1. Integrate 2015:
Big Data: Big Investment Opportunities
September 2015
Glenn Solomon
Managing Partner, GGV Capital
2. Table of Contents
Why Has Big Data Become a Big Target for VC
How We Look at the Big Data Marketplace
Where are VCs Investing
How is Big Data Being Used in Companies
Where Do We See the Opportunities
Big Data Risks and Opportunities
4. Why Has Big Data Become a Big Target for VC
Source: Domo, āData Never Sleeps 3.0ā
We are surrounded by a
wealth of data we create
from our everyday activities
5. Why Has Big Data Become a Big Target for VC
Source: BI Intelligence
Explosion of IoT
Mobile Devices
By 2016, IoT > Mobile + PC
combined
IoT devices are generating
billions of gigabytes of data
everyday
6. Big Data is Driving Value for Organizations across Different Areas
Source: Gartner 2014
Increasing Number of Organizations are
Investing in Big Data
Across Different Areas, >50% of
Respondents See the Value of Big Data
8. How We Look at the Big Data Marketplace
Storing Data
Making Decisions with Data
Analyzing Data
Vertical Market Uses of Data
Sales &
Marketing
Security
Finance
Advertising
Healthcare
Retail &
Supply Chain
13. How is Big Data Being Used in Companies
With huge user base and long hours of streaming,
Netflix can collect data from everyone on viewing
patterns:
ā¢ Time spent on shows / movies, and location
ā¢ Types of entertainment (documentary, comedy,
drama, horror, etc.)
ā¢ Favorite starring cast
ā¢ When users stop watching a show, etc.
Using Big Data to Create a Show People Will Like
62.3 million total Netflix users
people watch Netflix for ~ 90 minutes per day
Big
Data
Started in 1997 as a pay-
per-rental via mail company
Now has evolved into an on-
demand streaming service with
thousands of movies and TV shows
ā¢ Engage users better on current shows
ā¢ Recommend shows they may like
ā¢ And create a show people will like!
Directed by David Fincher Featuring Kevin Spacey
With Big Data, 70% of Netflix original shows are renewed for
second season, compared to 30% from traditional TV series
Source: Company website
14. How is Big Data Being Used in Companies (contād)
Source: Company website; āData Jujitsu: The Art of Turning Data into Productā
āPeople you may knowā
Asking a set of questions such as:
ā¢ āwhat do you doā
ā¢ āwhere do you liveā
ā¢ āwhere did you go to schoolā
ā¢ using friendās friend connections
ļ More friendly faces up front to keep
users engaged with the product
Collecting Data from Users
āWhoās viewed your profileā
ā¢ By giving data back to users,
LinkedIn creates a more engaging
experience for users
ā¢ Brings more revenue and make it
more profitable for both users and
company
Feeding Data Back to Users
15. How is Big Data Being Used in Companies (contād)
Disney created a big data platform to store, process,
analyze and visualize all data that is generated through
the MyMagic+ system
Disney MagicBands and MyMagic+ System
Disney collects tons of valuable data through MagicBands
and MyMagic+ system - a gigantic database that captures
every move of the visitors of the park
ā¢ Real-time location data
ā¢ Purchase history
ā¢ Information about the visitors
ā¢ Entertainment ride patterns, etc.
Insights from big data enables Disney to make smarter
decisions:
ā¢ Audience analysis & segmentation
ā¢ Recommendation engine based on in-park traffic flow
ā¢ Better and targeted marketing messages and offerings
ā¢ And many moreā¦The MagicBands (part of MyMagic+ System) are linked to credit
card and function as a park entry pass as well as a room key
Source: Company website
16. How is Big Data Being Used in Companies (contād)
Solution:
The Black Book model, which analyzes up to
1 quintillion decision variables and combines
various data sets such as:
ā¢ satellite imagery
ā¢ weather data
ā¢ expected crop yields
ā¢ acidity or sweetness rates
ā¢ regional consumer preferences
ā¢ 600 different flavors profiles of an orange
Results:
ā¢ Precise & dynamic formula on how to blend
orange juice for consistent taste, down to
pulp content, for the $2Bn orange juice
business
ā¢ After hurricane or freeze, this algorithm can
re-plan the business in 5-10 minutes
Problem:
Inconsistencies in orange juice due to
variations in orange crop, sourcing, and
seasonality, etc.
Goal:
Consistently deliver optimal blend of orange
juice, ādespite the whims of Mother Natureā
Orange Juice and the āBlack Book Modelā
Source: Company website
17. And Not Just Companiesā¦Even Municipalities Benefit
The use of big data has brought the following
benefits:
ā¢ Identify fire hazards based on algorithm
ā¢ Reduce the number of fires
ā¢ Fires are less severe as a result
ā¢ Save on personnel and firefighting resources
New York Fire Department has captured 60
different factors that could contribute to the
likeliness of having a fire, such as:
ā¢ Average neighborhood income
ā¢ Age of the building
ā¢ Whether it has electrical issues
ā¢ Number and location of sprinklers
ā¢ Presence of elevators
ā¢ Each one of the cityās 330,000 buildings is
ranked in order of the risk of fire
ā¢ New York Fire Department uses the risk score to
determine which buildings get inspected first
Present
ā¢ Inspections were almost random except for
high-priority buildings like schools and libraries
Past
Big Data
Source: Company website
Risk
Score
19. Where Do We See the Opportunities
ā¢ Targeting sophisticated data analysts on data-driven
teams
ā¢ Connects directly to databases
ā¢ Fast, customizable visualization, easy collaboration,
and superior SQL editing experience
ā¢ Business management platform
ā¢ Focuses on the needs of the decision-makers in a
business, as opposed to existing data management
procedures and policies
ā¢ Connect, Prepare, Visualize, Engage and Optimize
Tools for Data ScientistsTools for Business Executives
Disclosure: GGV is an investor in Domo
20. Where Do We See the Opportunities (contād)
Readying Massive Data Intelligently
ā¢ USM (Unified Security Management) that provides
comprehensive, centralized and affordable security
visibility
ā¢ Combines log management and SIEM with other
security features for complete security monitoring
ā¢ Single platform, easy to use and deploy, perfect fit for
mid-market enterprises
Solving Big Problems ā e.g. Security
Disclosure: GGV is an investor in Alienvault
ā¢ Curates massive variety of internal and external data
ā¢ Reduces time and effort required for analytics and
other applications critical for business growth
ā¢ Leverages machine learning algorithms to identify data
sources, understand the relationships between them,
and connects siloed data
21. Where Do We See the Opportunities (contād)
Data Scientist as a ServiceNuanced and Unstructured Data -> Insights
ā¢ Provides actionable insights, not more dashboard
reports
ā¢ Helps companies quickly understand what they need to
do based on the data shown, so companies can spend
less time analyzing and more time implementing
ā¢ Highly trained on-demand team of Data Scientists
backed by powerful tools
ā¢ Captures and analyzes feedback from social media, blogs,
forums, surveys, etc. to attain deep understanding of
customer and marketplace feedback
ā¢ Big data ļ big insights, helping companies understand
how customers feel by deriving meaning from the most
unstructured, unpredictable, and nuanced and subtlest
context, so they can take action with maximum impact
23. Big Data, Big Risks and Even Bigger Opportunities
ā¢ Trade-off between privacy / security and the benefits of
wider pool of data
ā¢ Behavioral data collected has more direct impact to the
end consumers, and can lead to more sophisticated
attacks on targeted consumers
ā¢ As information becomes more readily accessible across
sectors, it can threaten companies that have relied on
proprietary data as a competitive asset
ā¢ Companies that have benefited from information
asymmetries are prone to disruption
Information Asymmetries to be Disrupted
Privacy / Security vs. Benefits of Data
24. Big Data, Big Risks and Even Bigger Opportunities (contād)
ā¢ The purpose of collecting data is to better serve customers not
the other way around
ā¢ Consumers are becoming more aware of the value of their data
and less willing to give sensitive data for free
ā¢ To turn data annoyance into empowerment, companies should
engage users, enlist their help, give them control, and even
reward them with data / insights they like to see in return
ā¢ Getting the exact result vs. having a good set of options
ā¢ If data is presented to users directly, such as search engine,
should aim to maximize precision
ā¢ In the case of ads where the relationship between ads and your
interest is obfuscated, can compromise on precision to achieve
broader optionality
Customers Come First, Data Second
Tradeoff between Precision & Optionality
25. ā¢ More is not always better - more data can lead to more
data quality issues, confusion and lack of consistency in
business decision making, especially with conflicting data
ā¢ The challenge of getting the right information to the right
person at the right time is expanded due to the sheer size
of big data
ā¢ Storage is relatively cheap, and the technology to
process data is available on demand
ā¢ But what about people and skills? Having the right
people and right skills to analyze and take action on the
data is the new big challenge
Data++ = Confusion++ and Consistency--
People and Skills are the New Challenge
Big Data, Big Risks and Even Bigger Opportunities (contād)
Hereās an infographic from one of our portfolio companies ā Domo, on how much data we generate every minute of the day.
And the driving forces behind the wealth of data we generate, are the mobile devices, which is 2x of PC by 2019, and the explosion of IoT, which is 10x of PC and 5x of mobile by 2019.
Overall, companies are increasingly investing in Big Data, as can be seen from the left chart. The right chart shows where different industries stand in terms of their readiness to invest in big data. Across these industries, more than 50% of the respondents are seeing the value of big data.
This is how we look at the Big Data marketplace: companies that store data, companies that analyze data, companies that help you make decisions with data, and vertical uses of data, such as sale and marketing, advertising, security, retail, finance, and healthcare.
Most of the companies from this group are in the storage space.
Companies in the group are a combination of analyzing data, making decisions with data, and vertical uses of data.
Netflix has used the data they harnessed around viewers patterns to predict that a show directed by David Fincher featuring Kevin Spacey would be a big hit ā which gave birth to House of Cards!
With big data predictive analysis, Netflix has 70% success rate renewing shows for second season, versus 30% from traditional TV series.
Remember to bring up:
Stitch Fix (apparel store for women ran by 50 data scientists)
Opendoor and daughter wanting to be a data scientist
LinkedIn collects data from users to recommend people you may know as a way to lower the friction for user acquisition. They are also feeding some data they collected back to users, such as the āwhoās viewed your profileā feature, to create a more engaging experience for users.
Disney has made a huge push on the MyMagic+ system, which helps Disney understand visitors better, such as location data and purchase behavior. As a result they are able to manage in-park traffic flow and target marketing more efficiently.
Coca Cola and their āBlack Book Modelā has enabled consistent quality of orange juice despite the whims of Mother Nature. It takes into account 1 quintillion decision variables, and using precise and dynamic formula to decide where to source and how to blend for the best quality juice.
Munis are also benefiting from big data ā taking the New York Fire Department as an example, through an algorithm of 60 different factors, they calculate a risk score for the cityās 330k buildings, and rank them accordingly for more frequent inspection and more firefighting resources.
The blue bars are where we see the opportunities, and the companies underneath are examples in those spaces.
Domo ā business management platform for decision makers
Periscope ā sophisticated tools targeting data scientists and data-driven teams
The blue bars are where we see the opportunities, and the companies underneath are examples in those spaces.
AlienVault ā single platform that combines log management and SIEM, easy to use and deploy for mid-market enterprises
Tamr ā leverages machine learning to connect and curate massive variety of data, saving data scientists 80% of their time previously spent on prepping the data into usable format
The blue bars are where we see the opportunities, and the companies underneath are examples in those spaces.
Luminoso ā specializes in deriving insights from the most unstructured, unpredictable and nuance context - customer feedback
DataScience ā brings a highly qualified and on-demand team of data scientists to you
We are constantly battling between the benefits big data brings and the privacy and security concerns ā especially as data is getting more personal and behavioral
Data wants to be more open, but companies that have benefited from proprietary information are pushing back
80/20 rule for tradeoff between precision and optionality
Although we are talking about big data today, itās helpful to take a step back and emphasizing that data is here to better understand and better serve customers
Storage and tech to process data are readily available to us; people and skills are the new big challenge to derive insights from data
More data can potentially lead to more confusion and conflicting decisions, so getting the right information to the right person at the right time is crucial