Digital Marketing Final

automation

("if user looks at product X but doesn't buy anything, wait 8 hours then send a retargeting email") Don't need a person to do that, simple, automatic, very little risk involved This isnt even AI its just raw automation because it allows consultants to charge more money May or may not use machine learning, AI, deep learning, etc. Sufficiently complicated automation scripts give the appearance or illusion of AI Generally doesn't adapt or change or alter scripting unless acted on by a manager

analytical

("let me know when our daily sales versus competitor X rises or falls more than 5%") Lots of machine learning Tracking and synthesizing data about customers and markets

generative

("write me some upbeat copy for this instagram post announcing our holiday sale") Ex: ChatGPT

pixels

"Live" in the company's server Pixels are tiny, invisible images that websites use to track your online behavior, like a spy that follows you around different sites and collects information about you, such as the websites you visit, the time you spend on them, and even your IP address. Visit a site, when the page loads it loads a 1-pixel transparent image that tells the pixel server your IP address requested for that particular page Nearly impossible for the consumer to clear→ don't live on the computer so we don't have control, cannot disable them Theoretically uses a hash to protect individual anonymity but very bad at it A key driver of retargeting Tells you device type, browser type, exact time of visit, referrer URL, interactions with the site FTC: need to be careful with these EX: TikTok Pixel Tracking TikTok knows tour behavior on any site that has a TikTok pixel Sites that have TikTok pixels: Weight Watchers, mayo clinic, WebMD, GirlScouts, RiteAid Pharmacy, Planned Parenthood, RAINN.org (sexual assault support site) Don't even need a Tiktok account for Tiktok to track u on these sites

Cookies

"Live" in your browser Visit a site, downloads a tiny snippet of code that marks your browser as having been there, what you did, etc. Marks who you are and what you've done at that website Relatively easy for consumers to clear Can contain personally identifying information A key driver of personalization Cookies are very useful and allow sites to know who we are

vimeo deceptive patterns

"here's everything you're giving up when you cancel"] -can also do this when signing up--> bunch of paid options but free version is tiny and at the bottom so consumers do not notice it

How the Chinese Market is Different: BAT Changes Everything (Baidu, Alibaba, Tencent)

(i.e. "What if a media company just up and owned everything?) Ex. Tencent Largest gaming platform Multiple new agencies The two largest social media platforms Two major online payment platforms Largest video streaming service Largest online sports video platform 2nd largest owner of China's largest online retailer One of the key things that make the Chinese market different from a marketing perspective One company that does everything We think of Google as having a monopoly but this company has so much more It would be like if Google owned Amazon, Shopify, Citizen's Bank, Playstation, and the New York Post all at once Tencent knows everything about online users They can track that information from initial media exposure all the way to purchase Chinese market is already over 25% of all luxury and beauty consumption products worldwide In about 6-7 years, it is going be about 50-60% of the worldwide market for luxury dealers Chinese consumer averages ~7 hours per day looking at their phones, hyper majority of media travels through smartphone channel (Americans are about half the time) Tencent captures ~ 55% of all Chinese online time within their ecosystem Singles Day easily creates more sales than Black Friday and Cyber Monday combined, ~90% of all Singles Day purchases are on mobile devices In China, the smartphone is everything *digitally driven economy, mobile phones drive economy over competitors

Brand Trust and Reputation

-as much about setting proper expectations as it is about proper behavior -we wants brands to be consistent in their personality, behavior and consumer interactions -establish open lines of communication with your customers -brand reputation is a double edged sword -the stronger your reputation, the higher the expectations your customers will have towards your performance -little evidence that strong relevance or brand strength insulates from fallout after mistakes, if anything it can amplify it due to expectations not being met

ESG (Environmental, Social, and Governance)

-becoming increasingly important for businesses -are you good to the environment?, reduce environmental footprint, uses natural resources responsibly -social: cares for employees, cares for communities -governance: positive economic contribution, supply chain transparency, clear codes of ethics and fair business -soft metrics that turn out to predict a whole lot, not enough just to make a good product or service -corporate culture/reputation important rather than product

Incoming CPPA/CPRA Legislation

-new california legislation where consumers are forced to consent due to deceptive patterns 5 new requirements: 1. easy to understand--> no hiding behind technical jargon 2. symmetry in choice--> Yes vs Ask me later, accept all vs more information not allowed 3. no confusing language or elements--> no double negatives like yes/no toggle after "do not share my personal information", no confusingly labelled on off toggles 4. no manipulative language or complex architecture--> No "yes" vs "no", I hate saving money 5. easy to execute--> you can't hide links to pertinent security info -IF complaints are brought and found to be valid, $7500 per intentional violation, $2500 per unintentional violation that is per affected consumer

Real Time Targeted VPP

-now can do customized product placement--> ex: cooking show and its a consumer in US so they add a US olive oil brand -currently, real time targeted VPP does not require disclosure or notification on the part of platforms or content providers -Netflix does not have to disclose this information

Virtual Product Placement

-products added virtually that were not present in video shooting, very sly in product placement -mainly builds brand awareness, reminder marketing, actors cannot actively interact w/products since added virtually later -now can do customized product placement--> ex: cooking show and its a consumer in US so they add a US olive oil brand -everyone watches TV on streaming services not on cable TV anymore

Why is brand trust so important?

-we don't trust anything anymore, increasingly polarized -erosion of trust in all aspects of public life -people trust companies to be ethical more than the government

ensnarement

1. Systemization: think of products and services as ecosystems 2. Spawning: designing products or services that are easily shared

cadence

1. new products or product updates need to happen frequently, consider the pace of the competition 2. break apart longer content into smaller more digestible content that can be spaced out 3. encourage and boost user-generated content 4. communicate both the good and the bad, be open about failures or things that did not work

Livestream Commerce

2020: 1.2 trillion yuan, ~10% of online Chinese shopping market The US market is significantly smaller but growing Why is it popular? Generate immediate and quick sales uplift, immediate ROI checks Reach younger and new market segments Launch new products Generate buzz, educate consumers It is not subjected to short 30 second advertising Can actually educate the viewers on the product Amazon, Tik Tok, and Facebook take part in livestream shopping *US very slow to uptake livestream shopping but there is a massive marketplace in China*

Gen Z and the Collapse of the Purchase Funnel

68% of consumers are more price conscious than in prior years 57% are making fewer impulse purchases 58% are doing more research before they buy -impacting the entire purchasing ecosystem -everyone is looking at GEN Z now for product advice, branding advice -impacting everyone in the market 78% of consumers discover things about a brand that drive loyalty after the first purchase 1. initial purchase is a trial run 2. if product/service passes the quality bar, then increased engagement and learning takes place--> brand can build trust and rep at thiis point but not before 3. trust and rep then serve as vehicle for future growth--> trusted brands more likely to inspire word of mouth and reputational marketing in addition to increased purchases * have to pass both of these stages before any word of mouth

Brand Trust as a Source of General Security

9% increase in degree to which consumers are attracted to brands that provide a sense of safety and security rather than a sense of adventure

Algorithmic Bias in Marketing

An algorithm using the data to produce a targeting policy would also unproportionally target a specific population due to underlying correlations in the data Is this a problem? → legally no but ethically? EX: Facebook's ad delivery algorithm is discriminating based on race, gender and age in photos Meta tries to deliver relevant advertising, ads you are more likely to click on, relevancy is based on prior data, prior clicks Pictures of African Americans to AA consumers, pictures of older women to older women The advertiser didn't choose this/didn't ask for this, they are not trying to discriminate or add bias→ The algorithm is introducing discrimination based on prior behavior and revealed user preference Problems: people missing out on job opportunities on Facebook because of gender based on prior clicking behavior→ reinforces bias ex: "blue collar" job openings far less likely to be shown to women due to algorithm

urgency

Baseless countdown timer→ Hurry up sale ending in (timeline) False limited-time messages False discount claims

Electro-Encephalo Gram

Better than fMRI for moving images and dynamic content since you get data in "real time" Separating "signal" from "noise" is very hard Brain is a constant whirlwind of electrical activity, isolating what is due to your stimulus can be very hard Need to map data back onto large scale brain models of "what does what" Note that a "real" EEG headset is bulky, has many sensors, and needs contact using paste/gel to work well Can only measure activity in the cortex; no deep brain structures Also can't measure hormonal response, neuro-receptor response, etc Companies making those claims are a big signal that they're not to be trusted Generally best for engagement and more gestalt good/bad judgments "Brain Wave Marketing" via EEG research Brain waves are real, but almost all application of brain waves to marketing purposes is pure snake oil (not real) The ability of any given commercial, or marketing message, or store environment, to affect your overall brain wave ratio is very minimal "Targeting" your advertising towards a particular brain state also makes no sense

Dark Patterns

Bringing dark patterns to light workshop by the Federal Trade Commission (FTC) Little tricks websites can use to encourage people to buy things they normally wouldn't Endorsements: False activity messages Deceptive customer testimonials Deceptive celebrity endorsements Parasocial relationships pressure Ex: I love this product→ fake Scarcity: False low stock messages False high-demand messages These are scripts you can buy for your platform and they give fake numbers to urge consumers to purchase This is not illegal Ex: ONLY 1 ROOM LEFT AT PRICE, 57 PEOPLE LOOKED AT THIS IN THE PAST HOUR Urgency: Baseless countdown timer→ Hurry up sale ending in (timeline) False limited-time messages False discount claims Obstruction: Price comparison prevention Cancellation roadblocks Immortal accounts Ex: you have to go to the gym in person to cancel your membership Sneaking and info-hiding Sneak-into-basket Hidden information/costs/subscription Drip pricing Intermediate currency Interface interference Misdirection False hierarchy/ pressured upselling Disguised ads

Geo-Conquesting

Came out pre covid Using geofencing to deliver targeted advertising when the customer is near a competitor's store Ex: Whole Foods→ put geofences around competing higher-end grocery stores and health food marketplaces when Whole Foods was getting ready to open a store in a particular area→ near a Trader Joes "hey Whole Foods is opening down the street" Ads drive users to Amazon App, Whole Foods app or Facebook Post click conversion rate ~ 4.5%, 3x industry average Actively fighting competitors Ex: Burger King in 2019→ updated their smartphone app with simple in-app ordering and pick up options, wanted to drive people to the app They put geofences around 14,000 McDonalds If you have the BK app and get within 600 feet of a McDonalds, unlocks a promotion where you can order a Whopper for a penny from the closest BK and the app guides you there This was extremely effective. 1.5 million app downloads, 3.3 billion impressions, 500,000 whoppers redeemed

machine learning

Can update itself overtime→ "hey this script isn't working we need to improve it or add a new one" Generative AI Predictive AI Prescriptive AI Many different mathematical routes based on desired goal

Third Party Data:

Data collected from customers by a company that isn't directly involved in the transaction Ex: im interacting with girlscout webpage but random tiktok pixel is there, random third party trying to collect data End of 3rd party cookies→ companies are using a wide range of alternative solutions in preparation

Brand Trust Elements

Competence, Ethical Behavior, Relevance

Top Uses of AI in marketing

Content personalization predictive analytics Targeting decisions Customer segmentation Programmatic advertising and media buying

zero party data

Data knowingly volunteered by the customer upon direct request Surveys, quizzes, polls, pop-ups, contact forms, registration, etc. Generally, no privacy concerns as long as you do not share it with 3rd parties

First Party Data

Data you collect directly from your customers and potential customers by analyzing their interactions with product or service Web activity, site interactions, order history etc. If using customer data to improve marketing, need to tell consumer Ex: Someones interactions with a website

Contextual Video Marketing

Contextual Video Marketing is the process of using automated content recognition to analyze various dimensions of the content being watched, such as tone, energy level, aesthetics, and location, in order to serve relevant advertisements that match the context of the video. Al Scans content and codes it on a number of dimensions Roku calls this "Automated Content Recognition" but all major streaming platforms (Netflix, Google on YouTube & YouTubeTV, etc) use this 2X/sec, 1X/sec, once every 5sec, etc Constantly scanning the content being watched and taking photos Original purpose? Audience verification for ads TV w/ RokuOS knows exactly what UserID #24601 is watching whether it's on Netflix, Broadcast TV, Hulu, whatever (note that individual data is not shared with advertisers) Currently in testing? Using it to drive contextual marketing Affective tone Energy level Aesthetics (color grading, palette, etc) Content location (office, outdoors, home, etc) People (solo vs group, presence of children or elderly, ethnic mix) If we have all this information, we can potentially serve contextually relevant advertising Ex. upbeat video, multiple people watching, and location is in an office Advertisement can be for a briefcase or something similar that matches the tone of what is already being watched This is very hard to do in real time Roku only knows what is being watched every 25 seconds which is not fast enough to use that information for real time bidding

better customer understanding

Customer data platforms (CDPs) ML/AI affects both data organization and data queries Integrate data from multiple customer sources Easily match customers with similar profiles in database and historical data as to what does and doesn't work NLP allows for easier coding of rich text customer data NLP/IG creates a more natural interface for marketing managers AI helps synthesize data across multiple platforms AI can fit customers into buckets→ match customers with profile

real time decision support

Detect browser triggers (ad-blocker, etc) and serve alternative content "Next best action" Ex: customer comes to site, places product in cart but doesn't buy and leaves Many options now: reminder email? Push notification? Targeted offer/discount? Chatbot? Personalized discount banner? ML/AI uses what it knows about the customer and what it knows from other customers to predict which action leads to best outcome "Next Best Product Offer" Using customer online behavior, prior purchases and matching with similar customer profiles to deliver hyper-targeted follow up product recommendations Ex: someone always orders same coffee? Offer deal on upsize. Someone high variety seeker? Deal on higher profit margin subcategory they haven't tried yet Ex: someone buys a pickleball racket? Wait one week, then recommend tiger balm

Disclosures and Product Placement

Disclosure Currently real-time targeted VP does not require any disclosure or notification on the part of platforms or content providers. FCC requires network television stations to disclose overall whether product placement in television shows is taking place But this generally isn't applied to product placement in entertainment shows and is more for paid-programming, for entertainment usually a quick blurb in the end credits is acceptable FTC is far less clear. Doesn't require disclosure of product placement in traditional video content (streaming, etc) But influencers need to disclose if they are being compensated So how does Virtual Product Placement fit in here? Is the real-time targeting of VP an issue above & beyond traditional product placement? Is there an increased need to disclose?

EEG/fMRI

EEG: Electroencephalogram: measures electrical activity in parts of the brain. fMRI: functional Magnetic Resonance Imaging: measures blood flow to parts of the brain. In theory, allows you to "see" what a consumer is thinking.

Upfronts: Linear TV

Each major network hosts their own, sometime between late March and early May Shows off next season of programming (September-October thru late April-May) Ex. NBC will bring forth their next seasons programming content Goal is to pre-sell as much ad space as possible by the end of the Upfronts; commitments from major advertisers Note traditional linear TV advertising doesn't do programmatic / real-time; all that ad space is booked far in advance Ex. why you should put your advertising on NBC These ads are bought far ahead in time (March-May) prior to the beginning of new seasons This is also where new shows live and die; pilots are available by the Upfronts, if they don't sell many ads they don't get made into a series Trying to sell advertisers on their upcoming content *NBC, FOX etc. host presentation of next season of content and pre-sell ad space

Newfronts: Digital Video & Streaming

Early May Major streaming platforms, digital video platforms, Smart TV manufacturers, content providers, "sell" their services to brand managers, ad firms, demand side platform execs, etc. Usually not as much about individual shows and more about the advertising tech Still try to close pre-commitment deals; move Inventory out of the programmatic pool and into signed contracts Whatever they couldn't sell at newfronts Ex. Amazon's 2022 Newfront Announcements included 11 year NFL deal for Thursday Night Football Launch of Freevee as a competitor to Roku's FAST Launch of Virtual Product Placement Acquisition of the MGM Film and Television library of content New suite of creator monetization and brand integration tools on Twitch Launch of Amazon Marketing Cloud for marketing media data analysis

mastery

Endowment effect: if consumers exert effort for a brand it deepens their sense of ownership Affinity Effect: we overvalue things we are familiar with

MACE Fortnite Example

Ex: Fortnite→ MACE personified, launched in 2017, saw massive success for multiplayer battle royale format→ Repurposed engine to this in 2 months and was a massive success in the marketplace by the end of 2018 with 200 million users Fortnite: Mastery: Noticed when people played PUBG that playing defensive is the best strategy so repurposed base-building tools in the game to be able to build platforms, outperform shooters, and encourage you to play more and more Fortnite Accessible: Free-to-play, cartoony look for all ages appropriate, client can play on almost any device Cadence: New abilities, weapons, and skins weekly, and major map changes Ensnarement: Advanced social features, easy party chat, matchmaking algorithms

Endorsements

False activity messages Deceptive customer testimonials Deceptive celebrity endorsements Parasocial relationships pressure Ex: I love this product→ fake

scarcity

False low stock messages False high-demand messages These are scripts you can buy for your platform and they give fake numbers to urge consumers to purchase This is not illegal Ex: ONLY 1 ROOM LEFT AT PRICE, 57 PEOPLE LOOKED AT THIS IN THE PAST HOUR

MACE framework

How to build enduring brands in the digital age Mastery: reward consumers for demonstrating expertise in your brand Accessibility: make your brand as accessible to as many consumers as possible Cadence: have a meaningful message to share with your customers as frequently as possible, and rhythm and speed at which we communicate with customers→ customers screw this up when they communicate something not meaningful Ensnarement: make your brand as sticky as possible through switching costs and network effects, make an incentive to stay with our brand

AI in past 5 years:

Increasingly AI is integrated into full service platforms Big shift in past 5 years from isolated AI instances to integrated platforms Ex: generating marketing email to customer it can use AI from customer history, type of voice, previous purchases, all linked together now

KOC: Key Opinion Consumer

KOC doesn't necessarily need to have a huge online following, but for a particular category they exert a much higher-than-normal consumption pressure on their social circle and online contacts KOCs are much easier to identify when you have extreme vertical integration like you do in China that links media, local behavior, work behavior, shopping, etc. Ex. your friends ask your opinion about products you are well knowledged in Or someone has a small following but people trust their opinion about products Li Jiaqi, China's 'Lipstick King' Raised more than US $145 million in sales on Singles' Day 30 million subscribers All he does is demonstrate and try on lipstick His is a millionaire Extremely influential in the Chinese marketplace

KOL: Key Opinion Leader

KOL is a subset of influencers who are considered to be experts in a particular field Sandra Lee (Dr Pimple Popper): dermatology, skincare Loopop: Modular synthesizers Ron Scalzi: Science Fiction Often KOLs derived their original aura of expertise outside of social media and "influencing" Their expertise has caused them to become famous Key aspect of KOLs: their followers act on their advice, opinions, or recommendations at a very high rate

CTV as Data Resource

Kroger and Disney come together in a data sharing partnership Kroger will share the shopping data and Disney will share the viewing data Show ads for food and then Kroger will see if it impacts grocery purchases

coca cola AI

Limited edition sodas Year 3000 coke is supposed to taste like the future Used AI collaborations to design the color scheme, the can, imagery Create a bunch of insta tiktok style executions for this campaign Can take any old picture and it futurefies it What would coke look like in the future? Partnered with design firm in Hong Kong called Ambush (fashion goods, jewelry) and helped with the filters to futurefy your world This is what we think of for AI in marketing

interface interference

Misdirection False hierarchy/ pressured upselling Disguised ads

neuromarketing pros and cons

Neuromarketing is a fantastic tool for understanding consumer behavior by bypassing biases introduced by surveys and qualitative research, but it is less effective at directly manipulating behavior and tends to have subtle nonconscious effects. Neuromarketing as a tool for better understanding why consumers are doing what they're doing? Fantastic tool Bypasses cognitive filters Surveys & qualitative research both introduce tons of bias (temporal, social, memory, cognitive, etc etc etc) Lots of consumer behavior isn't rational & is routinized, low-conscious effort, and heuristic in nature Neuromarketing as a way to directly manipulate consumer behavior? Significantly less useful Nonconscious effects tend to be subtle They amplify or attenuate stuff that's already there Better at affecting emotion or cognition/ memory, direct effects on behavior are relatively rare

Subliminal advertising and memory insertion?

Neuromarketing is generally not this Nonconscious influence is real but marketers generally cannot "subliminally" make you do things you do not want to. Most 'studies' showing direct behavioral purchase outcomes of subliminal primes are not real or were faked Example used in class L'Oreal in partnership with global neurotech leader, Emotiv, launches new device to help consumers personalize their fragrance choices Not only a good research tactic but also a good marketing tactic The consumers who use this have a story to tell their peers It also helps to gain brand loyalty and brand recognition

Monitoring: Online Reputation Management Tools

Online review management Identify trends, respond to reviews, monitor review networks Social listening Combination of real-time dashboards and project-based studies Links with Social platform Customer experience Look at feedback volume, trends & sentiment, share of voice Links with CRM platform Crisis communication Organizing and delivering response to the relevant channels quickly Competitive analysis Monitoring brand distance, pos-neg ratios vis-a-vas competition Brand advocacy Amplifying positive information and voices Links with Social platform

mental simulation

Our brains are mentally simulate possible actions, so we design marketing elements to make that easy Ex. First person video makes us imagine ourselves in ad content Hands using a product makes us imagine using the product ourselves

The Changing Shape of Digital Video Advertising

SVOD: Subscription Video on Demand Netflix, Max, Disney+ AVOD: Ad-supported Video on Demand Hulu, Peacock FAST: Free Ad-supported Streaming TV Tubi, Roku, Pluto, Freevee vMVPD:Virtual Multichannel Video Programming Distributer Youtube TV, Sling, DirectTV Now, Hulu Live Today, Netflix has the largest amount of subscribers by ad-free/ad-supported subscriptions

Brand Rep Subdivisions

Performance: profitable Leadership: well organized, appealing leader, excellent management Products/Services: high quality, value for money, meets customer needs Citizenship: environmentally responsible, supports good causes Governance: open and transparent, behaves ethically, fair business Innovation: adapts to change quickly, innovative Workplace: rewards employees fairly, employee well-being, equal opportunities *lots of different aspects of brand reputation and brands score very differently on lots of different aspects

Obstruction

Price comparison prevention Cancellation roadblocks Immortal accounts Ex: you have to go to the gym in person to cancel your membership

Solution: Contextual Marketing

Rather than target ads on user behavior, it targets ads on the content and context of the ad location Ex: Tylenol runs ads on financial websites when the stock market drops more than 100 points Ex: NY times book section advertise reading glasses Ex: recipe website→ ads for food on website

When do influencers need to disclose information?

Relationship: financial, employment, personal or family relationships all need to be disclosed ex: best friends with CEO Actions: if you have a large enough following online, liking, tagging, pinning and other forms of social media activity can be considered endorsements ex: selena gomez liking another brand post -Global: if your posts reach a significant portion of the US audience, US law applies -Negative: you are not required to declare a lack of relationship if you are talking about a product/service with which you dont have a relationship and are not being compensated for talking about ex: no relationship not needed to disclose

Roku "Shoppable Ads"

Restaurant places Shoppable Ad using Roku's ad services During ad, customer clicks "OK" button Text message or email link sent to bring user directly to that restaurant within DoorDash Extremely direct way to measure ROI Can create special discounts that "trigger" with the link sent Ex. Wendy's ad activates $5 off orders of $15 or more

Types of WOM Programs (word of mouth)

Seeding: Product seeding: get product/service into initial group of influential people Viral marketing: seed influencers with product/service Referral: Referral reward: give customer reward when they get others to sign up Business reference (B2B clients) Affiliate marketing: $ incentive to drive others to site Recommendation Narrowband: via consumers social networks Broadband: via review sites like TripAdvisor/Amazon

Embodied cognition

Sensory metaphors that affect how you think Ex. Holding a warm beverage makes you think the people are talking to have warmer personalities Brand names with inward vocal movements (vs. outward) increases predicted taste for food products Inhalation vs. exhalation when saying a brand name The Holiday Market at Snowport Chase brand Activation Service: fee-free ATMs Goal: Build brand equity in 20-30 year old Boston Urban Proessional target market Task: design pop-up shop space that uses embodied cognition/sensory marketing to create a brand position and narrative for Chase Use 2 senses other than vision that can convey brand meaning (smell, hearing, touch, taste) What sensory trigger do you want to activate? How will you do it? What is the non-conscious association you want to create? How do we design a digital element on our website, smartphone app, or social media feed that reinforces that sensory association?

Reach Vs. Frequency "Smoothing"

Set desired Reach & Frequency goals for an ad campaign at launch Ex. "I want this campaign to reach 30% of my target audience and the average person that sees this ad sees it 4 times" CTV (connected television) programmatic algorithms can "spread" advertising out in real time across your desired target market since it knows a User ID for each user "oh , #24601 has already seen this ad the desired 3 times, let's not show it to him and instead show it to someone else" Have someone else in the target market see this ad instead Was very difficult to do well before Machine Learning took off This cannot be done with linear television

Warner Music AI

Signed a record deal with this virtual popstar Personalized soundscapes to help you focus, relax, and sleep, backed by neuroscience Music should be moving faster than video in terms of AI AI may be running behind in music

Solution: Zero Party Data

Smarter Customer Profiles and Personalization Yelp lets you enter dietary preferences, needs etc. Quiz questions, use data to change the cadence and tone of email communication Real-time segmentation "Lookalike marketing": uses a combo of 0 and 1st party data to make profiles of owned media usage and try to fit new users into those categories to customize the website experience before any personal information is given/requested Loyalty data ecosystems The loyalty program has options to earn extra points and discounts for volunteering personal information, then uses that information to deliver more targeted marketing and promotion, personalized product recommendations

sneaking and info-hiding

Sneak-into-basket Hidden information/costs/subscription Drip pricing Intermediate currency

"Social Proof"

Social likes and shares Long-form testimonials Customer reviews Trust signals (logos and star rating, review snippets from product reviews, google seller ratings from your paid ads in SERPS, etc) Case studies Celebrity or influencer endorsements Earned media OTHER PEOPLE LIKE IT SO IT'S PROBABLY GOOD

Second Party Data:

The first-party data collected by a "partner" company, made available for our use on some basis Ex: using visa to check out on amazon, data needs to be shared between these two companies

A/B Testing

This is the process of comparing two variations of a single variable to determine which performs best in order to help improve marketing efforts A/B tests can be used to test a global change or to identify heterogeneity in response to the intervention and implement a more personalized approach -simple statistical analyses can be used to analyze heterogeneity in the response to marketing actions -targeted strategies enable firms to efficiently use marketing resources and increase profitability -targeting and personalization practices may result in policies that discriminate against certain types of customers

Dominant Livestream Platforms

Twitch Has the ability and the popularity to enter the livestream shopping market but is very slow in doing so TikTok Trying to really expand into the livestream market Started paying livestreamers a larger payout Highlighting livestream shopping lives NPC trend on TikTok live Tipping people and sending them gifts on a livestream and they will do an act associated with the gift

Why Did This Blow Up So Fast in China but is Growing Much Slower in the US/EU

US/EU had bifurcation in online titans (Facebook / Insta / YouTube were media, Amazon was shopping), Taobao was both right out the gate There was a split between social media platforms and online shopping platforms Taobao was both initially Liveshopping 'feels' very natural in a unified platform Technology for natural link to sales already exists on the Chinese platforms links to buy on western social media feels bolted on video streams on platforms like Amazon feel somehow 'wrong' and remind us too much of HSN and QVC Commission-based cuts can happen automatically on Chinese platforms(KOL (an influencer) usually makes 20-30% of sale and negotiates discount on product for buyers during stream), need to be negotiated into contracts and handled post-hoc on US platforms China's got 5 years of hosting experience head-start. US may catch up.

Inbound and Outbound Data Paths Uber and Uber Eats Example

Uber unifies customer data over 50+ social media and marketing platforms Uber knows everything you do even when you are not on the uber app Can target you by destination If you're on the way they can use destination for your ride to give you relevant ad information→ ex: on the way to a bar they show you ads about nightlife, on the way to doctors they show you ads about medical stuff Inbound Datapath TO Uber Uber using purchase info from Stubhub or Ticketmaster to send notifications the day before or the day of to reserve rides to/from the show Sending you uber notifications the day before or day of to reserve a ride with them for the concert Ex: Airline info. → know ur taking a flight and send you notifications to book rides with Uber Ex: use people's social media and up charge ride fares if they know there are a ton of people in one area→ posting a story from a location they know where you are and use this data Outbound Datapath FROM Uber UberEats users who order an icecream sundae from McDonalds are 4x more likely than other users to order an apple pie, McDonalds then makes an Apple Pie Mcflurry Based on the data they offer this mcflurry exclusively on uber eats Uber helping other businesses get better

Reputation and Symbology

Use visual iconography to drive reputation We can put symbols on website to say hey trust us Payment iconography→ company that only accepts Venmo vs company that has Amex, discover, visa, etc. "US Search Awards"--> more validity

Biometrics

Using eyetracking, galvanic skin response, heart-rate, etc to measure variables like attention, engagement, excitement, involvement, etc Physiology signals As marketers, we have been looking at biometrics since the 1980s

What is Neuromarketing

Using physiological and neurological signals to help understand consumer behavior Combination of marketing, neuroscience, and psychology Key idea: a lot of consumer behavior isn't highly cognitive, or even conscious Snap judgments Emotional reactions Simplifying heuristics & habits Traditional marketing research (surveys, focus groups, depth interviews, behavioral experiments) are all extremely high-effort and conscious on the part of the consumer, so don't map onto how we normally process marketing information very well. Neuromarketing goes "around" those cognitive barriers to dig into people's more direct emotional, physiological, and neurochemical reactions to marketing content

emotion recognition

Using webcams and image processing to encode facial expressions into a suite of emotions in near real-time. Facial expressions reveal emotions far more accurately than verbal response. This is comparatively cheap in comparison to fMRI and EEGs

Sensory Marketing: Neuromarketing without the headbands

Visa created a signature sound that plays when someone completes a purchases using their card When people hear it, they will then recognize that it was a true Visa transaction and trust that it was a real transaction Want the noise to signify trust But wait... isn't this digital marketing? Can we engage the senses online? Sure Vision and hearing are easy Touch is constrained but possible Smell and taste... not so much But we can cue memory of it via other sensory information

Solution: First Party Data

Website visits: visit duration, how they came to site, bounce rate, return rate Email newsletters: open rate, link click rate, clickthroughs that result in action, ratio of subs to unsubs Content engagement: # of comments, likes to dislikes ratio, referral source In-store purchase: upsell rate, product/service popularity, abandonment and repurchase rate First-Party Data Challenges: It's pretty hard to get all of your 1st party data sources talking to each other and synthesizing insight Ex: smaller online retailer→ sales data from Shopify, web interaction from Google Analytics, social media data, loyalty data, inbound metrics Long way to go

Crossmodal correspondences

Where your sensations in one sense prime a different sense Ex. Objects with more saturated color feel heavier Snack chips with a louder crunch are viewed as having more flavor

Estee Lauder AI

You can upload a picture of your face to find the foundations and lipstick that match your skin tone Launched an app that allows visually impaired people to check their makeup Scans your face to figure out areas you did not blend or need more Uses advanced image processing

Geo-fencing

Your smartphone knows where you are at all times Using the customer's physical location to send targeted messaging via in-app advertising, social media advertising, or notifications Geofencing: general GPS info., company sets a virtual "fence" where the messages trigger within a particular area When someone enters a particular area they are now served particular messages/ads Beacons: close-range Bluetooth/NFC, usually used within-store or within a particular shopping environment

AVOD

ad-supported video on demand Hulu, peacock

Amazon deceptive pattern

amazon sets up subscription based option rather than just one time purchase -also makes cancellation impossible, multiple steps that are confusing for people to figure out how to cancel -erodes brand trust and consumership

"Marketing AI" is a Fuzzy Label: 3 categories

automation, generative, analytical

How to Disclose (Influencers)

be clear about how you were compensated, paid, free trip, whats the nature of the relationship Hard to miss: dont hide the disclosure behind a click or obstruct the visual or save the disclosure for the end of a video Simple and clear: be explicit as to the nature of the relationship Same language: your disclosure should use similar language, cadence, volume, visual elements etc as your traditional content Beyond the platform's tool: influencers should not trust that the platforms automated disclosure policies or tools mean they do not have to take any action ex: putting #AD is not enough

Benefits of Marketing AI

better customer understanding -real time decision support

beacons

close-range Bluetooth/NFC, usually used within-store or within a particular shopping environment

Three ways to digitally engage the senses online

crossmodal correspondences embodied cognition mental simulation

Functional Magnetic Resonance Imaging

fMRI measures blood flow to specific areas of the brain; think of it as what parts of the brain are demanding oxygen at any given moment Much, much more spatially detailed than EEG Much slower than EEG (EEG can detect changes in msec, fMRI several seconds) fMRI is much more expensive and difficult to run compared to EEG But can theoretically derive a lot more insight into what the person is thinking via more specific information about what portions of the brain are activating Key Q to ask is if the cost is worth it vs other research techniques

FAST

free ad supported streaming TV -tubi, roku, pluto, freevee

Is the real time targeting of VPP an issue above and beyond traditional product placement? Is there an increased need to disclose?

its Less we need to disclose product placement but the fact that we are being personally targeted

accessible

make entry level products as cheap as possible market more to younger and first time customers use as many sales channels as possible, meet the customer where they are

SVOD

subscription video on demand, i.e. a service like Netflix, Max, DIsney+

vMVPD

virtual multichannel video programming distributor -Youtube TV, Sling, DirecTV Now, Hulu Live