1. Joe Kinsella
MBA Guest Lecture, UMass Boston
October 13, 2011
Blog: HighTechInTheHub.com
LinkedIn: linkedin.com/joekinsella
Twitter: @joekinsella
2. My Background
Success Stories
Process for Iterating
Case Study
Parting Thoughts
Q&A
3. 20+ years in high tech software startups in
the Boston area with engineering focus
Currently VP of engineering for
Sonian, cloud startup
Formerly VP of engineering at
SilverBack, acquired by Dell in 2007
Technology enthusiast & startup junkie
Responsible for delivery of 18 software
products
5. Howard Shultz joined Seattle coffee retailer
Starbucks in 1982
On visit to Italy he becomes enamored with
social culture in the espresso bars
Tests concept at Starbucks, leaves to start his
own coffee shop, later acquires Starbucks
First stores launched outside Seattle in
1987, grows to 165 by 1992 IPO
2010: 17,000+ locations, 55 countries, $9.7B
revenue
6. Started with hypothesis: Italian coffee culture
could be transplanted to the U.S.
Pursued vision despite rejection
Iterated concept from 1983 to 1987
Visited over 500 espresso bars in Italy
Piloted concept with small 300 square foot
coffee bar in a retail store
Leveraged knowledge of industry experts
Iterated
seating, music, service, roasts, uniforms, and
menu in response to customer feedback
7. Michael Dell, a freshman at UT, identifies
market opportunity to sell computers
directly to businesses
Starts selling computers from dorm room
Quits school to build computer business
Grows company and goes public in 1987
Rapid revenue growth: $159M 1988, $2B
1993, $32B 2000
8. Started with hypothesis: could make money
and pursue hobby upgrading PCs
Pursued vision despite resistance from
parents
Iterated concept over 1983-1986
Sold expansion kits via small ads in papers
Adjusted business and products based on
customer feedback
When asked by customer if he could build
computers, decides to make prototype
Tests selling computers at PC show
On success, changes business focus
9. The idea is often the easiest step
The idea is often wrong
Must focus on identifying customer value
Must talk to and respond to customers
Break through happens when you start small
and incrementally adjust
Hard work and willingness to take risks will
not make up for real customer value
Need process to iteratively change features/
functions in response to customer needs
11. Introduced by Colonel John
Observe
Boyd in 1990s
Concept: opponent who can
move through cycle quickly
can gain advantage (get Act Orient
“inside decision cycle”)
First used in military circles,
but broadly applied today Decide
12. Winning is about agility
Requires iteratively adjusting to
circumstances “on the ground”
Cannot be achieved from behind a desk
Can disrupt competitors by getting “inside
their decision cycle”
Winners move faster than their competitors
13. Continuous customer Ideas
interaction
Continuously test and Learn Build
adjust hypothesis
Deliver value incrementally
Data Code
Grow based onproven
success Measure
14. Early forms used as early as the 1970s
Scrum early methodology in mid-1990s
First Scrum team came out of Boston area
(Easel Corporation, 1994) *
Methods owe allegiance to manufacturing
methodologies (e.g. Lean, Kanban)
Movement driven as reaction to
heavyweight methods of past (e.g.
waterfall)
* Easel was my first software job after college
16. Great products and services result from
epiphanies, e.g.
• Strategy – what if I sold computers instead of upgrades?
• Delivery – what if I added seating to my café?
• Product – what if I provided software to allow MSPs to
connect to their customers without a VPN?
Epiphanies come from
• Understanding customers and market
• Iterating & experimenting
• Continuous focus on customer value
It takes numerous epiphanies, often acquired over
years engaged with customers, to achieve success
17. 1. Has a Problem
+
2. Is Aware of Having a Problem
+
3. Has Been Actively Looking
For a Solution
+
4. Has Put Together a
Solution out of Piece Parts
+
5. Has or Can Acquire
a Budget
From The Four Steps to the Epiphany by Steven Gary Blank
19. Story
is one or more sentences that
capture what a customer wants to achieve
• E.g. User should be able to have comfortable and
social environment for drinking coffee
Should be sufficiently granular to be
actionable in short period of time
Stories should be recorded: e.g. 3x5
cards, agile board
Stories should receive weighting (story
points) based on effort to execute
20. Prioritizestories in order based on
potential business impact
Derive major themes (“epics”) from stories
to review with customers
Bring customers through prioritization
exercise to ensure ranking correct
Maintain “backlog” of unfinished stories
22. Sprint is unit of development that delivers
one or more stories to market
Typical lengths for software is from 1-4
weeks
Starts with planning and ends with delivery
of new version of product or service to
customer
Sprints can be applied to all businesses &
product
23. Each sprint starts with prioritization of
backlog
Development is timeboxed
Sprint results in usable deliverable
Larger changes decomposed into
incremental deliverables
Stories that do not get completed in sprint
move back to back log (“train leaving
station”)
24. Once new features launched, essential to
return back to Observe of OODA
Follow up with customers to gather
feedback on new services/features
Write new stories for backlog based on
feedback
Prioritize new and existing stories in next
sprint
Rinse & repeat
26. 1. Not talking to enough customers
2. Not talking to the right customers
3. Being overly vested in idea
4. Premature scaling
5. Not developing incrementally
6. Unrealistic expectations
7. Not differentiating customer must haves vs.
nice to haves
8. Lack of meaningful milestones to judge
progress
9. Not knowing when to give up on a direction
28. Author: Eric Ries, Silicon Valley
entrepreneur
Coined term “Lean startup” in
2008
Sample company using
methodology: Dropbox
Use agile software methods
Drive maniacally customer-
centric process
29. Author: Stephen Blank
Step by step process for
bringing successful products to
market
Drives key lean startups
concepts, including minimum
viable feature set
Introduced earlyvangelist
30. Author: Clayton Christensen
Shows companies exist in value
networks, interdependent social and
business resources within a market
These value networks
• Support driving sustaining changes in a product
or service
• Can disincentivize companies in responding to
disruptive changes
E.g. disk drives, digital
photography, mobile telephony, Internet
Protocol, outpatient clinics, ultrasound