|
|
Google Analytics is a very useful tool for Internet business especially online marketing. Google Analytics is probably the only free enterprise class tool available for website owners. It’s amazing the kind of features available for users at no cost. The amount of analysis, insights that can be accomplished is enormous considering the tool is available for free. The auxiliary benefits of Google analytics tool such Benchmarking & Google Trends is equally big.
Most website users have already used Google Analytics as online visitor reporting tool, however new features like Custom reports / Advance segmentation has upgraded it to an analysis tool. With Google Analytics you can achieve detailed analysis to make well informed decisions for online marketing & Internet business in general.
Some of the new features in Google Analytics are:
Advanced Segmentation: This is a most important feature to segment GA data as per your criteria.
a. Define segments based on any Site metrics, Traffic, Content, E-Commerce or Systems parameters available.
b. Apply these advanced segments on any report to view slice of data as per your criteria.
c. Apply more than one segment, along with overall visits for comparison.
d. Use predefined segments available by default.
e. Understand behavior of users for each segment and differentiate your action for each of them.
f. This feature along with regular expression can identify key pain points & opportunities that can make huge difference to your website.
g. E.g Segment1: Visitors landing on home page and bouncing from site.
h. E.g Segment2: Visitors landing on inner pages.
Custom Reports: This is a user defined report, where you can choose to identify elements in rows & columns in a data table.
a. Choose desired dimension parameters in rows and your desired metrics in columns.
b. Select other sub-dimensions for drilling down the reports. You can select drill down parameter to three levels.
c. Sub dimension chosen will be a slice of data from top level dimension. Since you have the ability to choose any dimensions for both levels, you can get absolutely wonderful segments.
d. E.g : First level dimension can be Geography, drill down to Top Landing pages & Network location.
e. E.g: First level can be Count of Visits, drill down to Traffic Sources, Top landing pages.
f. These reports can be saved, exported, sent to email and viewed anytime online.
Motion charts: Great Visualizing techniques like motion charts are available for users to pictorially view metrics over a time period. You can choose other site metrics as bubble sizes with different colors.
a. You need give Google motion chart API code access to your GA account.
Data Export API: Use Google Analytics API to import data from Google Servers. This is very useful as you can store GA information in your database or use it third party reporting software.
a. Export Data on Dimensions & metrics which can be used in combination for a custom report.
b. Common calculation such as formula for Bounce rate (bounces/entrances), Exit rate (exits/pageviews) and more can be found here http://code.google.com/apis/analytics/docs/gdata/gdataReferenceCommonCalculations.html
Flash & Dynamic Page tracking: Use ga.js new code for tracking dynamic pages with _trackpageview function.
a. Track Videos & various actions like pause, play, stop etc.
b. View Developer Docs present at http://code.google.com/apis/analytics/docs/
User Defined functions: You can configure Google Analytics to store user attributes like Age, Income, demographic, psychographic attributes. The website needs to capture these attributes in first place and using the setvar function in javascript code, you can pass these values to Google Servers. Setvar function overwrites new values for same cookie, to avoid this there is another function called supersetvar, which can be used to append future values to existing ones.
a. You could slice,dice, segment data on Age, Income & custom defined values.
b. View the list of functions which can be used to customized data representation in GA. http://code.google.com/apis/analytics/docs/gaJS/gaJSApi.html
Internal Site Search: Configuring your GA for site search can be done by enabling in main settings panel. You also have to give the parameter which stores the search keywords in URLs.
a. Search Exits, Bounce Rate, User behavior during search, post search and pre-search.
Ecommerce Tracking: You can use GA to store your e-commerce transactions. This is done by enabling e-commerce settings. You also need to place additional code in Thank You! Page to capture the elements of transaction such as price, product name, quantity etc.
a. You can find more info here about installing the Ecommerce tracking. You can even track 3rd party shopping transaction. (can be done by _linkByPost function) http://www.google.com/support/googleanalytics/bin/answer.py?hl=en&answer=55528 http://www.google.com/support/analytics/bin/topic.py?hl=en&topic=11001
Adwords Integration: Integrating Adwords keyword, impression, cost, cpc, clicks metrics on GA will be extremely useful as you can segment, slice, dice cost, cpc, impression via Geography, content, landing pages, visitor behavior, loyalty etc.
a. Adwords optimization is done based on conversion, however using GA we would be able to identify various segments of users such as Visitors doing research, Visitors browsing FAQs, Visitors commenting, browsing etc. Therefore we would be able to attach value to each segment and hence can better optimize Adwords campaigns based on different actions done by users rather than just conversion.
Coming articles will contain more information on GA regular expressions, using Goals, filters and advance implementation on pageviews etc.
This Article is originally available only in http://www.praveenkodur.com/blog/
This is the most widely used area of data mining techniques, in sales and marketing to understand customer and their behavior on various products offered by the company. It helps companies in understanding and predicting customer behavior for each specific situation and therefore introduce sophistication in targeting. There are lot of information generally collected about the customer such as demographic, geographic, lifestyle, attitudinal, behavior and many more. These data can be efficiently used to model customer behavior under different circumstances (situations). Through effective models we can improve ROI of marketing efforts & campaigns to make an impact in the overall profitability of the business.
Some of the common uses of analytical models are:
Acquiring new customers: Acquiring new customers is generally a big cost for the company, with increase in prices on various acquisition channels, companies find it hard to reduce cost in acquiring good profitable customers. Predictive modeling to certain extent becomes very useful in reducing costs in acquiring right customers and increase profitability. It also helps in designing marketing offers, special campaigns for customers to reach out to them.
Predictive models use historical data of customer attributes to understand relationships between attributes and their specific response or behavior. The output of predictive model is generally to predict a future response of the customer with their present data. These models are generally used to rank a list of prospective customers on the likelihood of their predicted response. This is very useful as it helps take present decisions. Other factors can also be fit into the response such as risk of acquiring customers like credit risk, or the cost of retaining customer. We can also predict response to each specific product, which helps in better targeting the customer and increases chances of acquiring them or even for cross selling products
(http://www.stochasticsolutions.com/pdf/CrossSell.pdf).
Additionally, there might be situations where it might not be easy to connect both customer and the response desired. Additionally there may be situations where various responses of customers are also valuable. For example, a customer might purchase a product after visiting the site 4 times, each time the customer performs a action like search, sending an enquiry, contacting user, reading knowledge material etc. Each action may be valuable to business therefore need to included in the model. In all such cases we need to use proxies to understand purchase behavior. Model building is a tedious task, but very worthy effort in increasing profitability.
In process there are additional outputs generated like customer profiling, customer segmentation, clustering and affinity pairs which is very valuable in developing products for customers.
http://en.wikipedia.org/wiki/Predictive_analytics
Customers Retention: The second problem is of retaining existing customers, companies are willing to provide offers to retain customers. The data mining problem will translate into finding the customers at risk (or customers who are looking to switch) and additionally identify those customers who are more likely to change behaviour due the marketing offer given by the company. (http://www.stochasticsolutions.com/pdf/FinanceRetention.pdf).
The data useful will be behavior of customers before attrition for certain time duration along with their demographics, attitudinal & behavioral patterns.
Predicting customer attrition rate is separate from predicting their behavioral change because of promotional offer. These models are extremely useful where markets are saturated and acquiring new customers becomes increasingly difficult.
Marketing techniques for retaining customer could backfire sometimes, resulting in loss of customer due to persuation from the company. There are also certain customers who would attrite irrespective of any marketing offered to them. Capturing this behavioral difference can be done through Control Groups, Test Groups & Hold Groups. This method is called as Differential Response modeling or Incremental impact model or Uplift model or Net model. Here is more info on the same: http://en.wikipedia.org/wiki/Uplift_modelling, and few white papers on the same http://www.stochasticsolutions.com/pdf/SavedAndDrivenAway.pdf. Customers in Control Group are randomly targeted, Customers in Test Groups are targeted based on model, Hold Group Customers are not considered for targeting of offer. Results of test and experiment is used in building the Net Model.
The industries where these techniques will be highly useful are Financial Services, Retail, Telecom, Internet Companies and Software Houses.
Do check back for more articles on this topic.
List of resources to find more information about Analytics
1) http://www.destinationcrm.com/Articles/CRM-News/Daily-News/Predictive-Analytics-Can-Pinpoint-Profitable-Customers-52164.aspx
2) http://scientificmarketer.com/search/label/response
3) http://www.redclaymedia.com/response_modeling.php
4) http://stochasticsolutions.com/retention.html
5) http://www.information-management.com/specialreports/2008_62/10000747-1.html?ET=dmreview:e323:1015879a:&st=email
6) http://www.information-management.com/issues/2007_52/10001990-1.html
7) http://www.predictiveanalyticsinsight.com/articles/callcenter.htm
 http://www.predictiveanalyticsworld.com/predictive_analytics.php
9) http://www.marketingprofs.com/4/shearer1.asp
10) http://semphonic.blogs.com/semangel/2009/01/predictive-analytics-getting-a-legup-on-where-analytics-is-headed-.html
The Original article is present on http://www.praveenkodur.com/blog
Data analytics involves a lot of transformations and therefore requires a careful attention to detail. The data generally contains many inconsistencies; the most common discrepancy is issue of Missing Values. Even a modest amount of missing values scattered throughout the data set will cause significant reduction in sample set. There are various methods by which you can handle missing values in the data. This process is known as imputation.
1) When the dependent variable contains missing values, simply eliminate the records.
2) Correctly Identify slices of data and Substitute with measure of central tendency like Median, Mean & Mode. Identifying the right slice is also important. You can group by various parameters and take a central tendency. Choose the one with highest bias (chi-square)
3) If the missing value forms a Normal distribution pattern, find the missing value by normal inverse function.
4) Treating the missing values as a dependent variable in a regression equation. Use the multiple linear regression function to impute the missing variable. You can try other methods instead of regression like classification, decision tree etc.
5) Use business logic to understand the missing values.
6) Check the data capturing process, there could a error present at source of data entry. Also it helps identify if the missing data points are at random or non-random. If it is random missing error then you can use simple imputations, however if it is non-random then you need advanced techniques to impute values. Also look at bias in the particular column, if the bias is significant then you need advanced techniques. If bias is minimal then you can proceed with simple imputation.
7) Identify the list of possible values for the missing data set. Try and replace each possible value and create different data sets and build the model. Calculate differences in accuracies and consistency based on different substitutes. This way you can even add variation of the values into the missing element and remove bias.
Use regression to determine the distribution of the values in place of missing values. Create a What-If scenario by imputing every range of value.
9) Do nothing remove missing values and duplicate records of sample data set to increase the size of the data set.
10) Measure similarlity of records like vectors. The similarity is the cosine function between records, and find similar records to the missing data values.
11) Use logistic regression to measure likelihood of observed or likelihood of missing. If value missing the output is 0, else 1. The rest of the variables (non-missing) act as independent variables. This does not predict anything but only a likelihood of finding the variable missing. Records with same probability or closest probability is considered similar and missing data is donated.
Multiple imputation generally yields better results but it requires high-end statistical software for computation. It becomes necessary to use the help of statistical software.
This article is originally found on Praveen Kodur http://www.praveenkodur.com/blog/.
Financial Institutions offer various products to its customers such as personal loans, home loans, savings account, credit cards and so on..
It is very essential for a financial company to know, which user to target for what product and when. The goal of data analytics and data mining department is to arrive at accurate answers to these 3 questions, (Which User, What Product, When), trust me it is a rewarding exercise for the bank, users and definitely the analytics manager.
The problem is very simple. A customer may have purchased a single product or multiple products from a bank. The company would ideally like if all its customers bought all of its products. This possibility however does not exist, therefore without any targeting model; the company would be wasting money, time and resources by randomly targeting customers for its products.
Using a little sophistication of data mining techniques, we would be able to build a targeting model or a response model on customer base. A response model is a predictive model which estimates the future behavior of a given customer even before it occurs. Even with a fair amount of accuracy it can be very useful for company in saving majority of costs in various departments/channels. There are various types of data available for analysis like demographic, behavioral, psychographic / attitudinal. The above data can be very useful in data mining methods, however the cost, usefulness and expertise to store / retrieve data has to be assessed and judged.
Giving an example: A customer has savings account in a bank, based on the information given to the bank along with account status and history, we could target her as a prospect for a personal loan product based on a logic / algorithm.
The logic / algorithm could be as simple as if a customer from metro with account balance of Rs 1,00,000 or more for over 6 months should be a prospect for home loans Rs 20,00,000 or personal loans over Rs 3,00,000. This is a simple example; however it could be a complicated neural network or logistic regression model.
Cross Sell Models: A customer of a product A, is a likely prospect for product B if his account activity positively predicts likelihood of purchase. Target model has to predict purchase of product B.
Up Sell Models: A customer can be targeted for higher segment product in the same category based on his account status and demographics. Normal saving accounts customers can be sold for Premium segments customers, if his account balance is over a period of time.
Deep Sell Models: A customer can be sold more of same product based on his need. Target model has to predict his need.
These target models are also called response models that only predict the purchase behavior of a customer for a given product. This model is ineffective as the cost of randomly targeting a customer for product is LESS THAN cost of targeting using predictive response model.
The main reason is because of below limitations of traditional response model
A) All customers behave in two ways: Purchase or DO NOT Purchase
B) Behavior of customers during absence of any marketing activity.
C) Maximum lifetime value to be derived from customer is unkown, this will result in underselling or selling lower value products when higher products could have been sold.
IF these issues are taken care of by strong statistical modeling then loss making response model can be made profitable.
The original article was found on http://www.praveenkodur.com/, Copy rights are reserved.
Disclaimer: These are my personal views and not a view of any organization or body.
Internet Banking is a different arena of web analytics, which require a totally different perspective and problem solving capabilities. The rest of the article will talk about some problems and solutions in such areas.
In traditional ecommerce sites, KPIs and key metrics are tracked by setting up page-tracking and Goals. Users are tracked through funnel process, like movement from one page to another page until the sales confirmation page. Usability and productivity of the page is evaluated by bounce rate, exit rate, conversion rate and click stream behavior on each of these pages.
However, consider a case of Internet Banking, it is very interesting that such funnel navigation or bounce rate gives a very little information about its effectiveness, here is why? Lets say a user has a bank account and is using Internet Banking for transaction.
There are various things are tracked to understand his behavior, a person who wants to transfer funds from one account to another will do it even if it is a difficult process and does not fully understand, a user will try and figure out the ways to understand process and transfer funds online. Therefore the drop out rate in funnel navigation is very less, unlike ecommerce sites. Most users will inevitable complete the funnel and navigate all important pages to reach thank you page, even if they do face difficulties. They would invariably figure out how to do a transaction online.
The problem for web analytics guy is, a lot of users seem to be using internet banking very well but how do you find out where exactly is the problem, what needs to be mended.
Here are some areas of data to be looked into, that gives us insights in this direction.
1) Pageviews of Error pages.
a. How many times error pages were thrown up?
b. Source of error pages, which page led to error pages most and their ranking.
c. Errors per transaction.
2) Pageviews of help section and FAQs, which part of help section was most viewed and their ranking.
3) Total Pageviews per page per transaction. Example, a fund transfer has 4 pages, how many times each of the pages were viewed per transaction.
4) Offline feedback on errors and complaints on Internet banking transaction.
5) Average time taken per page online
a. Rank pages on average time per transaction
6) Number of fraudulent transactions or security complaints.
These are my personal views and opinion on this subject. Please do let me know your views as comments.
This article is originally found on http://www.praveenkodur.com/ Copy right reserved on http://www.praveenkodur.com/
|
|
