There has been much research interest in the speed of innovation, although few consistent findings have emerged. We unpack the innovation process, focus on the commercialization stage and examine the determinants and impact of the speed of technology licensing. More specifically, using data for the biopharmaceutical industry, we examine: Which licensor and patent characteristics determine the speed of licensing? How does the speed of licensing impact the post agreement payments to licensors? We address these questions using a theoretical premise based on the effects of licensor prominence (size and experience), licensor knowledge structuration (technological depth, technological breadth and experience), and patent appeal (forward citations, scope and complexity). We predict and find that strength in these variables helps licensors discover and be discovered by licensees, resulting in a wealth of agreement options and a strong licensing-out position for licensors. This not only increases the size, complexity and duration of the licensing out-task, but allows licensors to take their time to review, negotiate and select the most attractive offer. This finding is counter to arguments for a fast-paced innovation strategy, as it suggests that the commercialization stage of the innovation process rewards a ‘less haste, more value’ approach. To understand why some patents get licensed and others do not we estimate a portfolio of firm- and patent-level determinants for why a particular licensor’s patent was licensed over all technologically similar patents held by other licensors. Using data for licensed biopharmaceutical patents, we build a set of alternate patents that could have been licensed-in using topic modeling techniques. This provides a more sophisticated way of controlling for patent characteristics and analysing the attractiveness of a licensor and the characteristics of the patent itself. We find that patents owned by licensors with technological prestige, experience at licensing, and combined technological depth and breadth have a greater chance at being chosen by licensees. This suggests that a licensor’s standing and organizational learning rather than the quality of its patent alone influence the success of outward licensing.

1. The propensity and speed
of technology licensing
Seminar at Ceará State University
September 30th 2017
Ian P. McCarthy

2. PAPER 1:

3. “Licensing speed: Its determinants and payoffs”
with Karen Ruckman, under review at the Journal
of Engineering and Technology Management
PAPER 2:

4. THE AUTHORS
Karen Ruckman
Beedie School of Business,
Simon Fraser University
Ian P. McCarthy
Beedie School of Business,
Simon Fraser University

5. 5
MY BACKGROUND: BUSINESS
Pentagon Radiators (Alcan)
Philips Electronics
Footprint Tools

6. 6
MY BACKGROUND - ACADEMIC

7. 7
MY BACKGROUND - OTHER STUFF

8. 8
MY BACKGROUND - OTHER STUFF

9. 9
MY BACKGROUND - OTHER STUFF

10. WHAT IS LICENSING?
• Licensing is an activity where the
owner of IP (the licensor), allows
another party (the licensee) the
rights to use, adapt and
commercialize the IP.
• A patented technology has IP
protection for a finite period of
time.
• Parties can earn monopoly profits
until this patent protection period
expires.

11. WHAT IS LICENSING?
As of 2017, the university had received more than $320
million from royalties (20%)
The 9 inventors (6 scientists, two trainers and a lab technician)
have received $1.3 billion.

12. LICENSING INCOME- US UNIVERSITIES: LIFE
SCIENCES 2015
B Huggett - Nature Biotechnology, 2017 - Nature Research

13. THE MARKET FOR PATENTED IDEAS: THE
ECONOMIST
• In 2005:
– U.S. patent licensing revenue is
$45 billion annually
– Worldwide, the figure $100 billion
and growing fast.
– IBM earning over $1 billion
annually from its intellectual-
property portfolio in 2005
– HP's revenue from licensing has
quadrupled in less than three
years
• The global licensing market was
estimated to be worth $200 billion in
2011 (Alvarez and Lopez, 2015)
• There continues to be increasing
supply and demand for technology

14. THE RISE OF LICENSING
Adapted from Chesborough 2004

15. THE RISE OF LICENSING
Adapted from Chesborough 2004

16. RESEARCH ON LICENSING (AND
PATENTING)
• Why firms license technologies
(e.g., Rockett, 1990)
• What factors affect the propensity
to engage in licensing (e.g.,
Fosfuri, 2006; Nagaoka & Kwon,
2006)
• How different approaches to
licensing impact licensing
outcomes (e.g., Gallini & Wright,
1990).
• The duration and timing of the
patent application and approval
process (Popp, Juhl & Johnson,
2003; Gans, Hsu & Stern, 2008)

17. IN OUR RESEARCH WE …..
• Surprisingly no research attention has been paid
to:
– Why some patents get licensed and others do
not? (paper 1)
– The determinants and payoffs of licensing
speed (paper 2)

18. PAPER 2: WHY SOME PATENTS GET
LICENSED AND OTHERS DO NOT
• Despite the considerable market for
patented technologies, many patents
remain unlicensed.
• We seek to understand why some patents
get licensed, while others do not.

19. WE KNOW THAT PATENT
CHARACTERISTICS ARE IMPORTANT.
The complexity, breadth and perceived importance of a
patent make it more attractive to licensees and increase
its chances of being licensed.

20. BUT WHAT ABOUT THE OWNER OF THE
PATENT?
We address this gap.
It’s important that companies
know which characteristics to
capitalise on so as to be better at
licensing out.

21. Patent Characteristics
Patent complexity (+)
Patent age (+)
Patent scope (+)
Patent citations (+)
Licensor Characteristics
Licensor size (+/-)
Licensor experience (+)
Licensor research intensity (+)
Licensor research age (+)
Past relationships between a
licensor and licensee (+)
The likelihood a
patent will be
licensed.
WHAT WE LOOKED AT

22. OUR APPROACH
We identified a set of biopharmaceutical patents
that were licensed.
For each of these licensed patents we identified
a set of 19 alternate patents that could have
been licensed but were not.
To do this, we used topic modeling to analyse
each licensed patent and then identify
remarkably similar patents that were not
licensed

23. DATA
• Biopharmaceutical licensing agreements from the RECAP
database by Deloitte.
• An initial search = 297 non-university licensing agreements
(1993 and 2008)
• After data matching across three databases, the default
sample was reduced to 93 agreements.
• RECAP provided information about the past relationship
history of each partner.
• Accounting data (revenues and R&D expenditures) from
Compustat and filings with the US Securities and Exchange
Commission.
• Data for patent-based variables came from the NBER US
patent citation data file (Hall et al., 2001) and its 2006 update.

24. SOME OF THE LICENSOR VARIABLES
• Technological prestige = Licensor’s average number of non-
self forward citations on patent stock within 5 years after
patent granted
• Licensor depth = Number of patents granted to licensor in
same IPC as licensed patent during 5 years before
agreement, logged
• Licensor breadth = Number of different IPC classes in licensor
patents that were granted within 5 years before agreement
• Licensor experience = Number of license agreements during
5 years before agreement involving the licensor

25. THE PATENT VARIABLES
• Patent complexity = Number of technological claims made by
the patent
• Patent age = Year of agreement minus year patent was
granted
• Patent scope = Number of IPC categories listed on the patent
• Patent citations = Number of backwards citations listed on the
patent

26. LICENSOR CHARACTERISTICS THAT IMPACT
THE LIKELIHOOD A PATENT WILL BE LICENSED
HOW THE LICENSOR CHARACTERISTIC
IMPACTS THE LIKELIHOOD A PATENT
WILL BE LICENSED
Licensor prestige: The extent to which a
licensor is viewed favorably by licensees.
Prestigious licensors have a greater
chance of licensing-out as licensees
are more likely to know about and be
attracted to them due to the increases
in licensor standing, visibility,
credibility, and the benefits by
association
Licensor experience at licensing: The extent
to which a licensor is practiced at licensing.
Experienced licensors learn from prior
licensing activity which makes them
more proficient at licensing-out and in
turn increasing the likelihood their
patents will be licensed.
FINDINGS

27. LICENSOR CHARACTERISTICS THAT IMPACT
THE LIKELIHOOD A PATENT WILL BE LICENSED
HOW THE LICENSOR CHARACTERISTIC
IMPACTS THE LIKELIHOOD A PATENT
WILL BE LICENSED
Licensor technological depth: The extent to
which a licensor specializes in a particular
technological area.
Individually technological depth and
breadth have no impact on licensing
likelihood.
However, licensors strong in both
technological depth and breadth have
greater learning and innovation
capabilities and these also combine to
increase a licensor’s ability to license-
out.
Licensor technological breadth: The
technological scope of a licensor’s past
patenting efforts.
FINDINGS

28. THE TAKEAWAYS
1. The halo effect.
Licensees are not just making licensing
decisions based on quality of the patented
technology, but also on the reputation of
and technological fit with the licensor.
Prestige makes a licensor more visible to
licensees. It is also makes licensors more
legitimate and attractive to licensees.

29. THE TAKEAWAYS
2. Experience matters.
Experience allows licensors to accumulate
the specific knowledge, people, and
routines required to find and do deals with
licensees
Experience increases a licensor's ability to
be known to and selected by potential
licensees.

30. THE TAKEAWAYS
3. Organizational learning
enhances licensing likelihood.
A licensor’s technological depth and
breadth balance the knowledge
processes for technology transfer
transactions.
Depth and breadth combined, signal
to prospective licensees that a
licensor possesses the efficient
ability to convey and transmit
technical ideas.

31. “Licensing speed: Its determinants and payoffs”
with Karen Ruckman, under review at the Journal
of Engineering and Technology Management
PAPER 2:

32. INNOVATION SPEED
• Fast innovation =
– greater revenue returns (Ringel, Taylor and Zablit, 2015)
– more new product development (Acur, Kandemir,
Weerd‐Nederhof and Song, 2010)
– growth in sales and initial public offerings (Eisenhardt, 1989).
• Fast innovation =
– less impactful and profitable outcomes (Steen and Dhondt,
2010)
– mistakes and inefficiencies (Crawford, 1992).
• To help resolve this conflict, we explore the determinants and
impact of licensing speed.

33. INNOVATION SPEED
• Research on innovation speed has
focused on the speed for all three
stages of the innovation process:
– the conception of an idea
– the development that idea,
– and the eventual
commercialization of that idea

34. WHY THE SPEED OF LICENSING?
• The market for licensing patents is big and
growing.
• Patented technologies are protected for a
finite time period:
– licensing slowly reduces the value of
the technology to both licensor and
licensee (Hegde, 2014; Markman et al.
2005).
– fast licensing = hurried sub-optimal deal
and lower price and reduced returns
(Allain, Henry, and Kyle 2011; Mauleon,
Vannetelbosch and Vergari, 2013).

35. IN OUR RESEARCH WE …..
• Surprisingly no research attention has
been paid to licensing duration or
speed.
• In response we examine:
– Which licensor and patent
characteristics determine the
speed of licensing?
– How does the speed of licensing
impact the post agreement
payments to licensors?
Licensing speed = the elapsed time from when a patent for a technology is
filed to when the licensing deal for the technology is announced.
Patent
filed
Patent
approved
Patent
licensed

37. LICENSOR PROMINENCE
• Licensor prominence = a licensor’s standing or status relative
to other licensors, as perceived by potential licensees.
• It is a signal of the latent value of a licensor’s patent
technology offerings, which impacts licensing speed through
different kinds of visibility, appeal and power effects.
– Hypothesis 1a. Licensor age increases the time it takes to
reach a licensing agreement.
– Hypothesis 1b. Licensor size increases the time it takes to
reach a licensing agreement
– Hypothesis 1c. Licensor prestige increases the time it
takes to reach a licensing agreement.

38. LICENSOR KNOWLEDGE STRUCTURATION
• Knowledge structuration is the extent to which a licensor’s
knowledge portfolio is based on multiple or a few technology
domains (George at al 2008).
• Knowledge structuration indicates a licensor’s breadth and
depth of technological expertise, which is linked to learning
and exploration capabilities (e.g. Ahuja and Lampert, 2001;
Rosenkopf and Nerkar, 2001).
– Hypothesis 2a. Licensor technological depth increases the
time it takes to reach a licensing agreement.
– Hypothesis 2b. Licensor technological breadth increases
the time it takes to reach a licensing agreement.

39. PATENT DESIRABILITY
• Forward citations = knowledge impact of a patent (Duguet and
MacGarvie 2005), an indicator of the value and desirability of the
patent (Ceccagnoli et al. 2010; Harhoff, Scherer, and Vopel 2003).
• Patent scope = breadth of technological areas cited by the patent’s
technology (Ruckman and McCarthy, 2017; Sakakibara, 2010).
• Claims = building blocks of a patented invention, and they are
indicative of the purview of the invention (Hall, Jaffe, and Trajtenberg
2001).
– Hypothesis 3a. Patent forward citations increase the time it takes
to reach a licensing agreement.
– Hypothesis 3b. Patent scope increases the time it takes to reach
a licensing agreement.
– Hypothesis 3c. Patent complexity increases the time it takes to
reach a licensing agreement.

40. PAYOFFS TO LICENSOR
• The payoff to a licensor is typically an up-front lump sum
and/or royalty rate paid as a percentage of the final sales.
• Previous research looks at which type of payoff will
predominate (see: Kamien and Tauman, 1986).
• We focus on how licensing speed affects the payoffs.
– Hypothesis 4. A slower licensing speed increases the
payoffs to the licensor.

41. DATA
• Same industry and data sources as for Paper 1
• Different variables for the hypotheses

42. MODEL
• Dependent variable = (Time 1 – Time 0)
– Time 0: date of patent application
– Time 1: date of licensing agreement
• Survival analysis
• Log-logistic model

43. RESULTS
• We find that strength in licensor prominence (H1), licensor
knowledge structuration (H2) and patent desirability (H3), all
work to slow down the time it takes to licensing-out a patented
technology.
• This suggests that strength in these factors endow licensors
with an abundance of licensing opportunities that:
– increases the size, complexity and duration of the licensing
out-task, and
– allows licensors to take their time to review, negotiate and
select the most attractive offer

44. RESULTS
• The results show that slower licensing speeds are associated
with higher royalty rates (H4).
• There also evidence that licensor size, research intensity, and
prestige and patents with a high number of forward citations
and scope, reduce royalty rates.
• Licensee size and patent complexity act to increase lumpsums.

45. TAKEAWAYS
• While innovation speed is important, when it comes to
licensing, it is beneficial for licensors to slowly and carefully
license-out patented technologies.
• There are different stages and aspects to the innovation
process, and the benefits of being fast or slow can vary for
these.
• Licensing is not just a functional transaction based solely on
the quality and suitability of a patent. Perceptions about the
licensor matter.
• A licensor’s knowledge base and related organizational
learning capabilities (knowledge structuration) influence the
speed of licensing.

46. TIMELINE PAPER 1
• May 2009 = Idea for paper
• Dec 2012 = First submission to the Strategic Management Journal
• Aug 2013 = Second submission to the Strategic Management Journal
• May 2014 = First submission to Research Policy
• April 2015 = Second submission to Research Policy
• May 2015 = First submission to Industry and Corporate Change
• Jan 2016 = Second submission to Industry and Corporate Change
• June 2016 = Third submission to Industry and Corporate Change
• July 2016 = Accepted by Industry and Corporate Change !!!!
• July 2017 = Published in Industry and Corporate Change !!!!

47. Dr. Ian
McCarthy
@toffeemen68
Professor, Technology and
Operations Management
Beedie School of Business
Simon Fraser University
Ian McCarthy
It Depends Blog