Online-Offline Interaction & How to Find Lead Users and Facilitate Influence and Contagion
Why Offline is Still Important
Even though the Internet “liberates” consumers, there are still some “disutility costs” associated with buying online as more than 90% of what is sold to consumers is still sold offline.
This is because there are still frictions:
Uncertainty about “fit and feel” of certain products.
Cost of returns (time and money).
Competition is intense for popular products, but almost non-existent for niche products
For example, when an offline bookstore opens in a neighborhood, Amazon.com sales in that neighborhood decline; but only for popular books like Harry Potter and the Sorcerer’s Stone and not for niche books like 101 Years of All Black Trivia.
Friction Number 1: Search Friction
1987 Rugby World Cup … where do I buy a TV?
Who will have the best price and assortment?
2013 .. go to http://milo.com/
Friction Number 2: Geographic Friction
New York City versus Iowa City.
The location that you live in imposes certain costs and delivers particular benefits.
The Internet can deliver you from the “tyranny” of your local options.
The Long Tail
Q1: What is range of quality/satisfaction?
Q2: What does this imply for filtering?
Historically, we lived in a world of “hits” (and not just for content), but now live in a world of infinite slots.
The economics of distribution have been radically altered for content, and for product as well.
The Long Tail exists because the economics of storage and distribution have changed; this is the supply side.
The Long Tail is itself endogenous (demand side) as there are more ways for us to discover variety.
1897 Vilfredo Pareto found 20% of population owned 80% of the wealth (in the UK) giving rise to the Law of the Vital Few.
1949 George Zipf found that second most common word used ½ as much as the first; third ⅓ as much, etc.
The tyranny of locality (relate this idea to the concept of preference minorities).
An audience that is spread too thin geographically used to not be served, e.g., Lagaan: Once Upon a Time in India opened on only two screens.
The ratio of niche products to “hits” is changing.
Distribution efficiency is amplifying (digital, search, etc.).
Recommendations and reviews drive search so that consumers get into the tail.
Collective value of niches > hits.
“A Long Tail is just culturally unfiltered by scarcity”.
Disentangling the supply side (larger product availability) from the demand side (easier for consumers to find and sample new products).
MIT study obtained data from a retailer offering identical product selection and price through two different channels: Internet and catalog.
Descriptive evidence is that Internet sales are significantly less concentrated (eve after controlling for differences in customers and supply side factors such as taxes and shipping).
The results hold up when differences in the customer groups are controlled for too (Internet customers are younger, have more income).
The Internet channel has more evenly distributed sales and the difference cannot be attributed to differences in price or availability.
Unite sales of “niche” products: 14.8% v 12.7%
$ sales of “niche” products: 15.0% v 12.7%
(Differences are statistically significant)
The question is now whether there is a demand side explanation, e.g., customer use of search tools on the Internet.
Key variables to explain sales of niche products:
All computed as percentages in terms of page views. e.g. , Directed Search Views / Total Views.
Critiques and Summary
Law of natural monopoly (hits get disproportionate share of light users).
Law of double jeopardy (unfamiliar things are less well liked).
But the Long Tail is still a very powerful concept!
Million Short: http://tcrn.ch/SPIJ9B
Red Ties Versus Blue Ties
Imagine that you live in a town where everyone wears ties all the time. Then, imagine that you’d like a blue tie, but everyone else wants a red one.
Will you get what you want?
“Isolated” Prospects are Worth Pursuing
Preference Isolation brings shoppers online and explains geographic breakdown of online brand demand.
Unit volumes of different brands in zip code z in MSA m is an integer count, regress on demographics and Preference Minority Index
b value significant and > 0 for Preference Minority Index for all brands but largest effect is for niche brands.
Compare 90th percentile vs. 10th percentile PMI.
Category sales 50% higher.
Niche brand sales 125% higher
How Internet Retailing Startups Grow
Work with participating Internet Retailers
Gather sales information from inception
Customer ID, Date, Transaction Value, Zip Code
Geo-demographic “real world” data
Why do some locations have more customers than other locations do?
Customer Acceptance of Online Retail Depends on Offline Shopping Costs
Internet retailers can alter the cost-benefit trade-off shoppers by making things “closer” and more accessible, perhaps even at better prices.
Sales Evolution is Structured and Predictable.
Social Contagion from communication and observation affects online demand evolution.
Migration from “Good” to “Great” Requires Expansion to Niche Locations
Spatial Structure follows a pattern of proximity and similarity (spatial “Long Tail”).
There are two ways to think about distance:
Social, or demographic distance (people live far apart from each other still might be similar in other ways)
Two important patterns:
In the beginning, sales start out in larger cities and spread by proximity from person to person.
Later on, sales pick up in smaller areas that are quite far apart, but that contains “similar” kinds of people.
Customers and Digital Marketing
Goals and Tactics
Customer goals: Attract, engage, retain subject to:
Never pay more to acquire than you will recoup, i.e., Customer Lifetime Value (CLV) > Acquire Customer (AC)
CLV needs to incorporate “Referral Lifetime Value (RLV)”.
Key tactic: Encourage customers to refer and acquire other customers (e.g. www.Diapers.com).
[Attractive target customer]
Monologue to conversation
Amplification through virtual and real world synergy
“Long tail” leverage
[Selection and Treatment]
Marketing “Spend” as an Asset
Influence and How Information Spreads
“Pathways through which information, advice, resources, and support flow between people” (Aral 2012).
Networks can be physical or virtual.
Network usually exhibit homophily, either in characteristics of participants, or in preferences.
A “network” can be a simple as a dyas, e.g., two partners, or so complex as to encompass hundreds or thousands of people.
Connections (between people)
Ability to share information and resources
Constraints (geography, socioeconomics)
[More Background: http://www.youtube.com/watch?v=2U-tOghblfE]
Entering a Network
Participation in a network is a choice (often governed by some form of homophily).
We also decide how many people we want to connect to and how central we want to be.
When our relationships are transitive (our friends know each other) then we are deeply embedded.
Being Affected by a Network
Even “strangers” and loose connections in a network can affect us (Milgram 1968; 1 person = 4% influence, 15 people = 40% influence).
Separation and Influence
1960s: Few hundred business people send letters from Nebraska to Boston
2002: 98,000 people send emails to unknown targets.
Six degrees of separation
Three degrees of influence (decay, instability). The influence spread from you typically won’t go further than three steps.