website_measure - Module II Overview Why SEM Goal Analysis...

Info iconThis preview shows page 1. Sign up to view the full content.

View Full Document Right Arrow Icon
This is the end of the preview. Sign up to access the rest of the document.

Unformatted text preview: Module II Overview Why SEM? Goal Analysis How good is my site? Site Analysis PLANNING: Things to Know BEFORE You Start... How good is my search? Measure SEM performance How to do it? Strategic Planning How to sell it? SEM Proposal Fall 2006 Davison/Lin CSE 197/BIS 197: Search Engine Strategies 6-1 If you can't measure, you can't manage In this section, we discuss: - How to diagnose a web site's well being using multiple performance metrics [notice not just HOW, but HOW WELL!!] Counting visitors Conversion Rate Counting dollars And more... - How to measure web site performance based on your goals - Tools that help your web analytical needs Fall 2006 Davison/Lin CSE 197/BIS 197: Search Engine Strategies 6-2 Metric 1: # of Visits / Visitors Cookie: a piece of text that a Web server can store on a user's hard disk. The pieces of information are stored as name-value pairs - By using cookie, servers store the state of your machine Session: theoretically, one visit paid by the customer to your web site - Technically, hard to accurately measure - Often session = visit in SEM context session-id-time 1159167600l amazon.com/ Fall 2006 Davison/Lin CSE 197/BIS 197: Search Engine Strategies 6-3 Metric 1: # of Visits / Visitors Number of visits: the number of unique sessions as counted by the server Number of unique visitors: the number of unique machine IDs that visited the server Both were actively used to measure site performance around 2000 Fall 2006 Davison/Lin CSE 197/BIS 197: Search Engine Strategies 6-4 Metric 1: # of Visits / Visitors Issues with visits / unique visitor - Issue with session - How unique is "unique"? Financial analysts have become increasingly skeptical of nonfinancial metrics [Gupta et al. 2004] Fall 2006 Davison/Lin CSE 197/BIS 197: Search Engine Strategies 6-5 Metric 2: Stickiness - Related Visit duration / page views Textbook (as well as Brian ) explains in detail how page views are counted - Simple log for each requested file (but file != page) - Heuristics to solve the problem (but distributed environment!) - Single pixel tracking Stickiness = average duration / page view Fall 2006 Davison/Lin CSE 197/BIS 197: Search Engine Strategies 6-6 Metric 2: Stickiness - Related Conventional wisdom suggests stickiness to be a valuable metrics - Reflects high loyalty - Implicates higher likelihood to purchase Page views offer some explanatory power but do not appear affecting firms' net incomes [Trueman et al 2000] Stickiness is capable of explaining the share price of Internet firms [Demers and Lev 2001] Fall 2006 Davison/Lin CSE 197/BIS 197: Search Engine Strategies 6-7 Metric 2: Stickiness - Related Positive relationship between stickiness and purchase [Wu et al. 2005] Positive relationship between stickiness and conversion [Lin et al, 2006] - Duration significant for experience goods - Page views significant for search goods Fall 2006 Davison/Lin CSE 197/BIS 197: Search Engine Strategies 6-8 Solving Measurement Issues Server side solution vs. Client side solution - Companies could invest in technology and build stronger server side monitoring programs (limitations exist) - Or they could buy service from companies such as comScore to get client-side monitoring capability 100,000 213,356,003 342,706 174,990 1,392,713 46,942 601 330 Total number of participating households Total number of website visits Total number of online purchases made Total number of online purchases made in shopping websites Total number of websites visited Total number of websites belonging to shopping category Total number of websites that offer direct sales services Number of websites in shopping category that offer direct sales services Fall 2006 Davison/Lin CSE 197/BIS 197: Search Engine Strategies 6-9 Metric 3: Stochastic Models Purchase Behavior Purchase Rate Assumption Poisson Poisson Poisson Poisson with variation Inter-purchase time: Poisson Inter-purchase time: Gamma Gamma Gamma Gamma Inverse Generalized Gamma Gamma Gamma Models Heterogeneity in purchase rate Death Events Heterogeneity Death Rate Assumption in death rate None Exponential Geometric None * NBD (Gupta and Morrison 1997) None Gamma Beta None None None *** Pareto/NBD (Schmittlein 1987) BG/NBD (Fader et al 1004) Dynamic NBD (Moe and Fader 2004b) ** Joint Model (Boatwright, et al) None None Dynamic Model (Allenby et al 1999) Fall 2006 Davison/Lin CSE 197/BIS 197: Search Engine Strategies 6-10 Metric 3: Stochastic Models Enables individualized prediction Hard to implement Mostly academic research, no industry adoption yet Fall 2006 Davison/Lin CSE 197/BIS 197: Search Engine Strategies 6-11 Metric 4: Conversion Conversion = number of actions / number of visitors - Average conversion is at 5% and decreasing [Moe 2004] Fall 2006 Davison/Lin CSE 197/BIS 197: Search Engine Strategies 6-12 Metric 4: Conversion Conversion = number of actions / number of visitors - Average conversion is at 5% and decreasing [Moe 2004] A relatively well-accepted metric for measuring web site performance Depending on the goal of the web site (remember last chapter?), the meaning of "action" might differ Fall 2006 Davison/Lin CSE 197/BIS 197: Search Engine Strategies 6-13 Metric 4: Conversion Revenue Generation Simplest to calculate - Use e-commerce system to find number of transactions - Use web analysis program to find number of visits - Purchase / visits CSE 197/BIS 197: Search Engine Strategies 6-14 Fall 2006 Davison/Lin Metric 4: Conversion Lead Generation For lead generation with the goal of acquire new customer information - Count each visitor who fills in web contact form as an action - That means connect the form with lead management system - You can also continue to track these visitors and capture their purchase events as well Fall 2006 Davison/Lin CSE 197/BIS 197: Search Engine Strategies 6-15 Metric 4: Conversion Lead Generation For lead generation with the goal of offline sales: - Need innovative methods for the offline channel to identify traffic re-directed from online - Special phone number, special coupon, etc - "Call Me" button - Questionnaires at the offline locations Fall 2006 Davison/Lin CSE 197/BIS 197: Search Engine Strategies 6-16 Metric 4: Conversion Lead Generation For lead generation with the goal of offline sales: - Measuring the lead is only the first step... ? Fall 2006 Davison/Lin CSE 197/BIS 197: Search Engine Strategies 6-17 Metric 4: Conversion Brand Image Abstract and hard to measure - First, define what is a conversion based on the goal of your campaign - Then implement possible mechanism to capture conversion behaviour Fall 2006 Davison/Lin CSE 197/BIS 197: Search Engine Strategies 6-18 Metric 5: Monetary Measurements Cost per Action: the advertising cost you pay for one completed action - Advertising Cost / Total Completed Actions - For example, 1,000 visitors, $ 1 per visitor, 20 end up purchasing, then cost per purchase = ? Value of a Buyer: the average gross profit you earn from a completed action - Average Action Value x Gross Profit as % of Sales - Average Action Value = revenue / action - For example, last month you spent $1,000 on advertising to generate 2,000 visitors and 20 bought at an average of $100 per sale with a gross profit margin of 90. Fall 2006 Davison/Lin CSE 197/BIS 197: Search Engine Strategies 6-19 Web Analysis Tools Tools that help analyzing the web site performance - Basic Level: number of visitors to the site, unique/return visitors traffic referrers search engine referrers search keywords used page views visit paths average number of page views per visitor entry and exit pages CSE 197/BIS 197: Search Engine Strategies 6-20 Fall 2006 Davison/Lin Web Analysis Tools Tools that help analyzing the web site performance - Advanced Level: Conversion stats. Dividing the website into logical categories and monitoring each separately Bounce rates- the percentage of visitors who leave the website within the first x seconds of the visit. Example Fall 2006 Davison/Lin CSE 197/BIS 197: Search Engine Strategies 6-21 ...
View Full Document

This note was uploaded on 08/06/2008 for the course CSE 197 taught by Professor Hectormunoz-avila during the Fall '07 term at Lehigh University .

Ask a homework question - tutors are online