Digital Strategy and Marketing

With digital information, it was now possible to create endless and identitical (also qualitatively) copies

When it can be stored, accessed, analyzed and (ideally) transmitted outside the device

• Data

• Automated learning and production

• Conntected devices

• Robotization

• Interfaces

  
  

Former analogoues data becomes digitalized, allowing it to be stored, transmitted, processed and analyzed with unprecendeted fidelity and speed

More data and faster computers allow the automation of knowledge discovery, decision making and reporting

New communication protocols and components allow devices to be connected to one another and by extension individuals

New materials, batteries, actuators and software enable the emergence of dexterous automated machines (robots)

High quality images, video, 3D render, and sounds can be delivered with higher definition and realism creating new interfaces to interact with humans

• law remained in place for decades: computing power

• companies used the predication to plan their business model

• predication as a way to organize business

—> today:  everything can compute i.e. Airpods, Laptops, charger, coffeemachines

—> combination of computer power and algorithm

Economies of Scale refer to the cost advantages that result from large outputs.

Positive Feedback refers to the positive effect of past adoption of a product or a standard on the current adoption of a product or standard.

Market tipping refers the increasing advantage of one market player that earns an initial advantage enabling it to take over most or all the market

—> Network effects and switching costs can lead to market tipping

Network Effects are the increase (or decrease) in benefits that a user derives from a good or service when the number of users consuming this product or service increases.

Indirect Network Effects are the increase (or decrease) in benefits that a user derives from a good or service from complementary products and standards developed as a result of product adoption.

The value of a network is proportional to the size of the audience, N

The value of a network is proportional the number of possible connections, which for large N is approximately the square of the number of connected users of the network, N2.

1. There is still no rigorous way to validate Metcalfe’s law and other related laws.

2. Networks grow at different rates over their lifetime and the value can deviate greatly from any law. There are a critical mass of users that increases value faster than Metcalfe’s law.

3. Network externalities can be positive or negative at different points in time or for different markets. Congestion and diminishing marginal utility are better captured by Brescoe’s law.

4. Data availability and quality for testing Metcalfe’s and other laws is still poor. 5.Value of a network depends on the definition of the market 6.Value of a network depends on the business model of the network, which can change over time

Not all connections are equally valuable. Value of network is better approximated by N . log(N).

The value of a network is proportional to the number subgroups possible, which for large N is close to 2 N(hochgestellt)

A Platform is the set of rules, components, and technologies that creates value by facilitating a network of users to interact, connect, and transact*.

*This definition is in the specific context of network platform

A Two- or Multi-sided Platform is a platform that connects two or more different types of users (further discussion and refining of the definition later)

A Two- or Multi-sided Platform is an organization that creates value primarily by enabling direct interactions between two or more distinct types of users.

Two characteristics:

1. Each group of users is a customer of the platform in a meaningful way

2. The platform enables some direct interaction between the sides

Cross–group or network effects are not sufficient nor necessary for a multi-sided platform:

§ If we require cross-group network effects, the definition becomes under-inclusive: e.g., advertisement- based media

§ If we require cross-group network effects in only one direction, the definition becomes over-inclusive: e.g., retailers are not multi-sided platforms

• Connectivity Point-to-point transfer of information or goods between two partners. The sender knows the identity of the receiver before the transaction

• Variety Allows buyers with a need for variety to access suppliers with diverse assortments. The buyer knows the identity of the seller before the transaction

• Matching Allows users of the network to access other users with specific characteristics. They don’t know the identity of the matched partner before the transaction.

• Price Setting By matching traders the platform enables price formation.

Indirect Network Effects occur in networks with different types of units (buyers and sellers, two-sided markets), and refer to the benefits a member derives from increasing the number of members of a different type.

• Part of anti-trust regulation seeks to control monopoly power, which can lead to higher prices and worse outcomes for society. But users value larger platforms, so a monopoly platform might be efficient. Is it?

• Farronato, Fong and Fradkin (2021) explore this idea. They look at a merger between two platforms that specialize on pet-sitting and observe how users behave and how prices change.

• They observe several cities and the original two platforms have different relative market share.

“We find that the acquiring platform experiences network effects, as evidenced by increased participation of existing users who see their choice sets expand from the influx of users from the acquired platform.”

“However, on average at the market level, users are equally well off with one or two platforms, as evidenced by the constant number of transactions, match rates, and proxies for match quality. Combined with our evidence that platform prices did not increase post-acquisition, our results suggest that, on average, a single platform does not provide substantially larger consumer surplus than the sum of two competing platforms.”

What are implications? Prices don’t increase. Network effects are small

• Critical mass (—> network effects work here). You need large number of customers because of network effects and market thickness. Use existing platforms or public data to get initial critical mass

• Provide stand-alone value. Users join the platform even in the absence of network effects

• Manage asymmetric information. Users fear the platform might not survive or not attract enough users on the other side. Use freemium and subsidies.

• §  Manage risk. Users might not trust the other side

• §  Increase compatibility. Users do not adopt technologies that are not compatible with current use. You need to increase interoperability and allow for easy migration.

•  Control disintermediation

The Open vs. Closed dichotomy describes whether a participants needs to be authorized by a sponsor or a provider to participate in a platform, as well as how strict the rules of membership are.

• A platform does not help a struggling product —> not every business needs to be a plattform

• There is no need to go all-in. A platform can co-exist with non-platform products and service

  • Amazon —> Amazon Marketplace

  • Lego—Y Leg Mindstorm

  • Iphone—> App Store

  
  

• Be willing to include competitors in the platform

• Deter competitive imitation

• Involve users

Platform Sponsors design and hold rights to modify the platform and determine who can participate.

—> No obvious sponsor (Internet)

—> One sponsor (eBay, Skype, American Express)

—> Many sponsors (IEEE, Visa)

Platform Providers mediate the interaction among users, and serve as their primary point of contact.

—> One sponsor is the sole provider (Xbox)

—> One sponsor, many providers (Android, App Store)

—> Many sponsors, one provider (Orbitz) § Many sponsors, many providers (Visa, Linux)

—> you want to study trends / data traffic

A business model is a concise description of how an organization efficiently uses its resources to create value for its target customers and capture that value to deliver a return above the cost of capital.

business —> investment ( which costs sth) —> create a supply side to deliver benefit to customer

1. A description of the “demand side”

1. A description of the ability to supply that demand

1. description of how the operating profit will be generated, its

   investment rate and potential growt

Focus on the principles —> business is based on principles

A good framework is based on a fundamental business, psychological, or economic principle. We often see new frameworks emerge, but they are reincarnations of existing principles. So ...

—> five things to think about

• Data—> Lower Search Costs: The cost of gathering information about a the product, the price and other aspect of the transaction is lower than ever

  Data; search became easier / more transparent

• Automated learning —> Lower Replication Costs: Digital products and digital features have a marginal cost close to zero;

• Connected devices —> Lower Fulfilment Costs: Ordering, payment and distribution of goods is decreasing rapidly. Distributing digital goods is costless.

• Robotization —> Lower Tracking / Targeting Costs: It has become easier to learn about individual needs and develop customized products. Targeting individuals with messages and tracking the response is now common; individualization

• Interfaces —> Lower Verification Costs:The cost of veryfing the quality and performance of a product is decreasing. Buyers can verify the reputation of sellers and viceversa, i.e. user ratings at air bnb

-> based on market segmentation (horizontal), based on preferences

The Long Tail business model is to making a large assortment of niche product available and easily searchable.

• first, cost of fulfilment + search cost go down —> allows for the emergene of long tails

• long tail: we have the interest and we can the niche products

  —> making money by focusing unique / niche things

• Search —> Customers can easily search across a very broad assortment

• Replication —> Digital files have very low inventory costs

• Fulfilment —> Digital files can be sent over digital channels; the logistics of physical has become highly automatized

• Tracking —> Customers can received automated recommendations according to their tastes and behaviors

• Verfication —> Customers can read product reviews; customers can check the reputation of third-party sellers

—> framework in terms of costs / effiency

What happens if costs go down, what would that enable / what new ideas emerges?

—> based on segmentation: vertical (quality)

The Freemium business model offers vertically differentiated versions of the product, with the basic version being offered at no cost to the customer.

-> one product at zero price

perceived quality - perceived price

• Replication —> Digital products have zero marginal costs

• Fulfilment —>Digital files can be sent over digital channels at zero marginal costs

• Tracking —> Customers can be segmentead according to value; Target customers to upgrade

examples: Google Drive, ChatGDP, Spotify

—> idea of trial / learning: quality check, testing the product first

—> lock-in: one you use a platform, there is cost switching the platform

—> network effects i.e. Dropbox’s file sharing

—> Advertising

—> Data (whats the value of data / benefit)

An own digital channel refers to the segmentation of the market into customers targeted via digital (branded) channel and customers targeted via traditional high-touch channels

example Dow Corning:

-> invented silcons (which are everywhere)

-> had a segmentation

-> had to lower the cost of service —> website launch

—> having 2 brands when selling higher and lower quality products, it makes sense

—> attract price senstive customers

—> high tech vs. high touch (B2B)

—> create two very business models and brands

high tech: not a high services, lower prices

• Search —> Customers can easily find products without salespeople recommendations and help

• Fulfilment -> Payment is automated, delivery is controlled and managed

• Tracking —> Targeted offers

• Verficiation —> Customers can check quality online

—> emerged in the 80s

—> Idea: Business= combination of things

—> Business functions: we need Invoation / Operation / Customer Management

• everything combined in one bc of transactions costs i.e. negotiation contracts, personnel search costs

—> digital now allows parts of the company to talk to each other and lower transaction costs —> allows to have separate business models

—> i.e. food delivery is an example of unbundling the business model of restaurants

Unbundling is based on the idea that a business is composed of a few fundamental businesses that need to be coordinated. A digital technology takes over the customer interaction creating a direct relationship. Typically this relationship is scalable because the marginal cost to interact with customers is close to zero. They can then re-create the industry by unbundling the core infrastructure

Hotels

• Supply side: Hotels provides —> infrastructure / operations room (rooms)

• Customer Management side (booking)

—> must run it as one business due to transaction costs

—> however, with digital —> makes transaction so much easier and allows to split business model

—> specilization and scale i.e. outsourcing the “back office”

• 1 Business for “rooms”

• 1 Business for “booking”

Air BnB

—> Business Model Customer Management side

—> relys on infrastructure providers -> hosts

Access Economy business model is based on controlling the demand side of the business (i.e., the customer experience) and unbundling the supply side, typically by using underutilized or idle assets and labor.

—> i.e. Air BnB / Uber business models is own to customer relationship

—> outsourcing the infrastructure

The sets of interdependent organizations involved in the process of making a product or service available for use or consumption

—> intermediaries that help company produce / deliver the prodcut to the end user

Before transaction

—> pre-sale service (Informationto customer, Education, Promotion, Certification)

—> logistics ( company that takes care of Transportation of the productto the end customer, Assortment, Availability)

During the transaction

—> product transformation: Customization, BulkBreaking, Installation

After the transaction

—> transaction handling: Invoicing, Payment, Financing I Information: Negoiation

—> After-sale services: Fullfilment: Warranties, Repairs, Servicing I Informational: Support

—> Business model: which functions do you need and who takes care of what

—> online products: when I’m online which functions do I need to replace?

—> Saturn take care of all of the channel functions

—> Online stores cannot provide rich product experiences

—> other functions are necessary:

• subsitute the experience partially i.e. reviews

• Amazon delivers the product

• Join: Collaborate with Amazon Business on a 1P relationship (direct wholesale relationship with Amazon).

• Fight: Exclude Amazon as a channel and build its capabilities and e-commerce platform.

• Multichannel: Partner with Amazon while strengthening data integration from the wholesaler platforms.

• Keep the status quo: Stick to the existing sales channels (sell through traditional wholesalers) and ignore Amazon.

—> Digitizable goods markets were disrupted early because of distribution

• Even if digital Information is expensive to produce, it is cheap to reproduce, i.e. high fixed costs and low marginal costs. This is even true for paper-based information communication (i.e., books and newspapers).

• It is not cheap to distribute non-digital goods

  
  

Amazon

Pro

—> more access

Con

—> loss of control: price,

—> channel conflict: previous distributor

—> no (customer) service / no clear after-sales service

—> give power to Amazon

—> loss of data

Issues:

—> loss of control

—> knock offs: make its your brand

Disintermediation is the use of a digital medium to remove of one or more intermediaries between a manufacturer or a distributor and its customers.

• Example: Airlines selling direct to passengers i.e. buying airplane tickets directly by the airline

• Example: Magazines selling subscriptions online

—> try on the headphones in the store, get customer servies & then go an buy it on amazon -> free riding of services by users

Manfucturer has two distribution channel:

• high-service-high-cost channel (store)

• low-service-low-service (online store)

—> high-service-high-cost channel will stop providing the service (which you need)

—> is the most pressing issues for online channels

• “persuade” information price; recommondation retail price ; show price in advertising (big companies with a lot of power)

• Reward the high-cost channel via i.e. commission, backpayment / revenue sharing

• Assortment, different assortment by quality / seasons

  (i.e. more specialized /quality products are only in stores, more basic products online) by quality / seasons

• Don’t sell online i.e. Luxury brands

—> you want to be a monopolist when you sell your product

• in the offline world, you give territorities for example 1 or 2 distributors in an area -> online is more difficult

Possible strategies:

• “Branded variants” strategy: when there is not clear product line, online stores list different variants i.e. online store 1 lists variant 1, 3, 6 and online store 2 lists 2,4,5

  —> usually happens in i.e. fashion, selective products

An aggregator is a business that brings together the offering of multiple business under one site or platform

• Example: price comparison website

• Example: news aggregator

• Amazon marketplace

Ex. Buying a airplane ticket: Airlines —> websites like Skyscanner aggreagte the different airlines

• Offline: Traditional store, with no online presence. Hardly exists any more

• Pure digital: Goods are discovered, chosen and processed online. Direct delivery

• owned multichannel: Two options for the customer. Not necessarily integrated or with identical offerings i.e. Gap

• online info online: Website is for information and product

  advertising. Purchased fulfilled at store i.e. Rolex

• BOPS —> “Buy online pick up in the store” i.e. Ikea

• Showrooming: Search, information and transaction occur in physical stores. Direct delivery i.e. Eyewear

  
  

• Consumers can easily search for information and compare prices, before placing an order. Best of both worlds:

• Relative to a purely online setting, they get the immediate gratification of obtaining the item in the store, and don’t need to wait for the product.

• Relative to purely physical stores, shoppers have a good overview of what products are available.

• Sales in store actually go up [puzzle!] —> people buy online, but then they are in the store and buy other stuff

• Variant: ROPS (research online, purchase in store)

• Sales in store actually go up [puzzle!] —> people buy online, but then they are in the store and buy other stuff

• Sales online went down in areas where there was a store with a BOPS policy

—> People like about BOPS tells them actually what is in the store (transparency)

• Products that require experience (touching, feeling, trying, etc.) are typically difficult to sell online. Typical solution: good system of mailing and return. (e.g. Zalando)

• Open an offline retail location without inventory

• Low investment

• Easy to involve third parties

• In the case of Warby Parker, they opened showroom

  stores. Online sales within the online sales increased by 9% in locations close to the stores

• Showrooms overcome limitations of the online channel in selling products with uncertain fit, yet maintain the pooling benefits of centralized fulfilment

• Wararby Parker showrooms opened in 20 locations. The researchers collect online sales data for 823 postal code

  ine Showrooms in Omni-Channel Retail areas (ZIP codes) and compare them to the sales in similar postal code areas that did not see a showroom

  Results

  —> total sales go up, web sales go up, home-try-on goes down

  
  

Single Channel:

• Eitherpure online or pure offline

Multichannel:

• Online and Offline options available

• No connection between the channels

Cross-channel:

• Onlineand Offline options available

• Channel are integrated and probably share operations

• Somewhat consistent in terms of image

Omni-channel:

• OnlineandOffline options available

• Channelare integrated and consistent ——> online and offline channels are consistent

• Seamless customer experience along the customer journey

• Same assortment and price

  
  

Omni-channel refers to the synergetic management of the numerous available channels and customer touch-points (online and offline), in such a way that the customer journey across channels and the performance over channels is optimized.

• Seamless experience across all channels and platforms

• Focus on the whole customer journey

• Unified customer journey

• Consistency in message and visual identity

• Consistency in prices

• Tracking customer information

  
  

—> is it necessary? debabtable

today: different prices in different channels

M-Shopper: Consumers who have used their smartphones in retail environments to assist with their shopping

—> today: most / all of it is ecommerce

—> idea of m-shopper is becoming weaker, there is no distinction anymore mobile- desktop shopper

Study de Haan:

Main hypothesis: “The conversion rate is higher when consumers switch from a more mobile to a less mobile device, compared with starting out on the less mobile device. “

—> Main hypothesis confirmed. In addition, they show that: The effect is stronger for “risky” products The effect is stronger for higher priced products The effect is weaker for categories where the customer has more experience

• Chatbots and automated messages

• Augmented reality

• Payment

• Customer journey automation

• Branded apps

Any developer starting a new business (e-commerce, services, etc.) can use stripe, square or other provider to take care of payments:

• complex transactions

• billing

• subscriptions

• fraud

• analytics

• credit

Stripe:

• difficulty for small business payment cycles —> stripes offers a solution

• is a global player in comparison to other processors

• has advantaged features

• pricing is transparent

• Closed-loop systems control the whole payment system: they control the merchants that can participate and there are no links to banks.

• They obtain insights on purchases because they observe all the data

• Example: American Express

• Example: Transit cards (Oyster, Chip card)

• Example: Store cards (e.g., Starbucks) that serve also as loyalty programs

• Hard to scale

—> payment system that does not involve the bank i.e. Mensa card, money is saved on the card

• Related to the “Freemium” model

• “Zero is almost another world. The difference between two cents and one cent is small, between one cent and zero

  cents, however, enormous.” (Dan Ariely)

• Digital products are provided for free because they the digital environment enables advertising and other revenue generating features, such as data commercialization

• Free could be loss making, but necessary for creating network effects (Dropbox), barriers to entry (WhatsApp)

Cases:

—> Radiohead offered their album and asked to pay what they wished

—> due to low cost of replication of digital products

• It is price discrimination (remember, this is a good thing)

• It avoids ”menu costs” -> costs of changing prices i.e. changes prices in stores / data bases

• It leads to large number of unit sales

• It undermines competitors

• Low marginal costs

• Fair minded user § Product can be sold at different prices

• Strong relationship between buyers and sellers

• Competitive market

• Face to face interaction increases the average price paid

• Fairness increases the average price paid (e.g., large vs small company)

• Altruism matters (e.g., offer to donate)

• Make benchmarks salient

• Downplay market norms and emphasize social norms (e.g., Disney photographs)

• Emphasize the long-term

• Bitcoin is a secure ledger -> “big spreadsheet who says who sent what to whom?”; public “for everyone to see / to alter”

• Bitcoin controlled by anyone

• sending money w/o a bank

• Bitcoin users have a public key and a private key.

• The private key is combined with the message to create a “digital signature”.

• The public key, can be used to verify that the digital signature is authentic.

• Anyone can verify that the transaction came from Alice if they know her public key.

• Notice that the message is involved in the creation of the signature, so any change of the message would invalidate

  the signature.

Sufficient Funds

• Alice wants to send 5 bitcoins to Bob.

• The message must contain references to previous transaction where Alice received 5 bitcoins

• And the transactions before that until the origin of those bitcoins

• Every transaction can only be used once as input, so it cannot be used in other transactions.

• Software checks the input against an index of “unspent” transactions.

• Owning bitcoin means that there are unspent transactions that point to your name

—> but transactions are not immediate

• Transactions are grouped into blocks.

• Every 10 minutes, a new block is generated.

• Blocks are ordered in time in a “Block Chain” or simply “blockchain”

• Any node can collect unconfirmed transactions and put them into a block and submit it to the blockchain.

• If all nodes are collecting unconfirmed transactions and making blocks, who decides which block is added next?

--> block has a name, in the block is the number of the previous name included (ordered)

—> a hash can go only one way, a unique combination of numbers and letters; transforming “KG” creates a hash, “Kg” the hash is newly created (any change will completely change the hash)

—> every block contains transactions, linked to the next block

—> Digital signature

—> Transaction system controlls for sufficient funds

—> Blocks and Block chain

—> bitcoins get created by miners

= Money is any object that is accepted as payment for goods and services and repayment of debts in a given socioeconomic context.

Functions:

• Medium of exchange

• Unit of account -> how much value a product has i.e. Euros, to coun value of things

• Store of value ->

  
  

• Fungible -> one bitcoin/euro is the same amount as another bitcoin/euro

• Transferable -> able to move it around

• Divisible

• Durable

—> Theoretically it can be accepted as money: like stones, shells, gold or Euros.

—-> A little like gold: it belongs to the person who finds it.

—> A little like the Euro: it is accepted because we expected to be accepted in the future

• Money in the back is insured

• Bitcoin can be “taken”

—> almost like money

—> blockchain = is a public data base which is immutable & decentralised

• immutable. The blockchain cannot be changed. Attacks are detectable. Transactions are irrevocable.

• Reliable. (Resilient) If a node in the network fails, the network continues operating. This ensures network

  integrity and security.

• Distributed Power. (Disintermediation ; Decentralization) No single party can control the system

• Transparent. Transactions are visible to everyone.

• Private. Transactions occur between wallets without the need of the individual identity.

  (Distributed power + Transparency + Privacy) à freedom from censorship or state control.

• Value as Incentive. The stakeholders that work on the system receive a reward.

• Digital. Any digital good can be expressed as a transaction and thus into a blockchain.

• Automated

Proof-of-work mining is energy intensive

• Mining Bitcoin and other cryptocurrencies consumes large amounts of energy, which has negative environmental consequences.

• According to estimates, Bitcoin consumes as much energy as a country such as Israel or Greece (~60 TWh per year, Thailand, 200 TWhsource: digiconomist.net) We are back to 80- 100 TWh.

• The consumption of energy has been growing steadily but has stabilised.

  
  

Alternatives to proof-of-work consensus mechanism:

• Validating transactions using Proof-of-Stake. Proof-of-stake distributes mining power across the network according to the coins currently own by each node.

• Nodes with few coins cannot alter the chain

• Nodes with many coins have no incentive to alter the chain

• Bank transfers: many banks involved, everyone takes a fee

• Remittances:

• 7-12% commission

• Need for capital

• Exchange rate risk

—> fund a start up: fund a company & find investors / loan from the bank

—> instead of giving stock to an investor, they receive tokens

• Decentralized finance is a set of protocols that enable a exchange market to function without an order book or a market maker

• It uses an Automatic Market Maker (AMM)

• Transactions are sent and encoded into a blockchain (e.g., Ethereum)

  •  Rather than matching demand and supply, there is a transparent mathematical relationship between the asset reserves and the price is set according to those rules

  • No borrowing limits

  •  Incentives to have the market price match the underlying price leads to arbitrage and efficiency

“KYC refers to a process by which banks and other financial institutions collect identification and contact information from current and potential customers.” - Binance

To prevent

• fraud

• money laundering

• other illegal activities and the misuse of financial accounts

Problems with KYC:

• -  Double-recording of data

• -  Centralized by each bank

• -  Inconsistent process handling

  Benefits of using blockchain:

• +  Enhances the security of private data

• +  Eliminates the possibility of unauthorized access to data

• +  No double-recording

• +  Cost-efficiency and consistency

Information sent by sensors allowing real time monitoring

• Information about quality, location and security available

• Interconnected system provides fast, non-invasive and non-destructive analysis

  
  

  Benefits:

• Reduction of waste in the cold chain (temperature control)

• Data security guaranteed

• Reduced cost due to constant control

• Reduced downtime, lead time and stockouts

  QR-Code System to prevent Counterfeit products

•  Problem with cryptocurrencies: volatility

• Stablecoins are cryptocurrencies that are pegged to a currency, a basket of currencies, commodities,

  other cryptocurrencies, etc.

• Tether (USDT), as it is the most liquid and heavily traded of the stablecoins, with a market cap of $4.1 billion in late 2019, accounting for nearly 80 percent of the total market cap of the stablecoin market. (Lyons and Viswanath-Natr 2019)

• They require collateral

• When the USD price of the stablecoin rises above parity, investors have an incentive to deposit dollars to

  create new stablecoin tokens, and sell them in the secondary market. (Lyons and Viswanath-Natr 2019)

• Facebook’s libra (launched 2020 and closed in 2022)

• Some stable coins are controlled algorithmically. And they crash (biggest was Terra in May 2022). See the NBER paper “Anatomy of a Run: The Terra Luna Crash” https://www.nber.org/papers/w31160

• Awareness Capture the attention of a customer to a product, service, or idea. -> bit more content-heavy

• Information To convey factual information about a product, service, or idea.

• Attitude To persuade consumers to change their attitude regarding a brand.

• Call-to-action To persuade consumers to act through specific behavior.

-> when you see ads, think about whats their purpse

(search ads more cta-ads,

• Paid media -> brands pays the channel, i.e. display ads, search ads

  • Cons: low credibility, clutter

  • Pros: control, immediate

• Owned media -> brand controls the channel i.e. website, blog, twitter account

  • Cons: low credibility, low reach

  • Pros: high control, versatile

• Earned media -> customers are the channel i.e. eWOM, Buzz; most credible form

  • Cons: low control, can be negative

  • Pros: credible, cheap

-> Display (40%), one of the biggest forms

-> Search (30%)

-> Video (30%)

-> Inbound marketing: attracting and engaging customers by creating valuable content and experiences.

• should provide a benefit i.e. educational YT-chaannel

  
  

• Getting found via search (relevance + authority)

• Content creation (e.g. blogs)

• Optimize landing pages

• YouTube Channel

• Twitter Account, Facebook account, etc.

Chatbot: an interactive conversational tool that connects and supports customers that navigate a website. They are triggered by keywords, behaviors and other signals.

-> Advertorial: Advertising in the form of editorial content.

-> Native Advertising: Paid advertising in digital media that resembles the content of the site.

• Pros: Better targeting (cost effective) and less intrusion.

• Cons: Confusion, unethical, negative consequences for the publishe

Legally, it needs to be clear that the content has a commercial purpose or source. Ads should identifiable. Legal person behind should be clear.

Disclosure techniques:

• Labels (“Sponsored”, “Promoted”, “Ads”)

• Colors (background, headings)

• Design (fonts, boxes, ...)

Contract between advertiser and publisher where the publisher agrees to show a fixed number of impressions (CPMs) to a specific demographic during a specific period of time for a fixed price.

-> Revenue management:

as a publisher, how do you price it?

-> too high: nobody buys the ads

-> too low: full of ads, not enough money

—> Revenue management in digital, becomes more difficult

-> Guaranteed Delivery Contract between advertiser and publisher where the publisher agrees to show a fixed number of impressions (CPMs) to a specific demographic during a specific period of time for a fixed price.

• Advertisers wants to target a specific demographic through a particular type of website. Hundreds of variables to consider:

  • Demographics (age, gender, ...)

  • Context (location, mobile, time of day, ...)

  • Page type and dimensions

• Duration and desired number of impressions is specified

• Publisher takes the risk for number of impressions

• Contract signed in advance (often months in advance)

• Premium inventory at premium prices

• The publisher has an inventory of impressions where advertising can be placed Future inventory of impressions available for a campaign depends ont

  • traffic,

  • demographic mix,

  • ads used by other campaigns.

• Future ad space inventory for a particular demograhic determines price (demand and supply!).

• Future inventory is only known probabilistically and needs to be forecasted when accepting and writing a contract.

• Maximize revenue while keeping the quality of the user experience.

• Revenue management problem similar in nature to airlines, hotels, theatres.

• At any point in time, publisher is allocating pageviews to ads to fulfill the active contracts. Difficult

  problem with many constraints.

• Choose the right demographic and context

• How specific should the demographic description be?

• Can the demographic description be written and verified?

• Creative

  • Match to the website

  • Who is responsible?

-> probabaly 90% of todays ads

-> ad networks emerged (bigger service which collects ads from different services)

• Original aggregators, deal with small advertisers

• Sell remnant

• Categorize ads and match them to the right publisher

• Cut across multiple websites

• Hundreds!

• Exchange markets matching suppliers of impressions (publishers) to buyers of impressions (agencies)

• Second price auction

• Bulk buying also possible

• Google, Yahoo, Facebook are big players.

  
  

i.e. NYT, load the webpages that has an ad (not definite yet), browser sends message to an exchange, the exchange delivers the profile (i.e. Google, Safari, browsing history), advertising advertisings auctioning of the ad space, winner sends to the plattform, NYT displays the ads

• Ad agencies do not have the expertise to use exchanges

• Bidding process (what and how much?) is difficult, requires data and models.

• Unique access to several exchanges

• Budget and optimization

• Trading desks

-> Showing display ads to individuals that have potential interest in your product or website because they have browsed or searched for it (pixel retargeting); sometimes retargeting is used when targeting individuals that are part of a specific list, typically subscribers, followers fans (list retargeting)

• Useful for both awareness and conversion campaigns

• Many third-party providers

-> Open market: always available

-> closed market private + verything on the right side: need invitation

• Automated sells for guaranteed volume for a fixed price

• One buyer

• Premium

• Volume and prices agreed in advance

• Search engines are the gateway to the Internet

• In 2002 search engines starting selling paid

  results (“sponsored”)

• Words are “auctioned”. First auctions for key

  words Goto.com, Google.com, etc.

• Auctions were introduced by Overture in 1997

• First-price auctions do not have an equilibrium

• For one slot, the solution is a second-price auction

• For many slots, it gets complicated ...

• Two available ad slots with CTR 5% and 4% respectively

• Three bidders with valuations $20, $18, $10 per click

-> Bidding “truthfully” is not a dominant strategy in GSP auctions. In fact, no dominant strategy exists for GSP auctions.

Measure probability of click (= quality of ad)

Should you show ads from eBay relative to website X? ctrebay = 0.001, ctrx = 0.01

Idea: (proportional to revenue for search engine)

1. Assign slots in (decreasing) order of ctri ・bid

2. Pay minimum bid needed to stay ahead:

   pi = ctri+1 bidi+1/ctri

-> Bidder pays the minimum amount that would keep its rank

—> Google creates a Quality Score; to ensure paid high quality links for users

Search Engine Optimization Maximize the visits to a website by ensuring that the website shows up high on the organic results of a search engine via webpage design, choice of keywords, outgoing links and other technical features.

On-page SEO: First and foremost, find the right keyword.

• Transactional keywords, Instructional keywords, Navigational keywords, Local keywords

• There are tools to test them § Good keywords are high volume and high difficulty (e.g,. Four words) § Keywords should relate to your business

Also on-page SEO:

• url, Title tag, Multimedia, Subheadings , Mobile

Off-page SEO: backlinks from blogs, fora, etc. Technical SEO:

• Structure and internal links

• Duplicate pages and broken links

• A cookie (a.k.a, HTTP cookie, browser cookie, etc.) is a piece of data that is saved by a browser in an user’s device and is used to remember a state, or browsing information.

•  Same-site cookies (a.k.a first-party cookie) is information saved by the webpage the user is browsing. They are used to manage a session (actions, items in a shopping cart, etc.) and to keep users logged-in. They are not accessible by other websites.

• Third-party cookies are not set by the website but by an adtech platform loaded onto the website. These cookies are accessible by other websites and are used to target advertising for products based on recent browsing and other identifiers.

• Cookie blocking refers to third-party cookies.

• Social media buttons generate third-party cookies

• Re-targeting relies on third-party cookies

• Most pop-ups that you see when you access a website relate somehow to cookies and similar tech

Ghost ads is a technology to predict who in the control group would have seen the focal ad but did not. It uses a “simulated” assignment of ads before sending the ads to the programmatic advertising exchange. The result is a control group that matches the characteristics of the treated group.

• Solves the problem of biased selection

• It is more precise than the intent-to-treat approaches

• It provides the right baseline

This setting is ideal and we could obtain the treatment effect by loooking at the groups in the top half of the figure: those who said the ad, vs. those would have seen the ad but did not

Intent-to-treat: Consumers are assigned to t conditions and we compare the two conditions irrespective of whether they saw the ad. Users who do not see the ad do not contribute to the measurement and only add noise

(multi-touch) Attribution Determining the causal impact of the different communication channels, events, touchpoints, or media on customer responses (e.g. conversion, visits, etc.)

Wearout: The decrease in effectiveness of additional ads with repeated exposures. (Opportunity cost)

Weariness: Additional advertising exposure has a significant negative effect

(Business loss)

• Browsing behavior

• Reach and frequency

• Changes in browsing behavior

• Stop if conversion

eWOM as “any positive or negative statement made by potential, actual, or former customers about a product or company, which is made available to a multitude of people and institutions via the Internet.” (Hennig-Thurau et al., 2004)

Social Media as websites and internet-based applications that enable users to create and share content or to participate in social networking.

1. Dispersion

2. Enhanced volume

3. Persistence and observability

4. Salience of valence

5. Community engagement

6. Anonymity

= Dispersion: The extent to which product-related conversations are taking place across a broad range of communities (Godes & Mayzlin, 2004)

• nature of platforms influence appearance and dynamics of eWOM

•  Written word remains in public repository and is persistent

• Message content and source becomes salient in evaluations of eWOM

• Past eWOM impacts future eWOM

• eWOM is endogenous: eWOM influences consumer purchase behavior and is at the same time the outcome of consumer purchases.

• WOM can be misinterpreted as it is based on interaction between individuals

• eWOM often has assigend numerical rating

• Community engagement:

  -> platforms support specialized, non-geographically bound consumer communities

  -> vent frustration or learn from other consumers how to better use products/services

• Anonymity:

  -> self-interested behavior can reduce credibility and informativeness

  -> deceptive company behavior

  -> eWOM quality increasingly important (e.g. verified purchase reviews)

• §  Online opinions reflect vocal minority

• §  Under-reporting bias: opinions are more extreme

• §  Rich-get-richer phenomenon

• §  Self-selection: more experienced or those with high desire for being different avoid following the minority, less experienced go with popular opinion

  
  

Receivers infer credibility and quality of information from

•  prior activity of sender

•  two-sidedness/valence (moderate) of eWOM

• high response speed of sender

•  previous reviews of sender positively evaluated

• depth (providing valuable information of specific domain) for receivers with decision-making orientation

• breadth (comments across various domains) for consumers in learning mode

  
  

  
  

  Heuristics:

• Negative reviews have greater influence when many reviews are positive

• -  But sometimes negative reviews have less informational value as they convey individual emotion

• -  Variance of ratings: low rating products perceived as higher quality when variance is high

• Social capital and reputation (of communicators)

• Review quality and online citizenship (of communicators)

• Product and brand knowledge

• Customer engagement

• Purchase intention

• Product sales (volume, revenues, stock prices)

• Price premium and product spending § Trust and loyalty

Platforms:

•  Independent third party is more influential than seller website

• eWOM on social media platforms is stronger when eWOM receivers can assess their own similarity to eWOM senders

Products:

• Stronger effects for tangible goods new to the market,

• for high involvement products,

• and privately consumed, low-trialability products in less competitive industries

• The credibility of reviews can be undermined by businesses leaving fake reviews for themselves and the competition

• Some business hire people in Amazon Mechanical Turk to post 5 star reviews (0.25$ per review)

• Luca and Zervas (2016) analyze Yelp data and find out that fake reviews are prevalent in low reputations resturants

• Restaurants with more reviews are less likely to receive a fake review

• Increased competition increases the likelihood of receiving a negative review. Entrants for different cuisines has no effect

• Respond to complaints and negative eWOM

• But be aware that service interventions can encourage further complaints

• Improved targeting based on customer‘s underlying relationship to firm § Possible benefits of motivating consumers and create space for their opinions

• But should firms actively participate in eWOM?

• Rather engage in conversations that address functional needs such as solving problems or offering technical advice

Influencer as individual or group who built their own audience through social media platforms and influence others.

Influencer marketing as using social media influencers as a communication channel in the marketing mix.

• §  define target audience, e.g. broad mass vs. experts

• §  aims of campaigns, e.g. brand awareness vs.

  specific information

• §  communication message

• §  focus on domain breadth or social presence?

• §  which of the four influencer types?

• §  identify influencers of this type

• §  watch dynamics

• Social media listening (SML) involves identifying and analyzing information about a target (e.g., company, product) on the internet.

• SML became popular, with companies using internal or external resources to monitor and respond to online conversations and consumer trends.

1. Social media is full of information and we should listen with a purpose, either hypothesis-driven or with a structured approach to knowledge discovery

2. Social media listening is important for new products and for marketing mix decisions

3. Text mining is still bringing new insights and making new sources of data available

4. Search trends hide information about future demand

5. You cannot tell a fad from a trend, you can only make educated guesses

• Different opinions and discussions were found in the popular press, blogs, and online forums regarding hair trends.

• Trends like ombre, tie-dye, and rainbow hair gained attention and had both supporters and critics.

• Text mining techniques can be used to understand consumers' associative networks and market structure based on online discussions.

• Semantic relationships between brands can be identified using metrics like lift, which measures the strength of association between two brands.

A customer journey is the series of actions a customer takes to arrive at the moment of purchase, and the cognitive and emotional states along those steps (Lemon and Verhoef 2016, Schmitt 1999).

Customer experience refers to the multidimensional construct that involves sensory, physical, cognitive and emotional responses from the customers during the interactions with a company or brand (adapted from Lemon and Verhoef 2016).

• Awareness

• Interest

• Consideration

• Purchase

• Experience

• Post Exerpience

• Awareness has to do with all the underlying knowledge that customers has about their needs and available offerings. We can further distinguish between problem awareness, company awareness, brand awareness, product awareness, feature awareness.

• Interest refers to an active state of search for alternatives, research of pros and cons, and specific distinguishing features. It may also involve planning about the purchase or scheduling.

• Consideration is the stage where a set of competing offerings have been identified and they are evaluated and compared. The consideration set is not necessarily fixed, but in most cases the chosen alternative is already included at this stage. Around this stage there could be negotiations and quote requests.

• Purchase is the moment where the customer chooses one product alternative. In some cases, it may include some final negotiation, in others it may involve third-party channel partners, etc.

• The Experience stage encompasses the customer experience with the product or service. In the case of services, the experience could involve a rich interaction with the providing organization.

• We use the term Post-Experience to include all actions that occur as a result of the purchase and experience of the product. These could be “follow-up” interactions, customer advocacy and recommendations, and the establishing of repurchase relationships.

• Questions: What does the customer need to know?

• Customer Actions: What are customers doing at this point?

• Experience: What are the most important aspects of the customer experience

• Touchpoints: How are customers interacting with us? Individuals, media, third-party

• Information: What kind of information is reaching the customer?

• Responsibility: Who is in charge of the customer journey at this point?

• Firm actions: What are we doing?

• Bottlenecks: What is stopping /delaying the customer to continue with the journey?

• Metrics: Awareness

• Awareness:

  • Problem -> Journey starts here; make people understand they have problem

  • Brand

• Interest & Consideration -> in touch w/ expert

• 
  

Hedonic-utility scale

Channels

Early to late process

Conversion rate

• Consumers making hedonic purchases tend to utilize social media and product page views on the target retailer’s website more extensively.

•  Consumers making utilitarian purchases tend to use search engines, third-party reviews, deal sites, and product page views on the competing retailer’s website more frequently

• The H/U effect on the usage of social media and third-party review sites tends to be stronger earlier in the customer journey. Conversely, the effect of H/U differences on search engines and deal site usage is stronger closer to the point of purchase.

• Social and search is always high

• Product page becomes more important later in the process

Content marketing is the creation of valuable content to entertain, inform, educate and inspire the target audience, in a way that creates engagement, attracts and retains customers and ultimately drive sales.

• Notice that content marketing is not new to digital and brands have been providing content for a long time, especially in B2B

• Original content marketing was used to create awareness and increase usage

• In the digital domain, content can be distributed at very low cost, it provides value that customers seek within a cluttered advertising landscape,

• It can be distributed at low cost

• It provides intrinsic consumption value that make sought-after in a cluttered advertising landscape

• It can fulfil different roles along the customer journey

• It can be targeted

• It can be shared

• It can be created by the customers/users themselves

Educate (High involvement, functional) i.e. White Paper, E-books

• The goal here is to explain the benefits of a product and how it links to the customers’ current and future needs

• Appeals to rationality

• Accurate

• It requires the customers’ attention

  and interest

Inspire (high involment, emtional/ hedonic) i.e. Branded video

• The goal is to connect with the customer at an emotional level

• Content to move and persuade

• Empathy, authenticity, recognition and community are important factors

Inform (low involvement, functional) i.e. Infographics, Ratings

• Offers objective information about products and prices

• It allows customers to consider and compare alternatives

• It summarises information and explains basic points

Entertain (low involvment, emotional hedonic) i.e. Games, Quizzes

• The goal is to create awareness and build trust, turn visitors into fans

• Typically humorous posts that appeal on a humorous basis.

• Highly shareable content that can connect with a target audience.

• There is a strong focus on authenticity and visual stimuli

1. AirBnB

2. Booking.com

Engagement in digital contexts refer to interactions of the user with a particular page or account.

Outcome Metrics refer to a market or financial result and measure current performance. Typical outcome metrics are revenue, revenue growth, gross margin, market share, average prices, etc.

Impact Metrics link outcome metrics to marketing effort

Process Metrics refer to variables that measure the performance of the organization’s processes or the customer purchase process. These metrics should be causally linked to outcome metrics and used to predict future market performance. Typical process metrics are brand awareness, customer satisfaction, purchase intention, recommendation intention, etc

Brand Value refers to estimates of the value of a brand or the value of customers

Customer Value refers to estimates of the value of a brand or the value of customers

‘consent’ of the data subject means any freely given, specific, informed and unambiguous indication of the data subject’s wishes by which he or she, by a statement or by a clear affirmative action, signifies agreement to the processing of personal data relating to him or her;

Data subjects must be given a real choice and control.

• Subjects must be able to freely refuse processing without fear of detriment: the subject believes he or she will not face negative consequences.

• Consent cannot be bundled with other provision of a contract, in fact disguising data processing that is not necessary for fulfilling the contract (conditionality).

• Consent should be given separately for each purpose of the data processing and the extent of the processing (granularity)

• The reason for the data processing must be specific, unambiguous and concise.

• Ensure that consent is informed; subjects must understand what they are agreeing to.

The following rights are highlighted by the law

• Right of access: know whether data is being processed and obtain it partially or fully

• Right to portability: receive the personal data in a structured commonly used machine-readable format

• Right to be forgotten: erase all personal data if there are no legitimate grounds to keep it

• Right to data breach notification: to be informed within a short time frame if there are security problems, or the risk of a security problem, with their personal data.

GDPR reduces the number of vendors in the short run and increases concentration

• GDPR affects recorded outcomes via (1) user consent, (2) limiting marketing ability, and (3) privacy frictions.

• Result: GDPR reduces recorded pageviews and e-commerce revenue by 12%.

• Larger drops in users arriving from personalized marketing channels that are most directly impacted by the GDPR.

• Under weak assumptions, the real effect of the GDPR reduced real website pageviews by 0.4% and real e-commerce

  revenue by 0.6%

Privacy concerns: individual‘s beliefs about the risks and potential negative consequences associated with sharing information.

Privacy management: individual‘s adoption of privacy protective measures.

Barth and de Jong (2017): Users are concerned about their privacy but undertake little to protect their data. • risk-benefit

• biased risk assessment • no risk assessment

• knowledge deficiency

• benefit >> risk

• failed evaluation of risk

  Baruh, Secinti and Cemalcilar (2017): Using a meta- analysis find little evidence either way. But there is some evidence that users with strong privacy concerns share less information

Privacy preference is the preference to control information about oneself

• Intrinsic privacy preference: Privacy is a good in itself, an

  element of human dignity and autonomy (moral /

  philosophical view). Privacy is a final good.

• Instrumental privacy preference: privacy is necessary to

  protect one’s private information in market exchanges. Privacy is an intermediate good.

Algorithmic Bias occurs when an algorithm using individual or group-level information produces outcomes that are disadvantage (or privilege) a certain group of people based on gender, race, class, sexual orientation, religion or age, therefore discriminating against that group. Algorithmic bias can occur even situations where the group identity (e.g., gender) is not included in the algorithm.

Promotion

• Campaign targeting groups with specific characteristics

• Example from Lambrecht and Tucker (2018). Career ads for STEM jobs were shown on Facebook. The ads did not stipulate gender, but

  ended up being biased and shown more to male users. Why?

• Price

• Price discrimination is a prevalent practices and in most cases legal

• Example: Orbitz

• Example: Uber

  Product

• Facial recognition

• Apple credit card

• Epistemic bubbles and content

  Distribution

• Amazon same-day delivery

Causes

• Historical biases

• Incomplete or unrepresentative data

• Characteristics that interact with the algorithmSolutions

• Audits and bias detection

• Diverse teams

• Including fairness in the optimization

• Collect sensitive protected data

There are three main barriers to automation (computerisation)

1. Social Intelligence: Ability to persuade, care and negotiate.

2. Creativity: The ability to find or invent novel and valuable ideas and artifacts

3. Perception and Manipulation: Robots cannot match human perception and manipulation of objects in

highly unstructured environments.

• Technical feasibility (innovation, integration, adaptation)

• Cost of developing solutions (development + deployment; software + hardware)

• Labor market dynamics

• Economic benefits (cost savings? Increase production speed? Safety?)

• Regulatory acceptance