mustafa degerli - 2017 - technology entrepreneurship and lean startups - aardvark case
TRANSCRIPT
TECHNOLOGY ENTREPRENEURSHIP AND LEAN STARTUPS
Mustafa Değerli
2017
Mustafa Değerli – Aardvark Case Page 1 of 16
Table of Contents
1. The Business Model Canvas – Aardvark .................................................................................... 2
2. Financial Predictions, Revenue Model, and Profitability – Aardvark ........................................ 3
3. Ideation: Compare and Contrast – Aardvark & Dropbox ........................................................... 5
4. Vision vs. Experimentation – Aardvark...................................................................................... 6
5. Value Innovation – Aardvark...................................................................................................... 8
6. Multi-sided Markets – Aardvark ................................................................................................. 9
7. Lean Startup Methodology: Pros and (possible) Cons – Aardvark .......................................... 10
8. Lean Startup Methodology: Compare and Contrast – Aardvark & Dropbox ........................... 14
9. The Mechanical Turk – Aardvark ............................................................................................. 15
10. Dropbox, Aardvark, and ULAŞIM ......................................................................................... 16
Mustafa Değerli – Aardvark Case Page 2 of 16
1. The Business Model Canvas – Aardvark
The Aardvark’s business model canvas that I sketched out for the Aardvark is provided at the link:
https://bmfiddle.com/f/#/Fp6V8
Screenshot of the canvas that I sketched out for the Aardvark is as follows:
Mustafa Değerli – Aardvark Case Page 3 of 16
2. Financial Predictions, Revenue Model, and Profitability – Aardvark
The Aardvark’s revenue model is mostly based on generating advertising revenue. The Aardvark
hypothesized that generating leads and inserting affiliate links via high commercial potential
queries will generate enough monetization.
Principally, the Aardvark intended to generate advertising inventory in two main conducts. First,
sponsored links; those can be shown to users when they submitted queries. Latter, the Aardvark
might add links to responses and can receive compensation provided that a user clicked on the link
or purchased an item based on the Aardvark suggestion. Additionally, the Aardvark intended to
depend on third parties on the way to manage advertising inventory in the first year or two, and
then planned to develop its own ad sales team. In late 2009, the Aardvark forecasted revenues of
less than $1 million in 2010, assuming cost-per-thousand impressions advertising rates of $13 and
millions of users by year-end. To become a breakout success, the founders projected that the
Aardvark would have to reach users in the tens of millions with $35 cost-per-thousand impressions
in 2012. At those rates, with 60 million users, they estimated they could reach $5 million of
operating income that year on revenues exceeding $30 million.
With regard to Exhibit 7 and 8 given in the article of “Aardvark” by Thomas Eisenmann, Alison
Berkley Wagonfeld, and Lauren Barley, the Aardvark’s revenue and profit predictions for the
future were quite attainable. However, as of today, there is no financial information available for
the Aardvark on the web and even the Aardvark does not exist anymore. Therefore, we are not
able to exactly know whether the Aardvark is profitable today or not. On the other hand, we might
conclude that as they do not exist anymore, they might have become not profitable. Another
possibility may be that the Aardvark is purchased by an outside organization that lead to their end
as Aardvark. That is, in practice, the Aardvark may exist in the form of a group or unit in a new
organization with a different name.
The viability of the Aardvark’s financial model can be judged as moderately promising and
probable in theory, yet we don’t have real data to analyze the actual case. As I noted already, we
can conclude that regrading financial predictions, revenue model, and profitability, they seem
unsuccessful since they do not exist anymore. If it was not the case that the Aardvark was
Mustafa Değerli – Aardvark Case Page 4 of 16
purchased and integrated into a new organization, we can simply say that one probable reason for
the Aardvark’s failure is their failure in financial predictions, revenue model, and profitability. To
me, the assumptions related to financial details given in Exhibit 8 given in the article of “Aardvark”
by Thomas Eisenmann, Alison Berkley Wagonfeld, and Lauren Barley are reasonably reasonable.
On the other hand, it is still another time we appreciate that in theory there is no difference between
the theory and practice, but in practice there is. There are “other” factors and changing dynamics
that may truly shape our realities. Therefore, we need to be “agile” and have plan-b’s for whatever
the case is, including financial scenarios.
Mustafa Değerli – Aardvark Case Page 5 of 16
3. Ideation: Compare and Contrast – Aardvark & Dropbox
The ideation process for the Aardvark can be summarized as:
The Aardvark team identified the broad space in which they would search for a product
idea. They wanted to leverage the Internet to help make decisions that required human
judgment.
The Aardvark team wanted to create something that could be tested with users. They also
wanted to evaluate ideas quickly, so they needed concepts that required relatively little
programming before they could test them.
The Aardvark team’s goal for any idea was to have a prototype that could be user tested
within a month. After the Aardvark team saw what users thought, they could make
corrections or discard the idea.
The Aardvark team didn’t want to mine knowledge on the Web. They wanted to mine
knowledge in people’s heads. And how do people get into their heads? Through sentiment;
people are going to participate because people want to help other people.
The team explored many ideas and built five prototypes before coming up with a concept
for a social search engine, which eventually became Aardvark.
The ideation process for the Dropbox can be summarized as:
Houston got the idea for Dropbox while waiting for a bus at Boston’s South Station in
December 2006. He had planned to work on a programming project for Bit9 during the
four-hour ride to Manhattan, but left his USB flash drive at his Cambridge apartment.
Frustrated, he began designing a service to sync and share files between personal
computers over the Internet.
The Dropbox ideation and refinement was also happened in a very continuous fashion.
To me, regarding the details given in the pertinent articles, I conclude that the ideation process for
the Aardvark is more comprehensive and well-elaborated. Furthermore, the product idea for the
Dropbox was stemmed from a need, and the one for the Aardvark was stemmed from an
expectation.
Mustafa Değerli – Aardvark Case Page 6 of 16
4. Vision vs. Experimentation – Aardvark
Even though at some points they do have pretty different views, the founders of Aardvark
complemented each other, in general. For instance, Horowitz is 10 years older than Max and they
have very different interests and personalities. Different goals, as well. Max wanted to make a big
splash with the Aardvark. Horowitz wanted to build a product that reflects sensibilities about the
role of technology in the world. Max is plugged into the Silicon Valley culture; knowing everyone
in that world, which start-ups are hot, and how to dazzle investors and the press.
The roles of entrepreneurial vision and experimentation in Aardvark during both the ideation
process and the product development process are apparent. Basic reasons for my such view are
listed below as bulleted items:
They wanted to create something that could be tested with users. They also wanted to
evaluate ideas quickly, so they needed concepts that required relatively little programming
before they could test them.
Their goal for any idea was to have a prototype that could be user tested within a month.
After they saw what users thought, they could make corrections or discard the idea. They
were trying to shoot down ideas quickly, and most of them got shot down before the
prototype stage.
Horowitz Noted, “We didn’t want to mine knowledge on the Web. We wanted to mine
knowledge in people’s heads. And how do we get into their heads? Through sentiment;
people are going to participate because people want to help other people.”
Ventilla recalled, “We used smoke tests to see whether consumers would respond to an
invitation for a product that we hadn’t yet built. If they didn’t click through an invite, there
was no point in building a prototype. We also would fake functionality by substituting
humans for backend systems; these mechanical turks a pretended to be the product. We
would see if our early users—who didn’t know that they were using a turked product—
kept coming back.”
The cofounders established two product design principles. First, they agreed to seek user
validation at every step of the development process. As Spiro explained, “We wanted to be
Mustafa Değerli – Aardvark Case Page 7 of 16
user-driven—that is, we wanted to be informed by the interests and needs of people who
were actually using Aardvark.” Second, the team wanted to stay nimble with engineering
by being highly collaborative and adaptive.
Horowitz explained, “We had people in every week. We asked our friends to come in; we
grabbed people off the street; and we put listings on Craigslist offering a $15 Starbucks
card to look at our mock-ups.”
Building on their earlier experience with turk testing, Ventilla and Horowitz made a
commitment to a “Wizard of Oz”b infrastructure to learn about user behavior before the
engineering team built an automated question-and-response service.
The Aardvark team actively cultivated and managed the user community, designating 5%
of their most active members as “Aardvocates” and sending them t-shirts and requests to
vote on features.
Ventilla said, “The goal of our product development process was to ensure eventual
product-market fit.
The cofounders raised venture funding before Aardvark completed its end-to-end system.
Ventilla recalled, “We raised $2 million in seed financing from high-profile angels and
then a $5.1 million A round led by August Capital—all while we still had turks running
our backend.” Given their stage, the founders pursued investment from sources that would
allow the company to evolve gradually. Horowitz added, “This was not the kind of start-
up in which investors closely controlled strategy, personnel, and operating decisions. Both
our early angels and our VC pretty much let us do our own thing—though having them
available for advice at crucial points was invaluable.”
Mustafa Değerli – Aardvark Case Page 8 of 16
5. Value Innovation – Aardvark
To me, the Aardvark, at the time of its release, provided real value innovation. They created a
“social search” engine.
The reason for my position is that they created something new and quite niche. That is, thy created
the Aardvark which is effective with questions that required subjective responses, such as “What’s
a good book to read about Romantic poets?” The Aardvark differed itself from the Google and
Bing, traditional search engines, by this means. In addition, the Aardvark created a feeling like
users were engaging in a conversation with someone. Aardvark was not look like a Q&A website
with public information, like Yahoo! Answers. All of these reasons made it possible for the
Aardvark to generate a value innovation
Mustafa Değerli – Aardvark Case Page 9 of 16
6. Multi-sided Markets – Aardvark
In my opinion, the Aardvark is a multi-sided market platform. We should remember that
intermediaries are the actual actors that make two-sided markets or two-sided strategies promising.
Furthermore, pricing, openness, innovation, advertising, and quality investment disputes are the
leading strategies and issues that each of key actors of two-sided markets shall rigorously should
take into consideration. Predominantly, the pricing and openness mean a great deal. the network
effects lie simply positive.
My strategies about monetization and growth for the Aardvark:
Get shares for referrals (generating leads and inserting affiliate links via high commercial
potential queries).
Have adds on pages and applications.
Collaborate with expert users to improve the quality of the content.
Mustafa Değerli – Aardvark Case Page 10 of 16
7. Lean Startup Methodology: Pros and Cons – Aardvark
Components or
Tools from the
Lean Startup
Methodology
Used by the
Aardvark
(Yes / No)
Helped to the
Aardvark
(Yes / No)
Comments
Actionable metrics
and analytics
Fairly yes Fairly yes The Aardvark employed metrics.
However, metrics should be
actionable and they should be
analyzed for possible correction,
corrective action, and preventive
actions.
The Minimal
Viable Product
(MVP)
Yes Yes The Aardvark employed MVPs.
Alpha/Beta Test Yes Yes The Aardvark employed
Alpha/Beta Test.
Pivots Yes Yes The Aardvark employed pivots.
They actively listened customer
feedback.
Customer
Discovery and
Validation
Yes Yes The Aardvark employed
customer discovery and
validation. From ideation to the
end of product development they
Mustafa Değerli – Aardvark Case Page 11 of 16
have included the customers
actively.
Build-Measure-
Learn
Fairly yes Fairly yes Although the Aardvark fairly
employed Build-Measure-Learn
cycle, we don’t have enough
numbers to judge the
effectiveness and efficiency of
the cycle utilized by the
Aardvark.
Validated Learning Fairly yes Fairly yes Although the Aardvark fairly
employed Validated Learning
approach, we don’t have enough
numbers to judge the
effectiveness and efficiency of
the approach utilized by the
Aardvark.
Innovation
Accounting
Fairly yes Fairly yes Numbers related with some of the
constructs are available. But more
numbers (financial data,
predictions) are needed for full
characterization.
Continuous
Deployment
Yes Yes Adequately applied by the
Aardvark. They fine-tuned and
refined as per customer
comments and inputs.
Mustafa Değerli – Aardvark Case Page 12 of 16
8. Lean Startup Meth: Compare & Contrast – Aardvark & Dropbox
Components or
Tools from the
Lean Startup
Methodology
Employed by the
Aardvark
(Yes / No)
Employed by the
Dropbox
(Yes / No)
Comments
Actionable metrics
and analytics
Fairly yes Yes The Aardvark employed metrics.
However, metrics should be
actionable and they should be
analyzed for possible correction,
corrective action, and preventive
actions.
The Dropbox employed metrics
and analytics in a more
comprehensive way.
The Minimal
Viable Product
(MVP)
Yes Yes Both (the Aardvark and Dropbox)
used the MVPs effectively.
Alpha/Beta Test Yes Yes Both employed Alpha/Beta tests
meritoriously.
Pivots Yes Yes Both employed pivots. They
actively listened customer
feedback.
Mustafa Değerli – Aardvark Case Page 13 of 16
Customer
Discovery and
Validation
Yes Yes Both employed customer
discovery and validation. From
ideation to the end of product
development they have included
the customers actively.
Build-Measure-
Learn
Fairly yes Yes Although the Aardvark fairly
employed Build-Measure-Learn
cycle, we don’t have enough
numbers to judge the
effectiveness and efficiency of
the cycle utilized by the
Aardvark.
On the other hand, the Dropbox
appropriately applied Build-
Measure-Learn cycle. The
Dropbox was better especially for
measure and learn parts.
Validated Learning Fairly yes Yes Although the Aardvark fairly
employed Validated Learning
approach, we don’t have enough
numbers to judge the
effectiveness and efficiency of
the approach utilized by the
Aardvark.
Mustafa Değerli – Aardvark Case Page 14 of 16
On the other hand, the Dropbox
appropriately applied validated
learning approach.
Innovation
Accounting
Fairly yes Yes Numbers related with some of the
constructs are available for the
Aardvark. But more numbers
(financial data, predictions) are
needed for full characterization.
Yet, the Dropbox correctly
applied the approach.
Continuous
Deployment
Yes Yes Adequately applied by the both.
They fine-tuned and refined as
per customer comments and
inputs.
Mustafa Değerli – Aardvark Case Page 15 of 16
9. The Mechanical Turk – Aardvark
Mechanical turks are mainly used for machines or devices that can professedly do a fully
automated task, but which in reality is done by a hidden person.
Regarding the Aardvark, the Mechanical Turk is the medium to to fake functionality by
substituting humans for backend systems. They are used to see if early users - who didn’t know
that they were using a turked product - kept coming back.
It was moderately useful and it helped the Aardvark. The Aardvark used to the mechanical turk
learn about user behavior before the engineering team built an automated question-and-response
service. Additionally, by this means, they learned things that informed their work on both the user
interface and the backend systems. Moreover, with the help of the mechanical turk (Wizard of Oz),
the Aardvark automated portions of the work of classifying and routing queries, and they gradually
constrained the turks, so they did less and less of the work.
The mechanical turk approach is more useful for business cases which requires substituting
humans for systems. For example, if we would like to produce products or services which will
automate or to a certain extent replace humans, we can benefit from mechanical turk approach for
cost minimization and early feedback.
Mustafa Değerli – Aardvark Case Page 16 of 16
10. Dropbox, Aardvark, and ULAŞIM
In my project, I have used or I am to use following tools and components, to a certain extent similar
to ones done in the Aardvark and the Dropbox cases.
Actionable metrics and analytics
The Minimal Viable Product (MVP)
Pivots
Customer Discovery and Validation
Build-Measure-Learn
Validated Learning
Innovation Accounting
Continuous Deployment
After analyzing both the cases of the Aardvark and Dropbox, I see that I need to improve my
canvas and model regarding following dimensions:
Actionable metrics and analytics
Innovation Accounting
What is more, since my revenue stream reasonably matches with the one of the Aardvark (we both
mostly employ advertising for making money), I will analyze the details of the Aardvark to ensure
that I can benefit from these.
In addition, in my opinion, what the both (the Aardvark and the Dropbox) did great are pivots and
validated learning. These are the points that I also gave much importance during my project
development and improvement. Still another thing that I will benefit from the cases of the Aardvark
and the Dropbox is the more tangible and functional MVPs. I do know it is not a “must” for MVPs
to be that tangible and functional; yet, to me as I see in both of the cases, they (more tangible and
functional MVPs) are much more than “nice to have.” Therefore, I should have developed MVPs
a bit more tangible and functional than my sketches that I used for hypothesis testing with my
Category 1 [People using local public transportation facilities (Metro and AnkaRay)] customers.