Popular in Digital Marketing Strategies and Measurement
Popular in Department
verified elite notetaker
verified elite notetaker
Sin Yong Tan
verified elite notetaker
verified elite notetaker
verified elite notetaker
verified elite notetaker
This 15 page Study Guide was uploaded by Frederick Notetaker on Monday May 9, 2016. The Study Guide belongs to MKTG 5721 at University of Missouri - St. Louis taught by Ho Kim, Ph.D. in Spring 2016. Since its upload, it has received 16 views.
Reviews for Final Review
Report this Material
What is Karma?
Karma is the currency of StudySoup.
You can buy or earn more Karma at anytime and redeem it for class notes, study guides, flashcards, and more!
Date Created: 05/09/16
Three types of traffic Direct traffic; Referral traffic; Search traffic Referral traffic: paid vs. unpaid Search traffic: paid vs. unpaid Different websites may have different traffic profiles. Conversion A target action the website owner wishes visitors to take Depending on the types of websites, conversion can be defined in many different ways. Week 1 Case: The Big Four Amazon, Apple, Facebook, Google Battle for dominance Shift in perspective In information age, what matters is scale in data. This shift in perspective—from industry’s physical flow to information flow—is important to all subsequent analysis of internet business strategy. Week 2 Lecture: Web Metrics Definitions of various metrics Impression, pageviews, CTR, CPM, CPC, CPA (cost per acquisition), CPO (cost per order) Visits, Visitors, Bounce, Bounce rate Low CPM but high CPC A characteristic of badly targeted ads An ad may look inexpensive in that it has a relatively low CPM. But if the ad isn’t relevant it will end up with few clicks and so a high CPC. Badly targeted ads can be very expensive as measured by CPC. The quality and placement of an ad affects both CTR and the resulting CPC. Daily/Weekly/Monthly/Absolute unique visitors The time interval (daily, weekly, monthly, …) should be specified when reporting the number of visitors. Week 2 Case: BBVA Compass Online/offline budget allocation Online channel CPA (Cost per application): $81, Offline channel CPA: $20 Annual retention rate: online 55%; offline 65% Does it make sense to spend so much money online? How would you defend? Interaction between online and offline channels Week 3 Lecture: Search Engine Optimization Google’s search algorithm is mainly based on two factors: Relevance: how well the search phrase matches content on the webpage Font size, Position on the page, page ULR, …, page title tag. PageRank: popularity of the webpage (# of backlinks, PR of the referring page). SEO: Any efforts to improve the search engine rankings of a webspage. On-site SEO: changing the content of a website to increase relevance of particular keyword phrases. Off-site SEO: attracting backlinks from other websites to increase popularity. Week 3 Case: Air France A five-step procedure for optimizing SEM campaigns 1) Summarize metrics, 2) optimize publisher portfolio, 3) optimize campaigns, 4) calculate impact, 5) change and measure Understanding proper metric for objective E.g.) if you want to compare how much it costs to generate one click from an ad then, use CPC. Step 2) Optimizing publisher strategy A bubble chart of CPC and conversion rate (prob. of booking). Different strategy for each quadrant Step 3) Optimizing campaigns on publisher A bubble chart of CTR and TCR (transaction completion rate = no. transaction / no. click) Different strategy for each quadrant Week 4 Lecture: Paid Search Advertising CPC (cost-per-click) bidding: auction platform Because the actual CPC is determined not only by bid amount but also other factors, the advertisers are often times charged much less than their bid amount. Ad position is determined by ad rank, which in turn determined by bid amount, quality score, and expected impact of ad extensions and ad formats. Quality score is a measurement of the relevance of your ads, keywords, and landing pages to a person who will see the ad. Keyword match type and implications for keyword strategy Broad match, modified broad match, phrase match, exact match, and negative match Each match type can serve different needs of advertisers. Week 4 Case: Motorcowboy When keywords are extremely targeted (extremely specific), search volume can be very low. Exact match vs. phrase match vs. broad match Determinant of ad CTR: Ad position is an important determinant of ad CTR. Bid amount helps an ad show up on SERPs. But once ad shows up, the position of the ad is barely related to bid amount (at least for Motorcowboy case.) Quality Score is an important factor determining ad position. New advertisers (those who do search ad campaign for the first time) may initially have low Quality Score because they don’t have past results. New advertisers need absolute time. Weeks 5 –6 Lecture: Display Advertising/Landing Page Optimization Players of display ad network Advertiser, publisher, ad network Bidding system CPC bidding, CPM bidding, and CPA bidding Landing page: the first webpage that a visitor to a site sees. Subscription landing pages, long-copy landing pages, single-product landing pages, lead-generation landing pages. A/B testing can be used to find optimal landing pages. Week 6 Case: Thunderbird Thunderbird wants to evaluate the effectiveness of the school’s online media buying activity and the performance of the new website launched in 2012. To do this, they analyze Google Analytics data that contains the first three months of web traffic. From goal to KPIs To evaluate the current performance, we need to find relevant KPIs (variables in Google Analytics data). Then the KPIs should be analyzed from diverse perspectives. For example, one of the goals in the case is to increase traffic to the program page by 25%. Then, the relevant KPIs would be unique visitors, visits, news visits, and pageviews. Another goal is to improve conversion by 30%. The directly related KPIs are bounce rate and conversion rate, indirectly related KPIs are visitors, visits, and pageviews. Google Analytics data are not self-sufficient. For example, market potential cannot be inferred from Google Analytics data. Week 8 Lecture: Linear Regression Analysis Regression analysis is used to find relationship between variables. Marketing activities (e.g., bid amount, ad extensions, ad formats) are independent variables and marketing outcome (e.g., ad position, CTR) are dependent variables. By using multiple independent variable in a regression model, we can separately identify the effect of each independent variable on the dependent variable. The regression analysis results can be used to optimize marketing activities (e.g., finding best ad format to maximize CTR). Week 8 Case: Quality Alloys The objective is to analyze Google Analytics data (specifically the effects of brochure promotion on web traffic) using descriptive statistics and simple visualization tools in Excel. Simple descriptive analysis can provide many important findings about website performance and offline sales. Difficulty in measuring ROI of a B2B site. Google Analytics data is not self-sufficient because it lacks consumers’ intention. Week 9 Lecture: A/B Testing A fraction of users are diverted to different versions of a webpage. Their behaviors are compared against each other. A/B testing allows seemingly subjective questions of design—e.g., color, layout, image, text—to become matters of data-driven science. t-Test is used for A/B testing Real examples of A/B testing 2008 Obama campaign homepage 2012 Romney campaign homepage 2012 Obama campaign email marketing Week 9 Case: Star Digital The objective: to understand how a controlled experiment can identify ‘incremental’ effect of ad campaign. Frequency effect vs. heavy user effect Frequency effect: positive effect of frequency of advertising on purchase likelihood Heavy user effect: the phenomenon that a heavy internet user is more likely to purchase a product than a light internet users, whether he views the focal ad or not. Without control group, frequency effect cannot be separated from heavy user effect. Frequency effect exists if the slope of the test group is greater than the slope of the control group. Week 10 Case: Vungle The purpose of the case is to showcase how a simple A/B testing can be used to determine a more effective ad matching system. Statistical concepts Hypotheses: null hypothesis, alternative hypothesis Statistical significance: whether the statistical results from a sample can be generalized to the population Week 12 – 13 Lecture: Social Media Fives myths about social media 1) Social media is free; 2) Good social media must go viral by itself; 3) Social media will replace traditional media; 4) Good social media must reach millions of people; 5) Social media should function independently of traditional media. Social media vs. display advertising vs. paid-search advertising Social media is best for building brands. Display advertising can be helpful in building awareness of specific products. Paid-search advertising in generating immediate sales and closing the deal. Week 12 Case: Mekanism Mekanism’s unique approach to viral advertising Role of content: ad-in-story format, not story-in-ad format Use of influencers: from a risky earned media to a low risk paid media Platform management: not only social media but also traditional media For what brands, can viral marketing be most effective? Small brands, or strong brands with good stories? Will viral marketing be a mainstream promotional tool, why or why not? Week 13 Case: Ford Fiesta Ford Fiesta Movement’s viral marketing approach is similar to that of Mekanism The both use influencers or seeds/agents Mekanism’s approach to managing influencer network (identify Engage Value exchange Distribute message Optimize) Was FFM successful? Yes / No Marketers should look at how the buzz generated over time. Week 14 Lecture: Online Reputation Management Detrimental effects of negative content Review score and pricing power and sales in many industries Negative online content’s damaging effect is greater than positive content’s beneficial effect. The impact of one-star reviews is greater than the impact of five- star reviews (book reviews in Amazon.com). Negative online buzz has a significant negative effect on abnormal return of stock price. Positive content for the SERPs A way of managing negative online content is by producing a lot of positive content on owned media. GE case Week 14 Case: TripAdvisor Effects of hotel (brand)’s review management activities Listening/reacting: Hotels react to consumers’ suggestions or complaints Incentivizing: Hotels incentivize guests to write positive reviews Selecting: Hotels select certain guests to write positive reviews and withhold negative ones. Only listening/reacting has a positive long-run impact on hotels (brands). Given that selecting and incentivizing have negative long-run impact on brands, how should we understand Mekanism’s approach and Ford Fiesta Movement that utilized influencers or seeds/agents? Effects of reviews on performance vary with the quality hotels (brands) Different styles are recommended for luxury hotels and independent hotels. Week 15 Case: Bank of America Consumer multichannel behavior Substitution effect; augmentation effect; volume effect When new channels are introduced, companies often assume that consumers will substitute one channel for the other. However, consumers’ multichannel behavior is more complex. The net effect of a new channel is not straightforward; it depends on the size of the three effects. Final Exam This is the final exam of this course. Once you start, you should finish the exam in three hours. The Blackboard system will automatically send your answers if you don't finish it on time. As I spoke in the classroom, there are 10 questions. One of them (Q1) is the required question; the rest 9 questions are optional. As such, answer Q1, and choose 4 questions from Q2 - Q10 to answer them. Q1 asks you to evaluate your team members. Each question will be 20 points, so the perfect score of the final exam is 100 points (5 x 20 points). Good luck! QUESTION 1 (This is the required question) Evaluate your teammate for GOMC project. Do NOT grade yourself. Please write each group member’s name and rate the extent to which the person displayed or did the following things. Rate all items according to the following scale. 0 Not at all 1 To a small extent 2 To some extent 3 To a large extent 4 To a very large extent Download the attached answer table ( Q1 Answer Table(1).docx ), fill the table with your answer and submit it by attaching to this question. This is a required question. Your evaluation of team members will be used to determine their final grades. So answer conscientiously and honestly. Attach File Choose FileBrowse My Computer 20 points Save Answer QUESTION 2 Suppose there is a financial services company that does business on both online and offline channels (For example, you can think of Bank of America that provides services at its website as well as local branches). Suppose the online channel’s cost per acquisition (acquisition of one customer) is $100 and that of the offline channel is $20. Also, the annual retention rate of a customer acquired through the online channel is 50% while that of a customer acquired through the offline channel is 70%. In other words, the offline channel is more profitable than the online channel. Further suppose that this company spends about 30% of their total marketing budget on online channel. Given the online channel’s inferior performance, does it make sense to allocate 30% of the budget online? Why or why not? Path: pWords:0 20 points Save Answer QUESTION 3 The ‘big four’—i.e., Amazon, Apple, Facebook, and Google—are expanding their business areas to the extent that each of these firms competes in each other’s core business area. For example, Google entered online retailing business to compete with Amazon. This seems contradictory to the traditional marketing wisdom of segmentation, targeting and positioning. Explain why it is important for these companies to expand their businesses to the extents that their businesses overlap. Path: pWords:0 20 points Save Answer QUESTION 4 In paid-search advertising, why is it possible that the CPC of the first ranked paid- search ad (the paid search ad in the first positon) is smaller than that of the second ranked paid-search ad? Think how ad position in a SERP is determined. There are three factors. Bid amount is only one of them. Path: pWords:0 20 points Save Answer QUESTION 5 List the five myths of social media marketing and demystify them. Path: pWords:0 20 points Save Answer QUESTION 6 The Ford Fiesta Movement (FFM) resembles Mekanism’s viral marketing strategy. Explain how FFM applied Mekanism’s concept of “engineering viral marketing.” You may want to refer to the five-steps of Mekanism’s influencer management (i.e., Identify à Engage à Value exchange à Distribute message à Optimize). Path: pWords:0 20 points Save Answer QUESTION 7 Negative word-of-mouth on the Internet can have detrimental effects on a firm’s performance. Take three examples of such detrimental effects of negative word-of- mouth and explain them. Path: pWords:0 20 points Save Answer QUESTION 8 There are three types of activities that a firm can use to manage online reputation— i.e., listening/reacting, incentivizing, and selecting. Define each of the activities and explain their long-run effects on the brand, customer, and online review platform. Path: pWords:0 20 points Save Answer QUESTION 9 Given that selecting and incentivizing have negative long-run impact on the firm’s performance, how should we understand Mekanism’s strategy of using “influencers” and FFM’s method of utilizing “seeds/agents”? Do Mekanism and FFM use incentivizing and selecting by employing influencers and seeds/agents? Why or why not? Path: pWords:0 20 points Save Answer QUESTION 10 We tend to believe that a newly introduced channel will substitute the incumbent channels. For example, in the Bank of America case, BoA expected that transaction cost would decrease with the addition of less-costly online channel (compared to ATM, IVR, and branch offices.) To the contrary, the total transaction costs increased after the introduction of online channel. This is because channel addition may induce not only substitution effects but also augmentation and volume effects. Define these three effects and how each of these influenced BoA’s total transaction costs.
Are you sure you want to buy this material for
You're already Subscribed!
Looks like you've already subscribed to StudySoup, you won't need to purchase another subscription to get this material. To access this material simply click 'View Full Document'