Firm Generated Content & Viral Marketing Strategies PDF

License Results IMIs Virality Timeliness Ad Estimation Dramatic Study Flow Field Study Viral YouTubeCharacteristics Regression Main Takeaways Why Learning Where Today's Do The Improvised Marketing Motivation Summary Topic Lectures We Evaluations Exam What What's Email For G...

License Results IMIs Virality Timeliness Ad Estimation Dramatic Study Flow Field Study Viral YouTubeCharacteristics Regression Main Takeaways Why Learning Where Today's Do The Improvised Marketing Motivation Summary Topic Lectures We Evaluations Exam What What's Email For Going and Firm read wanted used the 2: 3: Lack of ofII& III IV Findings Takeaways Experiment "so Limited Marketing Drives can Results Mechanisms Virality Observation Observation: 1: 32: the Structure Coverage Policy Bigger Are Elements Oreo Citation Setup Results of Available Exam Topic Tweets much" Results & papers! Unanticipation Equation Paper: Advertising R! Goals YOU inTweet Academic Implications Weacademic Marketing Online Before Solution Coverage Now? bring for and Regression toofLab Findings data/stats? Matter?this Viral You? Sharing? the toData Emotions Studies Knowledge marketingAds Availability? Interventions the Week Exam Exam? Matters knowledge! Context: Study Suggested Same Can Weibo Almost There Link PART Course My IBe hope2:here: videomTurk Improvised Virality: All Drivers Strategic Why? Condense Superbowl explicitCitation: Information Effects: Multiple Data:Trade Emotional Unruly.co as off experience: Discuss IMI Company NO A: rm was the can≈ Survey studies Study calculators Lecture of always: 15 achieving generated ads Chinese Question:blackout, decreases emotional ads 2, but increase2013 shares "Typical" ads aChoice important → content Does What Can method, large (these shown curve Nov sharing vs consumer brand Interventions a acommon tweeting integralness characteristics connect tweets number within Questions 25, rm are learning that on one to2013 →create ads) social behaviour https://app.evalytics.nl/#/login?evaluationId=158180 sharing papers Multiple aclass help why Marketing learning between tweets NO Slides answers one company business has answers into Twitter understand good may type 48 structure: with of increase and can the hours (IMI): views -shareable media increase analytics of 30% an March generate answers of analysisstyle audience composition of in an demand to increase the a4,ad help demand short is ads 2014paper exams enhance demand, critical ads weary that for time product and get their bene for period of its in shared execution uential t due of to Informative demand? products? shareability? brands? marketing aconsumers summarize traditional retweets real-time @misc{deerdsms2023, strategy evenads increase marketing sharing advertising results more content from brand communication messages evaluation an online. experiment and purchase proximal that tests to the intention an external Study But Tweeting: Prominent at Erasmus 0know not Examine for viral U a positive new sharing ad & products tracking signi of online cant platform effect video ads Industry: Let. Why? event Approx This What Experiment We High Look Rather One Write, us Nothing wanted might've sounds Exception: Ineffective Emotional level at is documentary even than 1,962 theyour title={"Digital effectiveness stuff, the learning question? when (Firm to understanding a handled video thoughts! little calculate Marketing (brand) Emergencies latest less and controlled you ads too TV nitty goals! that info, don't Why shows causal that gritty tweets oncult better Strategies Integral Social of company adscommon is Mediamaterial know OR is it relevant? dif Quasi/Natural) combinethan the causal Strategies: answer, weFirm both demand didmethodologies...write effects Generated for your thoughts Content products And Tweeting Suggests less +are & Viral for mechanisms In high uential price Marketing"}, products retweet: driving main retweet effects gives acare signi cant boostcaveats Speci i.e. ads Gong, Humor... Marketing How Linear And Clearly Not De asc Mechanism: that only expect Thiswill Questions: can What Zhang, Regression express and is ne 2003 ain onebooks lot thedigital Improviseddo to Timeliness Zhao measure of next marketers yourself attracting their often "valuable / marketing and year Difference implications this? throw important Jiang via and a virality" Marketing assumptions, new (2017) causality managers in the audience Differences authors Interventions out care interpretations, of about the (IMI) window about and virality? assess Time:. Do Info PART Lab Leads June certain author={Lachlan Misalignment One We NOAssignments they B: towant Dictionaries 7 show low get Short totypes via gradeswith broadcast know Answerof what Youtube's ads Deer}, generate Consumer questions per you API: day Needs think more across - 70% virality seven in local"real" data? channels Approx Strategic And What Just where Positive the then They did50% Question: needed emotions argue year={2023}, of beginning we marketers professionals nd? Are lead this of is IMI's to a by: in larger wonder more the effective set shares eld why ofat have tools they not generating don't studied see online largemarketing sharing? effects What's Main aspects results: the of an mechanism? IMI that are (subtle) important for its success Firm Generated Content & Viral Marketing What The But If you Study What Keeping Ability Does Viral: Improvised Particularly you We Outcome: authors helps enjoyed url 1: have =highly information is"https: can't (a) tothe ads awant reach aclose picking grade Experiment Number creating Marketing causal valuable if shared, goyou usviral? vast eye to focussed have effect up and on audiences believe: skill the Interventions github.com/deer of what positive toshares test we ad trends constructive of skills can't content inferences in type 6 company and in along marketing see months on a online about short limit shares are ways after tweets theheavily sharing? period way: chatter to lab/dsms persuasion, and launch improve on brand shared demand? lectureof time the fgc when outcomes and atclass low they virality" When Improvised. Speci Even My Poor mantra: Context In (b) Tweeting is }. Especially Measuring High Doifuential cost Does Valuable: are cretweets the timely What Dedocuments, Shares Each NO 6 Questions: brand viral "latest is What Understanding nequestion Notesthe There'sarousal Ifimpact you king and Parting prominence by increasing tweetscan't info" involveCourse content Preliminaries helps Marketing as Drives Advertising term on outside a large Recap emotions...inbrand the KPIs uential withwas measure more of humor 5-10ayears is"Viral Facebook, approxworth ofThoughts of interest decrease users you broadcast Wrap Marketing ads it,and knowledge valuable not Virality? Marketing Effects explain you help? time go Marketing" Twitter, multiple 100 knowledge more sharing? aren't old can't info Interventions Up 10questions! points, Google...your viral? choice gap anTool Metrics each effective anticipated build for in +,the situation attract new than rms? questions LinkedIn question eld effective low over the easier viewers roughly arousal rst 30 days DigitalIdentify Moldovan, Less That corresponds Pursue Covariates: Study strategy Use. Timing 2: Ethical Which Increases Do we your and Attribution Models positive understand around time features...Enriches information employers and an being slows brand Concerns Social analytics Steinhart to your Study wisely emotions it oneof1 an able care down toMedia visibility (and ad career, lecture & to about, Lehmann progress... thoughts increase in that formulate and there's and/or proportion Strategies consumers in(2019) alots recognition sharing? witty rmsto of the more likely post/response employer → weight brand of share it demand equity a question) Info Mock ALL Informative (... of answers about Exam butAds:I can't brand aretweets sample in see the (broadcast appearence aofcompeting slides360 and +of time) length papers ads story also of adbeing attract collected, thoroughly matters new followers covers 79 tested) brands to How?: Why Explain Do More Features Which Helps Split... relevant? even Large IMI's Media the Mix tension Modelsbetween "viral-" and "valuable" marketing company i.e.into positive you was among I'mresult scale of multiple ifrarely ads happy page you emotions in experiment that rms move greater the to parts who are case! read into associated to virality? don't on help Master's longer, Weibo know you or lesswith study answer it w/yet aincreased media positive whatcomments company we're virality: looking that for When ALab Lachlan How?: It's Warning: Rule What is (Increase IMore Ideally, a don't Reduced critical viral Practitioners. Informative out Experiment: Summarize Deer marketing Online eye: Need careful type see with content alternative Field in vs video of it sales?) causal class Innovation They to often Emotional Experiment Undergrads the main summarised brand IMI are adsbevaluable yields cutting careful reluctant explanations @on Basics features measurement placement on greater YouTube the afor in data 7that Wharton topoint booksof rms? don't engage scale "valuably virality? many ways offend with marketing forviral" only 346 academia adsobservations! Course Same Mainly Themes: Gotprovide codersstructure, context the slides (i.e. paid differentsomequestionspeople) to watch ads and write down IfTilburg you Note: Rush Test found Feature Integral-ness Do for (again) external University Appreciating IMIs highly it enhancemore positive This shareable validity of Incrementalitythe painful brand paper howrm emotions Sentence to is than ad very narrative Experiments value? helps you content fun, LIMITS cool but in apply...persisted: Advertising contextualize results This (Paper Emotional Akpinar whats Revise course Example: also Randomly in by adheres Reasons: & ads itlooking Berger discusses assigned are toBurger the shared (2017) for King the principles complexity, predictive an ad, more on answer of power then the perceived than International Open of asked + informative discussing Science relevance, model) ads Women's with Community peers Day of abstract Tilburg natureUniversity... Diminishing Updated: are Lurking Remark: At That important Lab.initiative Note: the Are 2024-10-09 uses expense Measuring Experiment: User Think ofless Consumer differences: Paper antoinformative analytic promote is of Advertising thisUndergrads Generated asit's a Content descriptive Experience descriptive advertising Seeded toolkit online @&MEF virality Effects Socialto rather effectiveness sharing regression answer vs Mediathannon-seeded a causal question virality This Otherwise advocates for My memorizing Goal: transparency 50%and stylized response accessibility facts rate Study Speci When3:. you c questions? Investigate Willingness work to with Email mechanisms share an analytics us... behind use team theeffect the oninaresearch COURSE projectand email you teaching can to all address How?: levels But AThere be. Informative User mix of society open of Generated Onlineare and experiment, ads minded exceptions areContent Reviews thus with shared creating morequasi-experiment here & what Social more -when 'marketing' new accountability products, Mediabrand and is expensive observational integral and impact. products to narrative data I hope this appreciate "Content Not. Brand They'll class also our gave how personal aspects Evaluation use it for you the insights sausage that onesincrease prediction onwasto these made aspects sharingsample an (unseen) seem to decrease CarefulOnline placement Reviews Word of Mouth & In uencers of brand name "Are We'll Brand these (Why?) It's look. Purchase notfeatures at doesn't easy, integral-ness a subsetpredictive Likelihoodit get is time not of lost ofinshares?" their effectiveness" importantconsuming daily results, email in adds focusing volume that on the emotional emphasize rst two Word Firmthey of Mouth Generated Emotional & In Content integral uencers & though, "bestViral of Marketing both My Q: questions Trade appeal offIf between really shareability want to predict whyworlds" and KPI performance not go machine of adslearning This work Competing route?Firm Generated is licensed under a of effects Content Creative & Viral Commons informational vs Marketing Attribution-ShareAlike emotional appeals4.0 International 60 40 20 30 44 46 48 49 50 64 10 24 26 28 29 34 36 38 39 42 43 45 47 54 56 58 59 62 63 65 14 16 18 19 22 23 25 27 32 33 35 37 52 53 55 57 12 13 15 17 21 31 41 61 11 51 47532 / 65 6 8 9 Learning Goals for this Week Discuss why company tweets may increase demand and summarize results from an experiment that tests the effectiveness of company tweets on demand for products De ne Improvised Marketing Interventions (IMI) and assess aspects of an IMI that are important for its success De ne the term "Viral Marketing" Identify features of an ad that consumers more likely share it Explain the tension between "viral-" and "valuable" marketing Summarize the main features of "valuably viral" ads 2 / 65 Preliminaries 3 / 65 Where Are We Now? Course Themes:. Measuring Advertising Effects. User Generated Content & Social Media Online Reviews Word of Mouth & In uencers Firm Generated Content & Viral Marketing 4 / 65 Today's Topic Coverage Can rm generated content on social media increase product demand? Gong, Zhang, Zhao and Jiang (2017) What helps ads go viral? Moldovan, Steinhart & Lehmann (2019) When is viral content valuable for rms? Akpinar & Berger (2017) 5 / 65 Tweeting as a Marketing Tool 6 / 65 Do Firm Tweets Matter? Strategic Question: Does tweeting increase demand for their products? Speci c Questions: What is the causal effect of company tweets on demand? Do retweets by in uential users help? How?: Large scale experiment on Weibo w/ a media company Note: (again) This paper is very cool 7 / 65 The Experiment Weibo ≈ Chinese Twitter Industry: documentary TV shows One show broadcast per day across seven local channels 8 / 65 Results 9 / 65 Results II 10 / 65 Results III 11 / 65 Results IV 12 / 65 Main Takeaways Effects: Tweeting: positive & signi cant effect Tweeting + In uential retweet: retweet gives a signi cant boost What's the mechanism? (subtle) In uential tweets with broadcast time info attract new viewers Informative tweets (broadcast time) also attract new followers to company page 13 / 65 Improvised Marketing Interventions 14 / 65 Improvised Marketing Interventions Improvised Marketing Interventions (IMI): composition and execution of a real-time marketing communication proximal to an external event Strategic Question: Are IMI's effective at generating online sharing? Speci c Questions: Do IMI's result in greater virality? What type of IMI yields greater virality? Do IMIs enhance rm value? How?: A mix of experiment, quasi-experiment and observational data We'll look at a subset of their results, focusing on the rst two questions 15 / 65 IMIs and Virality Context: Superbowl blackout, 2013 16 / 65 Virality of Oreo Tweet 17 / 65 Virality of Oreo Tweet 18 / 65 Timeliness & Unanticipation Study 2: mTurk Survey 19 / 65 Marketing Implications IMI can help business connect with an audience weary of traditional advertising messages Humor and Timeliness important Keeping a close eye on trends and online chatter... and being able to formulate a witty post/response Warning: Need to be careful that don't offend Example: Burger King on International Women's Day 20 / 65 What helps ads go viral? 21 / 65 Viral Marketing Virality: achieving a large number of views in a short time period due to consumers sharing content online. What do marketers and managers care about virality? 22 / 65 Viral Marketing Virality: achieving a large number of views in a short time period due to consumers sharing content online. What do marketers and managers care about virality? Ability to reach vast audiences in a short period of time at low cost Increases brand visibility and recognition → brand equity (Increase in sales?) Lurking differences: Seeded virality vs non-seeded virality 23 / 65 What Drives Virality? 24 / 65 What Drives Online Sharing? Strategic Question: What characteristics of an ad enhance its shareability? Speci c Questions: Does information focussed content limit sharing? Does brand prominence decrease sharing? Do positive emotions increase sharing? How?: Online video ads on YouTube Remark: Paper is descriptive rather than causal 25 / 65 YouTube Advertising Data All video ads between Nov 25, 2013 and March 4, 2014 Approx 1,962 video ads Info they get via Youtube's API: Shares on Facebook, Twitter, Google +, LinkedIn over rst 30 days Info about Ads: sample of 360 of ads collected, covers 79 brands Got coders (i.e. paid some people) to watch ads and write down whats in it 26 / 65 Ad Characteristics 27 / 65 Regression Equation Note: Think of this as a descriptive regression They'll also use it for prediction on an (unseen) sample "Are these features predictive of shares?" My Q: If they really want to predict though, why not go machine learning route? 28 / 65 Estimation Results 29 / 65 Main Findings in Regression Information decreases shares But not for a new products And less for high price products Positive emotions lead to more shares High arousal emotions not more effective than low arousal Timing of brand appearence and length of ad matters A critical eye: They are cutting the data many ways for only 346 observations! (Paper also discusses predictive power of model) 30 / 65 Dramatic Elements and Emotions 31 / 65 Marketing Implications Drivers of sharing important to understand to help ads get shared This sounds a little too causal... as will a lot of their implications The authors want us to believe: Less information More positive emotions More careful brand placement are important to promote online sharing 32 / 65 When is viral content valuable for rms? 33 / 65 Motivation Strategic Question: Can a rm create shareable ads that bene t brands? i.e. ads that are Viral: highly shared, and Valuable: impact KPIs of interest Why relevant? Rush for highly shareable ad content... At the expense of advertising effectiveness "Content aspects that increase sharing seem to decrease effectiveness" Competing effects of informational vs emotional appeals 34 / 65 Flow of the Paper: Lab Studies Multiple studies → "Typical" consumer behaviour style paper Study 0 Examine sharing of online video ads Do certain types of ads generate more virality in "real" data? Study 1: Experiment to test ad type on shares and brand outcomes Study 2: Enriches Study 1 to... Rule out alternative explanations Test external validity Study 3: Investigate mechanisms behind the effect 35 / 65 Field Observation of Viral Ads Data: Unruly.co Prominent viral ad tracking platform Time: June 2003 Outcome: Number of shares in 6 months after launch Covariates: Informative vs Emotional on a 7 point scale Integral-ness of brand to narrative 36 / 65 Field Observation: Findings Trade off emotional ads vs brand integralness Emotional ads are shared more than informative ads Informative ads are shared more when brand integral to narrative Brand integral-ness not important in adds that emphasize emotional appeal 37 / 65 Study 1: Setup Lab Experiment: Undergrads @ Wharton Randomly assigned an ad, then asked. Willingness to share. Brand Evaluation. Purchase Likelihood 38 / 65 Study 1: Results Emotional integral "best of both worlds" 39 / 65 Study 2: Setup Lab Experiment: Undergrads @ MEF 40 / 65 Study 2: Results 41 / 65 Study 3 Same as Study 2, but at Erasmus U Suggests mechanisms driving main effects Main results: 42 / 65 Study 3: Mechanisms 43 / 65 Takeaways Emotional ads increase sharing Informative ads increase brand evaluation and purchase intention Emotional (brand) Integral ads combine both effects This is the "valuable virality" the authors care about They argue this is by: (a) creating positive inferences about persuasion, and (b) increasing brand knowledge (... but I can't see a competing story being thoroughly tested) 44 / 65 Recap 45 / 65 Summary Company tweets (these are ads) can generate demand, in uential retweets even more Mechanism: attracting a new audience Improvised Marketing Interventions are heavily shared when they are timely and involve humor and aren't anticipated Features of ads that are associated with increased virality: Feature positive emotions Are less informative There are exceptions here - new products, expensive products Careful placement of brand name Trade off between shareability and KPI performance of ads 46 / 65 Course Wrap Up 47 / 65 Topic Coverage. Measuring Advertising Effects Attribution Models Media Mix Models Field Experiment Basics Incrementality Experiments in Advertising. User Generated Content & Social Media Online Reviews Word of Mouth & In uencers Firm Generated Content & Viral Marketing 48 / 65 We read papers! Why? We wanted the latest info, common causal methodologies Marketing books often throw causality out of the window And then professionals wonder why they don't see large effects Even if the "latest info" was 5-10 years old... Which was rarely the case! I don't see it summarised in books 49 / 65 We wanted academic marketing knowledge! Why? Approx 50% of marketers in the eld have not studied marketing There's a large knowledge gap in the eld Which slows down progress in rms Practitioners often reluctant to engage with marketing academia Reasons: complexity, perceived relevance, abstract nature 50 / 65 Why Lack of Academic Knowledge Matters. Ineffective Marketing Strategies. Misalignment with Consumer Needs. Poor Understanding of Marketing Metrics. Ethical Concerns. Reduced Innovation. Diminishing Consumer Experience I hope this class gave you insights to these aspects 51 / 65 Lectures Condense papers into a common structure: What is the question? Why is it relevant? How can one measure this? What did we nd? 52 / 65 Why "so much" data/stats? Almost always: one method, one type of analysis Experiment (Firm controlled OR Quasi/Natural) Linear Regression / Difference in Differences Just the beginning of a larger set of tools My mantra: If you can't measure it, you can't build an effective strategy around it Ideally, with causal measurement Appreciating the how helps you contextualize results Otherwise it's memorizing stylized facts 53 / 65 We used R! There was a learning curve We might've handled that better than we did... And expect to next year But you have a valuable skill That employers care about, and/or Helps you if you move into Master's study 54 / 65 Evaluations Link here: https://app.evalytics.nl/#/login?evaluationId=158180 Let us know your thoughts! We want to know what you think Particularly if you have constructive ways to improve the class Context is king... the more you explain your situation the easier we understand your thoughts i.e. I'm happy to read longer, or less positive comments that provide context My Goal: 50% response rate 55 / 65 Exam Structure PART A: 15 Multiple Choice Questions - 30% High level stuff, less nitty gritty PART B: 7 Short Answer questions - 70% Each question is worth 10 points, each question roughly corresponds to one lecture Split into multiple parts to help you answer what we're looking for Sentence LIMITS apply 56 / 65 What can YOU bring to the Exam? NO calculators Nothing to calculate that is dif cult NO Dictionaries NO Notes 57 / 65 What's Available to You? Course Lecture Slides Look at the learning goals! Lab Assignments 6 documents, approx 100 questions! Mock Exam Same structure, different questions 58 / 65 Why Limited Solution Availability? My experience: answers → learning the answers Rather than understanding material Leads to low grades Especially outside of multiple choice questions ALL answers are in the slides + papers Mainly the slides Revise by looking for the answer + discussing with peers Speci c questions? Email us... use the COURSE email address Not our personal ones (Why?) doesn't get lost in daily email volume 59 / 65 Email Policy Before the Exam NO answers within 48 hours of the exams One Exception: Emergencies 60 / 65 Parting Thoughts 61 / 65 For the Exam Be explicit Write, even when you don't know the answer, write your thoughts Clearly express yourself via assumptions, interpretations, caveats where needed We can't grade what we can't see Use your time wisely (and in proportion to the weight of a question) It's a marketing class That uses an analytic toolkit to answer a question But be open minded with what 'marketing' is 62 / 65 Going Bigger I hope the class has shown that good analytics is critical for marketing strategy Not only in digital marketing If you enjoyed picking up the skills along the way: Pursue an analytics career, there's lots of employer demand... even among rms who don't know it yet If you found it more painful than fun, but persisted: When you work with an analytics team on a project you can appreciate how the sausage was made It's not easy, it is time consuming 63 / 65 64 / 65 License & Citation Suggested Citation: @misc{deerdsms2023, title={"Digital and Social Media Strategies: Firm Generated Content & Viral Marketing"}, author={Lachlan Deer}, year={2023}, url = "https: github.com/deer marketing lab/dsms lecture fgc virality" } This course adheres to the principles of the Open Science Community of Tilburg University. This initiative advocates for transparency and accessibility in research and teaching to all levels of society and thus creating more accountability and impact. This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International 65 / 65