache Hadoop™. Hadapt is adapting and expanding the Hadoop architecture to bring a more complete SQL interface, a patent-pending Adaptive Query Execution™ capability, and a hybrid storage engine to handle structured as well as unstructured data in a single platform.
Datasheet here
White paper here
The company’s product approach is a full integration of the open source parallel data processing framework, Apache Hadoop™. Hadapt is adapting and expanding the Hadoop architecture to bring a more complete SQL interface, a patent-pending Adaptive Query Execution™ capability, and a hybrid storage engine to handle structured as well as unstructured data in a single platform.
Datasheet here
White paper here
Thursday, May 5, 2011
Agile BI
Notes on Forrester 2011 Trends and beyond
90s was the time of silos development, 2000s the decade of centralization and 2010s the decade of agile
Centralization of BI address issues related to "one single vision of the truth" and provide uniform tools across the enterprise. The problem is that with consolidation comes additional processes, administration and middle people to go through to reach knowledgeable player/developers
Thus the idea of agility on top of centralization. The idea is to be flexible enough to address unique problematic of evolving business processes which induce small changes requests. These type of requests don't go well with traditional SDLC (software development lifecycle). BPM suite seem to be well suited to agile development. Business must have ownership of the BI initiative. Create a hub and spoke environment i.e. what needs to be centralize and what needs to belong to satellites.
Forester define 4 major sub-category to agility:
90s was the time of silos development, 2000s the decade of centralization and 2010s the decade of agile
Centralization of BI address issues related to "one single vision of the truth" and provide uniform tools across the enterprise. The problem is that with consolidation comes additional processes, administration and middle people to go through to reach knowledgeable player/developers
Thus the idea of agility on top of centralization. The idea is to be flexible enough to address unique problematic of evolving business processes which induce small changes requests. These type of requests don't go well with traditional SDLC (software development lifecycle). BPM suite seem to be well suited to agile development. Business must have ownership of the BI initiative. Create a hub and spoke environment i.e. what needs to be centralize and what needs to belong to satellites.
Forester define 4 major sub-category to agility:
- Automation: Free up human resource and process time by automating what can. Will leave more time for custom/last minute/ad'hoc request. Example: Auto information discovery - data profiling -> speedup ETL/Schema built; All stages of BI (ETL, data management, reporting) in one tool can enable life-cycle automation; optimize usage/speed/availability for most used elements/entities/tables/reports; See and share what other do; Auto decision
- Unify toolset: have a common choice across company for BI tools. Example of unification: SaaS logical linking to data; structure, unstructured, predictive, complex data structure like unbalanced, ragged or sparse hierarchy
- Pervasive: Spread widely information; whenever, however, whatever
- Limitless: adaptive DM, multi-dim
TIBCO Spotfire
Endica
Quiterian
Attivio
Subscribe to:
Posts (Atom)