- Email: [email protected]
Linking technology intelligence to open innovation
Technological Forecasting & Social Change 77 (2010) 335–343
Contents lists available at ScienceDirect
Technological Forecasting & Social Change
Linking technology intelligence to open innovation Mark Veugelers a,⁎, Jo Bury a, Stijn Viaene b a b
VIB, Rijvisschestraat 120, 9052 Ghent, Belgium Vlerick Leuven Gent Management School & K.U.Leuven, Vlamingenstraat 83, 3000 Leuven, Belgium
a r t i c l e
i n f o
Article history: Received 20 April 2009 Received in revised form 3 August 2009 Accepted 18 September 2009 Keywords: Technology intelligence Open innovation Life sciences Real options reasoning Technology radar
a b s t r a c t The explosive growth of the Internet has led to a dramatic increase in data sources for (competitive) technology intelligence. Appropriate implementation and use of IT tools to gather and analyze these data is of key importance for the creation of actionable technology intelligence. A strategy to optimize investments in the identified technologies becomes of paramount importance if an organization wants to match knowledge and ideas originating from outside of the organization with internal core competences. Such a strategy can create competitive advantage by effectively linking technology intelligence to open innovation. We show how VIB, a life sciences research organization, has established technology intelligence processes to identify a multitude of external technologies of interest, which are subsequently “probed” for their potential and fit with VIB using real options reasoning, thereby supporting open innovation. Our methodology may be useful for other organizations which are considering implementing open innovation approaches. © 2009 Elsevier Inc. All rights reserved.
1. Introduction Changing technology, driven forward by relentless innovation, affects everybody's business. Smart organizations do not wait for change to happen but actively monitor and take advantage of changing environments and new innovations. “Open innovation” approaches, which implement externally developed technologies or ideas within an organization, have been proposed as a remedy for decreasing product half-lives, faster technology cycles, and increased global competition [1]. To enable open innovation, innovation intermediaries such as online technology transfer exchanges, which list technologies available for licensing to interested parties, or technology brokers, who solicit new innovations by posting problems requiring a solution, have emerged to address specific technology needs. However, these intermediaries typically concentrate only on a fraction of the technology intelligence relevant to an organization's strategy. Technology intelligence has been defined as “the capture and delivery of technological information as part of the process whereby an organization develops an awareness of technological threats and opportunities” [2]. To identify, as comprehensively as possible, all options for introducing new, external technologies/innovations in the organization, it becomes necessary to analyze large amounts of technology data [3], originating from disparate sources outside the organization. Use of appropriate IT tools and extensive data mining analyses, such as text mining [4], can generate actionable technology intelligence. Technology intelligence can have many uses (e.g. in strategy, marketing or human resources). In this paper, we will show how we have generated technology intelligence that allows for the systematic identification of externally developed disruptive technologies, which are probed for their potential value to and fit with our organization, using real options reasoning based technology investments. Our work shows how technology intelligence can be linked to open innovation in practice. ⁎ Corresponding author. Tel.: + 32 9 2446611; fax: + 32 9 2446610. E-mail addresses: [email protected] (M. Veugelers), [email protected] (J. Bury), [email protected] (S. Viaene). 0040-1625/$ – see front matter © 2009 Elsevier Inc. All rights reserved. doi:10.1016/j.techfore.2009.09.003
336
M. Veugelers et al. / Technological Forecasting & Social Change 77 (2010) 335–343
2. VIB: focus on excellence in life sciences research and technology transfer VIB (http://www.vib.be) is a Belgian non-profit life sciences research institute established in 1996. A total of 65 research groups, encompassing more than 1100 scientists, are distributed over six research campuses. Using advanced gene technologies, VIB researchers study a wide variety of biological processes in the human body, in plants and in micro-organisms. VIB focuses on excellence in science and in technology transfer. Technology transfer of in-house generated inventions to external parties and start-ups has been very strong at VIB. However, the adoption of technologies developed outside VIB has been suboptimal in certain cases (e.g. the technology was acquired too late compared to the competition or the acquisition was too expensive compared to other options). In these cases, technology acquisition decisions were frequently made on an ad hoc, reactive basis by individual research groups, who did not sufficiently consider institutional strategy and synergies or, in some cases, alternative technologies. Such decisions could be explained as a legacy of VIB's origins and structure, as VIB was set up as a joint venture of individual research departments. Since several departments often acquired identical, expensive scientific instruments, leading to excess capacity, it became clear that an institutional technology management strategy could improve technology acquisition, while simultaneously optimizing technology adoption and spotting of new technology opportunities. To this end, several barriers had to be overcome, such as the lack of appropriate processes to scan for novel technology opportunities and evaluate technology acquisitions from an institutional perspective, the preference for conservative risk aversion versus opportunity recognition for novel technologies, and the lack of information on the progress of technology adoptions by other organizations. These barriers significantly slowed down the acquisition of several platform technologies, even when these technologies had already established a low-risk profile and proven their potential to early adopters in other organizations. 3. A framework for technology intelligence at VIB Several life sciences technologies have had a dramatic impact on scientific productivity since their introduction [5]. These enabling technologies led to bursts of scientific discoveries, publications, patents and commercial activity. A system to gather early intelligence on disruptive technologies and a correct timing of access or acquisition might facilitate successful adoption of these technologies, leading to a competitive advantage. A retrospective analysis of VIB technology adoptions, as well as of VIB decisions not to adopt, identified several processes that were lacking or could be improved in order to support both external technology identification and access to the identified technologies. In 2006, VIB's management therefore decided to support the establishment of technology intelligence processes to identify technologies of interest to VIB. However, in order to make the intelligence actionable for technology sourcing, VIB's management also decided to allocate a limited budget to facilitating access to the identified technologies. Linking actionable technology intelligence to a technology investment strategy could potentially facilitate VIB's technology management. As the time dimension is important in technology intelligence, our aim was to identify and access fast-moving disruptive technologies in a timely manner. Early access to the right high-potential, high-risk technologies could offer VIB the opportunity to increase its scientific and intellectual property output. 4. Technology intelligence identifies technology opportunities At the outset, we identified two key questions to address: - What is possible with technology: which technologies are out there? - What technology is valuable and needed by our organization: which technologies fit with our organization? In order to answer the question “What is possible with technology?” we needed to gather and organize information on all current life sciences technologies (current state of the art). These “baseline technologies” form the basis for the development of a technology intelligence process that allows for the continuous, cheap and efficient updating of the organization on technologies with accelerating growth rates as well as for the detection of signals of new, emerging technologies. To ensure comprehensive identification of technologies developed by both recent start-ups and large companies, we perform worldwide, continuous technology scanning. In the life sciences field, there are thousands of companies that develop tools and instruments of potential interest to VIB, with the number of academic research groups being a multiple of this number. In addition to scanning for technologies, we also decided to scan for technology adoptions by other organizations (competitive technology intelligence). Establishment of a (competitive) technology intelligence process, allowing for the collection of real-time information on all technologies of interest and their state of adoption by other organizations, requires distinct steps, defined as the ‘technology intelligence cycle’ (Fig. 1). Five stages can be distinguished in the technology intelligence cycle (adapted from [6]): • Planning and management: consulting with decision makers to define which intelligence should be produced [for VIB: (new) life science technologies with disruptive potential]. • Identification of sources: locating sources for data collection [for VIB: academic literature (Pubmed), patents and the Internet]. • Primary source collection: actual capturing of the data from the sources [for VIB: use of IT tools and databases such as ThomsonISI Web of Science, Matheopatent and Matheoweb].
M. Veugelers et al. / Technological Forecasting & Social Change 77 (2010) 335–343
337
Fig. 1. The technology intelligence cycle at VIB.
• Analysis and production: the primary data, consisting of thousands of documents, is analyzed for its usefulness and filtered to facilitate the analysis. Filtered data is transformed into actionable technology intelligence [for VIB: a ‘technology scout’ performs data mining using software tools and synthesizes conclusions into technology intelligence profiles]. • Reporting and information: presentation of the findings to decision makers and decision support for strategy. Feedback of decision makers is used to fine-tune the cycle [for VIB: technology intelligence profiles presented to management and employees through seminars and Intranet]. The output of the technology intelligence process for VIB would be a set of technology intelligence profiles on new disruptive life sciences technologies of interest to VIB. Access to these technologies and/or co-development with VIB could lead to competitive advantage by facilitating the generation of new intellectual property, start-up companies with new business models, or joint ventures (open innovation in practice). In order to optimize matching our internal technology needs and absorptive capacity with the technology intelligence obtained on external technologies, we opted to start on a small scale with a single, internal ‘technology scout,’ allocated part-time, with an initial focus on gathering science and technology data and performing analyses leading to the production of technology intelligence.
5. IT tools and databases support the technology intelligence cycle Establishing technology intelligence processes from scratch can be a serious challenge. Devising technology identification strategies is not a simple task, and scratching only the surface can be dangerous, as it may lead to overlooking key technologies of strategic importance to the organization [8]. The usual channels (market reports, suppliers, etc.) for identifying new technology opportunities are usually already well known to the organization, but these are far removed from a systematic process that continuously updates the intelligence on technologies of interest as well as identifies off-radar technologies, currently unknown and outside the normal field of view. In the case of the latter, “You don't know what you don't know” too often gets in the way of acquiring information about technologies currently outside the radar of the organization. Effective search strategies can be a source of competitive advantage [7], and the smart use of search engines such as Google may turn up interesting results. However, the Internet is growing so large that quick keyword searches at low search depth are likely to miss a number of technologies of interest to the organization. Data sources for technology intelligence contain millions of documents and to automate the collecting, processing and analysis of millions of textual data sources, one needs appropriate IT tools. To select the right tools, an organization needs to understand the role and possibilities of IT in the technology intelligence process, as this will allow the identification, selection and installation of the appropriate, cost-effective software [9,10]. Until a few years ago, the market for IT tools supporting technology intelligence was considered a niche market with just a few tools available. The accelerating growth of data on the Internet has fuelled the growth of a wide variety — in terms of their price, capabilities and focus — of software tools for multiple aspects of technology intelligence [6]. Software packages as well as consulting services are available for data mining, text retrieval and classification, semantic analysis and document summarization, patent searching, webpage tracking, and Internet monitoring using syndication technology. The situation is made more complex by software able to cover single or multiple steps of the technology intelligence cycle. An incorrect implementation of these tools may either lead to ‘information overload’ or to ‘myopia,’ due to the non-capture of essential information. As no monitoring system or software is perfect, a decision will have to be made on the trade-off between price and performance of the software and between its levels of precision and recall of technology data. Luckily, an organization with no technology intelligence processes in place can get one of the ground at a relatively low cost, as there are a variety of reasonably inexpensive yet highly efficient technology intelligence software packages available, which can be installed on any desktop computer.
338
M. Veugelers et al. / Technological Forecasting & Social Change 77 (2010) 335–343
In addition to software for technology intelligence, several databases, such as the Science Citation Index from Thomson-ISI, compile information on academic research, while information on technological innovations is available through publicly accessible patent databases (i.e. WIPO, USPTO, EPO, JPO), as well as from commercial service providers (e.g. Thomson, STN, etc.). Several tools (e.g. LexisNexis, Dialog, Google Alerts, Twitter Alerts, etc.) are also available for tracking companies and their press releases as well as new technology developments. Several of these service providers charge access fees, which may be well worth the money as they avoid wasting time on analyzing unstructured and inferior-quality data. VIB started with test-running a few software packages and databases to identify a combination of tools offering a good fit with the organization and high precision and recall. We acquired low-cost technology intelligence software and subscribed to a few commercial database providers to be well positioned for technology data mining. 6. Data-mining software processes data collected from information sources Hundreds of thousands of research articles and patents are published each year, and an immense wealth of new ideas is being made available on the Internet on a daily basis through media such as press releases about product launches, etc. All this information is furthermore growing exponentially in size, making it nearly impossible to stay abreast of rapid changes in technology landscapes without a structured approach to searching and scanning this data overflow. Buried within this data are gold nuggets: disruptive technologies that present as a weak signal in a background of noise and are therefore difficult to detect and interpret. Hence, the technology intelligence process needs to be designed such that it facilitates a broad scan of the periphery while being equipped with the appropriate filters that will allow identification of weak signals from disruptive technologies [8]. This involves a delicate balance between an appropriate filter/cut-off and an increase in analysis workload. The sheer number of technologies, companies and applications to track, and the identification of weak signals from disruptive technologies can stretch scarce resources even thinner, leaving organizations unable to respond quickly to novel research discoveries and technology opportunities. To identify technologies of interest to VIB, a workflow was set up to acquire and analyze large, textual datasets from three major sources of new technology information: academic literature, patents and the Internet (Fig. 2). The general methodology for technology identification was similar for all three data sources. Specific keywords from a ‘technology thesaurus’ — a set of keywords on life sciences technologies — were used for querying the data source. The returned data (abstracts, patent claims and webpage text) was subsequently downloaded with appropriate software, and stored either in text format or in an Excel database for further analysis and filtering. In order to reduce the size of the large dataset, Omniviz software was used for clustering texts (there are other packages with similar functionalities available, such as VantagePoint [11]). Several types of queries were performed: broad searches for macro-level technology landscape analysis as well as more detailed queries for specific technologies, inventors, companies or universities. As each data source has different properties, different strategies are needed to maximize information harvesting from each source: a) Academic literature The academic literature (~ 600,000 new publications a year in the life sciences) contains a wealth of information on peerreviewed publications describing technologies of interest. We scanned the academic literature using Pubmed and Thomson-ISI Web of Science. By searching for abstracts using our technology thesaurus, it became possible to filter the literature for technology-related publications. Abstracts were clustered with Omniviz into individual technology clusters (Fig. 3). While some technology management systems focus exclusively on text mining [12], we also included citations as a key technology indicator. As citations
Fig. 2. Workflow for identifying technologies of interest.
M. Veugelers et al. / Technological Forecasting & Social Change 77 (2010) 335–343
339
are indicative of the innovativeness of a technology, temporal analysis of the most frequently cited publications offered an insight into which technologies were considered critical breakthroughs. Temporal analysis, iterative clustering, identification of authors with company affiliations, and other detailed analyses identified novel technology developments. Taken together, the combination of citation analysis and clustering into technology fields identified several technology opportunities of interest to VIB. b) Patent literature The patent literature (~150,000 new patents published every year) contains key information on novel technologies and inventions [13]. Patent mining requires the analyst to actively scan and analyze all patents that have the potential to affect the organization [14]. A patent-mining strategy for the entire field of the life sciences (and converging science areas such as nanobiotechnology, bioinformatics, etc.) requires the analysis of large datasets, necessitating in turn effective mining, visualization and filtering strategies to arrive at a workable set of data and downstream actions [15]. Software analysis alone cannot be a replacement for a subject analysis of a patent. To ensure and assess the efficiency of the filtering step, it becomes necessary to review part or all of the text, especially the breadth of the claims. Devising a cost-effective and efficient approach to analyzing patent data is also challenging due to the availability of a broad spectrum of database and software resources for the collection of reliable data [16]. Several patent tools and software packages are available to assist in patent analysis, offering tools such as automated patent text downloads, categorization, integration and visualization [17–20]. Patent abstracts of interest were clustered using Omniviz for more in-detail analysis. Patent citation analysis as well as sizing of patent families was performed with Matheopatent software [21]. c) Internet Technology intelligence has traditionally focused primarily on data from patent and research publication abstract databases [22]. This resource base can be extended with data mined from the Internet. Although patent and academic publication data can also be obtained from the Internet, the latter also offers a wealth of additional technology information via company websites, press releases, blogs, message boards, etc. not found in the patent and literature databases. Although rich in content (45 billion web pages and doubling each year), there is a drawback: textual information on Internet websites is presented in a wide variety of text structures. We scanned for technologies of interest on the Internet using Matheoweb software. By using targeted keyword searches and software to extract information from thousands of web pages automatically, we were able to identify a number of technologies with disruptive potential. Data from all three search domains was stored in an Excel data warehouse for further processing and clustering. 7. Transforming data into actionable intelligence In a final step, the most complex one in the technology intelligence process, data is transformed into actionable intelligence. This step is not very suitable for automation and requires a significant human effort: an understanding of the technology and the end user's or organization's requirements with respect to the information is the key. In our case, the meta-analysis from the three different data sources identified intelligence on fast-moving technology fields as evidenced by topic bursts within clusters of technologies. The search strategy also identified multiple new technologies of interest to VIB. Access to these technologies could accelerate research programs at VIB, thereby offering competitive advantage and high returns on technology investments. However, a strategy to minimize the risk of purchasing mistakes and/or investing in unknown technologies with low end user potential, leading to excess capacity, would be very helpful.
Fig. 3. Clustering of life sciences technology abstracts from the academic literature (publication years 2004–2008). A Pubmed search for life sciences technologies and methods revealed hundreds of abstracts, which were downloaded, organized in a text database and fed into the Omniviz data mining tool. Omniviz returns clustered abstracts, with keywords defining each cluster and the number of abstracts for each cluster. The figure shows the clustering of all abstracts. A cluster can group several technologies. Iterative clustering, using for example abstracts belonging to a single cluster, can be used to obtain more granularity.
340
M. Veugelers et al. / Technological Forecasting & Social Change 77 (2010) 335–343
In order to support decision-making and present the identified technologies to VIB's management, we decided to use the technology radar framework in use at Deutsche Telekom AG [23]. A technology radar offers a visual overview of the information obtained from the technology intelligence process. The technology radar screen is a powerful format for the aggregation of technology intelligence on broad technology fields, categorized in different taxonomies, and enables the management of a large number of technologies simultaneously in a single view. Compared to newsletters or technology briefings, which typically would be sent out as individual technology topics over a period of several months, thereby diluting the information over a time period, the technology radar aggregates all data at once. As we identified dozens of technologies of interest to VIB, it became necessary to apply filters that would ensure that the VIB technology radar was populated with the technologies of most interest to VIB scientists. Therefore, an internal Technology Watch team made up of VIB scientists and members of VIB's management was set up to prioritize technologies. The VIB scientists on the Tech Watch team, together with their departmental colleagues, monitor technologies closely related to their expertise. They form the nucleus of an ‘intelligence network’ distributed throughout the organization. The Technology Watch team also contains two members of VIB's management: one patent expert and one technology scout. The Technology Watch team is not full-time allocated to the technology intelligence process but regularly convenes to review and discuss the technologies identified on the VIB technology radar screen (Fig. 4). The Technology Watch team also generates a technology intelligence profile (Fig. 5) for each technology on the VIB technology radar screen. The technology intelligence profile is a review-style one-page document featuring a broad summary of the technology, including key patents, publications, and Internet intelligence data (market status, usage by early adopters, etc.). The Technology Watch team generated over 50 technology intelligence profiles. Each of these highlighted specific technologies or instruments in development at a life sciences company and/or showing accelerating technology capability. These technology intelligence profiles were made available to the entire organization on the VIB Intranet. For presentation purposes, the profiles include mostly visual information and are limited to just a half page of text. The profiles also include links to patents, company websites and literature, and are an excellent tool to alert and inform VIB scientists about disruptive technologies identified through the technology intelligence process. 8. Creating actionable intelligence: assessing technology value and fit with VIB The VIB technology radar allowed us to visualize several options for investment in disruptive technologies, making it evident that the acquired technology intelligence could be a major tool to support technology sourcing at VIB. However, the cumulative cost of investing in all technologies featured on the VIB technology radar would greatly exceed the annual budget allocated to accessing identified technologies. To prioritize investments in the technologies with the highest potential for VIB, we tried to use a technologyrating matrix, ranking technologies according to potential impact and implementation risk. After numerous discussions, this approach was abandoned as the rating of technologies, especially of high-potential/high-risk technologies, was seen as too complex. To reduce the inherent risk of investing in high-risk/high-reward disruptive technologies, we opted for a real options reasoning approach [24]. This approach entails investing in several research projects, which use the technologies of interest, to obtain
Fig. 4. Technology radar for VIB. Meta-analysis of academic literature data, patent data and Internet data identified technology clusters with temporal topic bursts. These highly active technologies were visualized on a technology radar, which depicts technologies with their risk profile (blue: low risk, green: medium risk, red: high risk) and market stage. Technologies on the market are on the inner circle, technologies nearing market launch in the next circle, and technologies further away from daily lab use in the two outer circles.
M. Veugelers et al. / Technological Forecasting & Social Change 77 (2010) 335–343
341
Fig. 5. Technology intelligence profile. The technology intelligence profile is a review-style one-page document which focuses more in detail (key patents, publications, and Internet intelligence) on a technology present on the technology radar for VIB. This example highlights the technology “gene synthesis,” in the broad technology field “Synthetic Biology.” The profile starts with a technology snapshot, an in-house made, short 10-line overview of the technology. It is followed by a table which features quantitative technology indicators (no. of Pubmed hits (hits for query “gene synthesis,” “synthetic biology”), no. of companies (identified by patent assignees and Internet search), Google alert rate (increase in Google Alerts per year, comparing 2008 to 2004), no. of patents (2004–2008) (total number of WIPO patent applications filed between 2004 and 2008), patent acceleration rate (% increase in annual WIPO patent applications, comparing 2008 to 2004)). These indicators allow quantitative comparisons between different technologies. The technology indicators are followed by a concise summary of results from academic, patent and Internet data. Key publications highlight highly cited articles on “gene synthesis” technology. Key companies in the field of gene synthesis were identified by analyzing patent databases and querying the Internet. Internet keywords were generated by generating a tagcloud from all Google Alerts. Patent landscape was generated by analyzing patent databases for “gene synthesis” patents and assignees for these patents; number of patent applications is indicated. Recent Google Alert displays recent hits for a Google Alert which scans the web for new webpage hits on “gene synthesis.”
342
M. Veugelers et al. / Technological Forecasting & Social Change 77 (2010) 335–343
information on the potential of the technologies and their fit with VIB. The information obtained from these projects would then determine follow-on investment decisions. We reasoned that such an approach would also be of value for a company developing a technology, as it would allow the company to establish market intelligence as well as validation of their technologies by VIB scientists. Companies with disruptive technologies frequently look for validation of their technology through research collaborations or services which offer companies visibility, marketing opportunities and customer feedback [25]. We therefore set up a system of annual ‘Technology Grants,’ limited in budget and number. These Technology Grants allow VIB scientists to test the potential and fit of a fixed set of technologies, all identified through the technology intelligence process and featured on the technology radar. The Technology Grants function as real options, or Stage 1 investments. In case of the successful application of the technology in Stage 1, as evaluated by our Tech Watch team, the option is exercised through a larger, follow-on Stage 2 investment in the technology. If use of the technology by VIB scientists fails to deliver promising results in Stage 1, no follow-on investment is made at Stage 2 and the option is abandoned. Each year, VIB scientists can submit proposals for Technology Grants, with a budget allowing for 5–10 real options. A decision to move to Stage 2 is made within one year. This limited timeframe allows for the collection of sufficient information on the technology's potential to facilitate decision-making about exercising or abandoning the option. Information on the importance of the Technology Grants and technology intelligence as a tool to gain access to disruptive technologies as well as a summary of the intelligence on technologies of interest to VIB scientists was presented to all VIB employees through a seminar series. Subsequently a call for projects to be financed by Technology Grants was issued to all 65 VIB research groups. By working with a fixed set of accessible technologies and limiting the number of projects a group can submit for funding, we were able to answer the question “What is needed ?,” as the number of projects submitted for a particular technology functioned as ‘votes’ for that technology. 9. Results of the first iterations of the technology intelligence cycle The response of both scientists and VIB's management to the technology intelligence presented through the technology radar framework has been very positive. The Technology Watch team continues to receive numerous requests from VIB scientists for additional information on the identified technologies. The Technology Grants program has allowed us to establish numerous collaborations with other organizations letting us plug into their technologies (open innovation in action) and leading to an increased output in scientific publications and intellectual property (patent applications). Our Technology Grants program has also worked as an internal voting system, whereby technologies that attracted interest from a large proportion of research departments justified a technology acquisition by VIB. This has reduced the risk of investing in high-potential technologies for which interest would turn out to be limited, and whose capacity would therefore not be optimally used upon their acquisition. In several cases, results from accessing the technologies through Technology Grants projects were excellent, which led VIB scientists to continue investing in these technologies with their own funds. Some of the projects did not provide a fit with VIB scientists because the market strategy of some of the companies did not match the interests of the VIB scientists. However, obtaining this information was also valuable as these projects could be ‘killed off’ quickly. 10. Discussion Setting up a technology intelligence process aimed at identifying, prioritizing and exploiting external high-potential technologies through real options reasoning facilitates open innovation and strategic decision-making, leading to competitive advantage. Investing in external, early-stage or fast-moving technologies is inherently risky as the acquirer will always have incomplete information about the potential of the external technology and its fit with the acquiring organization. Consistent with this, the reaction to the spotting of a disruptive technology is often inertia and a wait-and-see strategy. Interviews with our scientists made it clear that this was often due to risk aversion: early-stage technologies were often seen as too risky and costly. In order to reduce risk, validation or adoption of a technology by peers was often required before a scientist decided to invest in a new technology. Another roadblock turned out to be the time, cost and effort required to establish relationships with external technology providers, an example being the administrative burden associated with accessing external technology (generating CDAs, MTAs, IP-issues, collaboration agreements, etc.). Correct timing in terms of identification, prioritization and technology adoption remains critical to reducing risk exposure. The real options reasoning approach provided by our Technology Grants program partially solves this problem. By investing in several technologies via a centrally administered budget, and thus taking options on multiple technologies, VIB can increase the information about the different technologies accessed during the option period and hence reduce its risk. Taking several options by accessing multiple technologies also increases flexibility. A well-functioning technology intelligence process, which presents multiple technologies as options on a technology radar, regularly updated and presented to top management, also leads to increased opportunity awareness. Knowledge of a broad portfolio of technology opportunities probably also reduces risk exposure, as the number of entry options increases. Technology options allow for the identification of technologies that are at the right stage in terms of their technology potential and fit with VIB, so that first-mover advantage for access to these technologies can be maximized.
M. Veugelers et al. / Technological Forecasting & Social Change 77 (2010) 335–343
343
In the past, several technologies were only assessed by VIB after they were on the market for several years. With the technology intelligence process and Technology Grants program in place, this has dramatically changed: new technologies are spotted early on through our intelligence process and a significant portion of ‘innovator risk’ is removed by allowing a real options reasoning approach to probing external technologies. Since the implementation of our technology intelligence process and Technology Grants program, VIB has invested in several external technologies from companies developing early-stage technologies not yet on the market. Moreover, discussion of the technology intelligence within VIB has greatly stimulated technology information awareness as well as early spotting of technologies of interest. Open innovation requires that organizations use external knowledge, ideas, technologies and Intellectual Property to accelerate innovation.VIB identifies external technology opportunities through technology intelligence and accesses these external technologies using a real options reasoning based approach. Access to external technologies and Intellectual Property is covered by negotiated research collaboration agreements and licenses that allow VIB scientists to pursue specific, innovative applications of a partner's technology. VIB scientists' use of these technologies has led to the creation of new Intellectual Property for VIB. VIB's patent portfolio is used for generating new biotechnology start-ups or to generate revenue via out-licensing of Intellectual Property to external partners. In conclusion, this paper frames the acquired technology intelligence in the context of a technology radar, from which external technologies are selected for investment through a real options based reasoning approach. Real options reasoning have previously been connected to open innovation in the context of corporate venturing and the creation of new business models. The approach detailed in this paper extends the use of real options reasoning in open innovation to technology sourcing based on technology intelligence. Our approach, which links technology intelligence to open innovation through real options reasoning, could be of prime interest to any organization aiming to increase the value of technology intelligence for open innovation. Pharmaceutical/biotechnology research labs and life sciences institutes in particular may benefit from following approaches similar to the ones described in this paper. References [1] H. Chesbrough, The era of open innovation, MIT Sloan Management Review 44 (3) (2003) 35–41 Spring. [2] C.I.V. Kerr, L. Mortara, R. Phaal, D.R. Probert, A conceptual model for technology intelligence, Int. J. Technol. Intel. Plan. 2 (1) (2006) 73–93. [3] D. Zhu, A.L. Porter, Automated extraction and visualization of information for technological intelligence and forecasting, Technol. Forecast. Soc. Change 69 (2002) 495–506. [4] R.N. Kostoff, D.R. Toothman, H.J. Eberhart, J.A. Humenik, Text mining using database tomography and bibliometrics: a review, Technol. Forecast. Soc. Change 68 (3) (2001) 223–253. [5] J.M. Rothberg, J.H. Leamon, The development and impact of 454 sequencing, Nat. Biotechnol. 26 (10) (2008) 1117–1124. [6] Fuld&Company, Intelligence Software Report 2008, 2008 Accessible on line at http://www.fuld.com/Products/ISR2008/. [7] J. Battelle, The Search: How Google and Its Rivals Rewrote the Rules of Business and Transformed Our Culture, Portfolio Trade, New York, NY, 2005. [8] G.S. Day, P.J.H. Schoemaker, Scanning the periphery, Harvard Bus. Rev. 83 (11) (November 2005) 135–148. [9] D. Vriens, The role of information and communication technology in competitive intelligence, in: D. Vriens (Ed.), Information and Communication Technology for Competitive Intelligence, Idea Group Publishing, Hershey, PA, 2004, p. 2003. [10] S. Gordon, M. Tarafdar, R. Cook, R. Maksimoski, B. Rogowitz, Improving the front end of innovation with Information Technology, Res. Technol. Manag. 51 (3) (2008) 50–58. [11] A.L. Porter, QTIP: quick technology intelligence processes, Technol. Forecast. Soc. Change 72 (2005) 1070–1081. [12] H. Chang, J. Gausemeier, S. Ihmels, C. Wenzelmann, Innovative technology management system with bibliometrics in the context of technology intelligence, in: O. Castillo, L. Xu, S.-I. Ao (Eds.), Trends in Intelligent Systems and Computer Engineering; Lecture Notes in Electrical Engineering, Springer, New York, NY, 2008. [13] K.G. Rivette, D. Kline, Discovering new value in intellectual property, Harvard Bus. Rev. 78 (1) (January–February 2000) 54–66. [14] E. Kahn, Patent mining in a changing world of technology and product development, Intellect. Asset Mgmt (July/Aug. 2003) 7–10. [15] W. Barrett, C. Price, T. Hunt, iProperty: Profiting from Ideas in an Age of Global Innovation, Wiley, Hoboken, NJ, 2008. [16] R. Simmer, The Patent Searcher's Dilemma: Where is the Value? How to Manage Costs Conducting Patent Searching for Competitive Intelligence, 2004 Accessible on line at http://www.patex.ca/pdf/publications/Patent%20Searchers%20Dilemma%20FPTT.pdf. [17] A.J. Trippe, Patinformatics: tasks to tools, World Patent Inf. 25 (3) (2003) 211–221. [18] H. Dou, V. Leveillé, S. Manullang, J.M. Jr Dou, Benchmarking R&D and companies through patent analysis using free databases and special software: a tool to improve innovative thinking, World Patent Inf. 26 (4) (2004) 297–309. [19] J.C. Vergara, A. Comai, J.T. Millan, Software for Technological Patent Intelligence: Evaluation of Software and Technological Intelligence Needs, EMECOM Ediciones, Barcelona, Spain, 2006. [20] N. Grandjean, B. Charpiot, C.A. Pena, M.C. Peitsch, Competitive intelligence and patent analysis in drug discovery: mining the competitive knowledge bases and patents, Drug Discovery Today: Technologies 2 (3) (2005) 211–215. [21] H. Dou, V. Leveillé, S. Manullang, J.M. Jr Dou, Patent analysis for competitive technical intelligence and innovative thinking, Data Sci. J. 4 (2005) 209–237. [22] A.L. Porter, D. Schoeneck, P.R. Frey, D.M. Hicks, D.P. Libaers, Mining the internet for competitive technical intelligence, Competitive Intelligence Magazine, Society of Competitive Intelligence Professionals, vol. 10 (5), 2007, pp. 24–28. [23] R Rohrbeck, J. Heuer, H.M. Arnold, The Technology Radar — An Instrument of Technology Intelligence and Innovation Strategy, The 3rd IEEE International Conference on Management of Innovation and Technology: Singapore, IEEE Conference Publishing, Piscataway, NJ, 2006, pp. 978–983. [24] K.D. Miller, A.T. Arikan, Technology search investments: evolutionary, options reasoning, and option pricing approaches, Strateg. Manage. J. 25 (2004) 473–485. [25] J.S. MacNeil, Guide to a beta life, Genome technol. (2004) 24–27.
Mark Veugelers is Integration Manager at VIB since August 2004. In this position, he is responsible for stimulating knowledge and technology diffusion throughout VIB. He has a Ph.D. in medical sciences from the K.U. Leuven and an MBA from Vlerick Leuven Gent Management School. Jo Bury is Managing Director of VIB, the Flanders Institute for Biotechnology. VIB, established as a joint venture between 4 Flemish Universities in 1996, counts 1100 scientists dispersed over 6 Departments. He has a Ph.D. from Ghent University and an MBA from Vlerick Leuven Gent Management School. Stijn Viaene is an associate professor of management and IT at K.U. Leuven, Belgium, and a partner of the Vlerick Leuven Gent Management School, where he heads the Competence Center Operations and Technology Management. He is holder of the Vlerick Research Chair on IT and corporate performance management endowed by Deloitte, and is in charge of the Vlerick Research Center on business intelligence sponsored by SAS Institute. His research interests include managerial issues in business-IT alignment and information systems management, with a particular interest in realizing the benefits from BI systems.
Contents lists available at ScienceDirect
Technological Forecasting & Social Change
Linking technology intelligence to open innovation Mark Veugelers a,⁎, Jo Bury a, Stijn Viaene b a b
VIB, Rijvisschestraat 120, 9052 Ghent, Belgium Vlerick Leuven Gent Management School & K.U.Leuven, Vlamingenstraat 83, 3000 Leuven, Belgium
a r t i c l e
i n f o
Article history: Received 20 April 2009 Received in revised form 3 August 2009 Accepted 18 September 2009 Keywords: Technology intelligence Open innovation Life sciences Real options reasoning Technology radar
a b s t r a c t The explosive growth of the Internet has led to a dramatic increase in data sources for (competitive) technology intelligence. Appropriate implementation and use of IT tools to gather and analyze these data is of key importance for the creation of actionable technology intelligence. A strategy to optimize investments in the identified technologies becomes of paramount importance if an organization wants to match knowledge and ideas originating from outside of the organization with internal core competences. Such a strategy can create competitive advantage by effectively linking technology intelligence to open innovation. We show how VIB, a life sciences research organization, has established technology intelligence processes to identify a multitude of external technologies of interest, which are subsequently “probed” for their potential and fit with VIB using real options reasoning, thereby supporting open innovation. Our methodology may be useful for other organizations which are considering implementing open innovation approaches. © 2009 Elsevier Inc. All rights reserved.
1. Introduction Changing technology, driven forward by relentless innovation, affects everybody's business. Smart organizations do not wait for change to happen but actively monitor and take advantage of changing environments and new innovations. “Open innovation” approaches, which implement externally developed technologies or ideas within an organization, have been proposed as a remedy for decreasing product half-lives, faster technology cycles, and increased global competition [1]. To enable open innovation, innovation intermediaries such as online technology transfer exchanges, which list technologies available for licensing to interested parties, or technology brokers, who solicit new innovations by posting problems requiring a solution, have emerged to address specific technology needs. However, these intermediaries typically concentrate only on a fraction of the technology intelligence relevant to an organization's strategy. Technology intelligence has been defined as “the capture and delivery of technological information as part of the process whereby an organization develops an awareness of technological threats and opportunities” [2]. To identify, as comprehensively as possible, all options for introducing new, external technologies/innovations in the organization, it becomes necessary to analyze large amounts of technology data [3], originating from disparate sources outside the organization. Use of appropriate IT tools and extensive data mining analyses, such as text mining [4], can generate actionable technology intelligence. Technology intelligence can have many uses (e.g. in strategy, marketing or human resources). In this paper, we will show how we have generated technology intelligence that allows for the systematic identification of externally developed disruptive technologies, which are probed for their potential value to and fit with our organization, using real options reasoning based technology investments. Our work shows how technology intelligence can be linked to open innovation in practice. ⁎ Corresponding author. Tel.: + 32 9 2446611; fax: + 32 9 2446610. E-mail addresses: [email protected] (M. Veugelers), [email protected] (J. Bury), [email protected] (S. Viaene). 0040-1625/$ – see front matter © 2009 Elsevier Inc. All rights reserved. doi:10.1016/j.techfore.2009.09.003
336
M. Veugelers et al. / Technological Forecasting & Social Change 77 (2010) 335–343
2. VIB: focus on excellence in life sciences research and technology transfer VIB (http://www.vib.be) is a Belgian non-profit life sciences research institute established in 1996. A total of 65 research groups, encompassing more than 1100 scientists, are distributed over six research campuses. Using advanced gene technologies, VIB researchers study a wide variety of biological processes in the human body, in plants and in micro-organisms. VIB focuses on excellence in science and in technology transfer. Technology transfer of in-house generated inventions to external parties and start-ups has been very strong at VIB. However, the adoption of technologies developed outside VIB has been suboptimal in certain cases (e.g. the technology was acquired too late compared to the competition or the acquisition was too expensive compared to other options). In these cases, technology acquisition decisions were frequently made on an ad hoc, reactive basis by individual research groups, who did not sufficiently consider institutional strategy and synergies or, in some cases, alternative technologies. Such decisions could be explained as a legacy of VIB's origins and structure, as VIB was set up as a joint venture of individual research departments. Since several departments often acquired identical, expensive scientific instruments, leading to excess capacity, it became clear that an institutional technology management strategy could improve technology acquisition, while simultaneously optimizing technology adoption and spotting of new technology opportunities. To this end, several barriers had to be overcome, such as the lack of appropriate processes to scan for novel technology opportunities and evaluate technology acquisitions from an institutional perspective, the preference for conservative risk aversion versus opportunity recognition for novel technologies, and the lack of information on the progress of technology adoptions by other organizations. These barriers significantly slowed down the acquisition of several platform technologies, even when these technologies had already established a low-risk profile and proven their potential to early adopters in other organizations. 3. A framework for technology intelligence at VIB Several life sciences technologies have had a dramatic impact on scientific productivity since their introduction [5]. These enabling technologies led to bursts of scientific discoveries, publications, patents and commercial activity. A system to gather early intelligence on disruptive technologies and a correct timing of access or acquisition might facilitate successful adoption of these technologies, leading to a competitive advantage. A retrospective analysis of VIB technology adoptions, as well as of VIB decisions not to adopt, identified several processes that were lacking or could be improved in order to support both external technology identification and access to the identified technologies. In 2006, VIB's management therefore decided to support the establishment of technology intelligence processes to identify technologies of interest to VIB. However, in order to make the intelligence actionable for technology sourcing, VIB's management also decided to allocate a limited budget to facilitating access to the identified technologies. Linking actionable technology intelligence to a technology investment strategy could potentially facilitate VIB's technology management. As the time dimension is important in technology intelligence, our aim was to identify and access fast-moving disruptive technologies in a timely manner. Early access to the right high-potential, high-risk technologies could offer VIB the opportunity to increase its scientific and intellectual property output. 4. Technology intelligence identifies technology opportunities At the outset, we identified two key questions to address: - What is possible with technology: which technologies are out there? - What technology is valuable and needed by our organization: which technologies fit with our organization? In order to answer the question “What is possible with technology?” we needed to gather and organize information on all current life sciences technologies (current state of the art). These “baseline technologies” form the basis for the development of a technology intelligence process that allows for the continuous, cheap and efficient updating of the organization on technologies with accelerating growth rates as well as for the detection of signals of new, emerging technologies. To ensure comprehensive identification of technologies developed by both recent start-ups and large companies, we perform worldwide, continuous technology scanning. In the life sciences field, there are thousands of companies that develop tools and instruments of potential interest to VIB, with the number of academic research groups being a multiple of this number. In addition to scanning for technologies, we also decided to scan for technology adoptions by other organizations (competitive technology intelligence). Establishment of a (competitive) technology intelligence process, allowing for the collection of real-time information on all technologies of interest and their state of adoption by other organizations, requires distinct steps, defined as the ‘technology intelligence cycle’ (Fig. 1). Five stages can be distinguished in the technology intelligence cycle (adapted from [6]): • Planning and management: consulting with decision makers to define which intelligence should be produced [for VIB: (new) life science technologies with disruptive potential]. • Identification of sources: locating sources for data collection [for VIB: academic literature (Pubmed), patents and the Internet]. • Primary source collection: actual capturing of the data from the sources [for VIB: use of IT tools and databases such as ThomsonISI Web of Science, Matheopatent and Matheoweb].
M. Veugelers et al. / Technological Forecasting & Social Change 77 (2010) 335–343
337
Fig. 1. The technology intelligence cycle at VIB.
• Analysis and production: the primary data, consisting of thousands of documents, is analyzed for its usefulness and filtered to facilitate the analysis. Filtered data is transformed into actionable technology intelligence [for VIB: a ‘technology scout’ performs data mining using software tools and synthesizes conclusions into technology intelligence profiles]. • Reporting and information: presentation of the findings to decision makers and decision support for strategy. Feedback of decision makers is used to fine-tune the cycle [for VIB: technology intelligence profiles presented to management and employees through seminars and Intranet]. The output of the technology intelligence process for VIB would be a set of technology intelligence profiles on new disruptive life sciences technologies of interest to VIB. Access to these technologies and/or co-development with VIB could lead to competitive advantage by facilitating the generation of new intellectual property, start-up companies with new business models, or joint ventures (open innovation in practice). In order to optimize matching our internal technology needs and absorptive capacity with the technology intelligence obtained on external technologies, we opted to start on a small scale with a single, internal ‘technology scout,’ allocated part-time, with an initial focus on gathering science and technology data and performing analyses leading to the production of technology intelligence.
5. IT tools and databases support the technology intelligence cycle Establishing technology intelligence processes from scratch can be a serious challenge. Devising technology identification strategies is not a simple task, and scratching only the surface can be dangerous, as it may lead to overlooking key technologies of strategic importance to the organization [8]. The usual channels (market reports, suppliers, etc.) for identifying new technology opportunities are usually already well known to the organization, but these are far removed from a systematic process that continuously updates the intelligence on technologies of interest as well as identifies off-radar technologies, currently unknown and outside the normal field of view. In the case of the latter, “You don't know what you don't know” too often gets in the way of acquiring information about technologies currently outside the radar of the organization. Effective search strategies can be a source of competitive advantage [7], and the smart use of search engines such as Google may turn up interesting results. However, the Internet is growing so large that quick keyword searches at low search depth are likely to miss a number of technologies of interest to the organization. Data sources for technology intelligence contain millions of documents and to automate the collecting, processing and analysis of millions of textual data sources, one needs appropriate IT tools. To select the right tools, an organization needs to understand the role and possibilities of IT in the technology intelligence process, as this will allow the identification, selection and installation of the appropriate, cost-effective software [9,10]. Until a few years ago, the market for IT tools supporting technology intelligence was considered a niche market with just a few tools available. The accelerating growth of data on the Internet has fuelled the growth of a wide variety — in terms of their price, capabilities and focus — of software tools for multiple aspects of technology intelligence [6]. Software packages as well as consulting services are available for data mining, text retrieval and classification, semantic analysis and document summarization, patent searching, webpage tracking, and Internet monitoring using syndication technology. The situation is made more complex by software able to cover single or multiple steps of the technology intelligence cycle. An incorrect implementation of these tools may either lead to ‘information overload’ or to ‘myopia,’ due to the non-capture of essential information. As no monitoring system or software is perfect, a decision will have to be made on the trade-off between price and performance of the software and between its levels of precision and recall of technology data. Luckily, an organization with no technology intelligence processes in place can get one of the ground at a relatively low cost, as there are a variety of reasonably inexpensive yet highly efficient technology intelligence software packages available, which can be installed on any desktop computer.
338
M. Veugelers et al. / Technological Forecasting & Social Change 77 (2010) 335–343
In addition to software for technology intelligence, several databases, such as the Science Citation Index from Thomson-ISI, compile information on academic research, while information on technological innovations is available through publicly accessible patent databases (i.e. WIPO, USPTO, EPO, JPO), as well as from commercial service providers (e.g. Thomson, STN, etc.). Several tools (e.g. LexisNexis, Dialog, Google Alerts, Twitter Alerts, etc.) are also available for tracking companies and their press releases as well as new technology developments. Several of these service providers charge access fees, which may be well worth the money as they avoid wasting time on analyzing unstructured and inferior-quality data. VIB started with test-running a few software packages and databases to identify a combination of tools offering a good fit with the organization and high precision and recall. We acquired low-cost technology intelligence software and subscribed to a few commercial database providers to be well positioned for technology data mining. 6. Data-mining software processes data collected from information sources Hundreds of thousands of research articles and patents are published each year, and an immense wealth of new ideas is being made available on the Internet on a daily basis through media such as press releases about product launches, etc. All this information is furthermore growing exponentially in size, making it nearly impossible to stay abreast of rapid changes in technology landscapes without a structured approach to searching and scanning this data overflow. Buried within this data are gold nuggets: disruptive technologies that present as a weak signal in a background of noise and are therefore difficult to detect and interpret. Hence, the technology intelligence process needs to be designed such that it facilitates a broad scan of the periphery while being equipped with the appropriate filters that will allow identification of weak signals from disruptive technologies [8]. This involves a delicate balance between an appropriate filter/cut-off and an increase in analysis workload. The sheer number of technologies, companies and applications to track, and the identification of weak signals from disruptive technologies can stretch scarce resources even thinner, leaving organizations unable to respond quickly to novel research discoveries and technology opportunities. To identify technologies of interest to VIB, a workflow was set up to acquire and analyze large, textual datasets from three major sources of new technology information: academic literature, patents and the Internet (Fig. 2). The general methodology for technology identification was similar for all three data sources. Specific keywords from a ‘technology thesaurus’ — a set of keywords on life sciences technologies — were used for querying the data source. The returned data (abstracts, patent claims and webpage text) was subsequently downloaded with appropriate software, and stored either in text format or in an Excel database for further analysis and filtering. In order to reduce the size of the large dataset, Omniviz software was used for clustering texts (there are other packages with similar functionalities available, such as VantagePoint [11]). Several types of queries were performed: broad searches for macro-level technology landscape analysis as well as more detailed queries for specific technologies, inventors, companies or universities. As each data source has different properties, different strategies are needed to maximize information harvesting from each source: a) Academic literature The academic literature (~ 600,000 new publications a year in the life sciences) contains a wealth of information on peerreviewed publications describing technologies of interest. We scanned the academic literature using Pubmed and Thomson-ISI Web of Science. By searching for abstracts using our technology thesaurus, it became possible to filter the literature for technology-related publications. Abstracts were clustered with Omniviz into individual technology clusters (Fig. 3). While some technology management systems focus exclusively on text mining [12], we also included citations as a key technology indicator. As citations
Fig. 2. Workflow for identifying technologies of interest.
M. Veugelers et al. / Technological Forecasting & Social Change 77 (2010) 335–343
339
are indicative of the innovativeness of a technology, temporal analysis of the most frequently cited publications offered an insight into which technologies were considered critical breakthroughs. Temporal analysis, iterative clustering, identification of authors with company affiliations, and other detailed analyses identified novel technology developments. Taken together, the combination of citation analysis and clustering into technology fields identified several technology opportunities of interest to VIB. b) Patent literature The patent literature (~150,000 new patents published every year) contains key information on novel technologies and inventions [13]. Patent mining requires the analyst to actively scan and analyze all patents that have the potential to affect the organization [14]. A patent-mining strategy for the entire field of the life sciences (and converging science areas such as nanobiotechnology, bioinformatics, etc.) requires the analysis of large datasets, necessitating in turn effective mining, visualization and filtering strategies to arrive at a workable set of data and downstream actions [15]. Software analysis alone cannot be a replacement for a subject analysis of a patent. To ensure and assess the efficiency of the filtering step, it becomes necessary to review part or all of the text, especially the breadth of the claims. Devising a cost-effective and efficient approach to analyzing patent data is also challenging due to the availability of a broad spectrum of database and software resources for the collection of reliable data [16]. Several patent tools and software packages are available to assist in patent analysis, offering tools such as automated patent text downloads, categorization, integration and visualization [17–20]. Patent abstracts of interest were clustered using Omniviz for more in-detail analysis. Patent citation analysis as well as sizing of patent families was performed with Matheopatent software [21]. c) Internet Technology intelligence has traditionally focused primarily on data from patent and research publication abstract databases [22]. This resource base can be extended with data mined from the Internet. Although patent and academic publication data can also be obtained from the Internet, the latter also offers a wealth of additional technology information via company websites, press releases, blogs, message boards, etc. not found in the patent and literature databases. Although rich in content (45 billion web pages and doubling each year), there is a drawback: textual information on Internet websites is presented in a wide variety of text structures. We scanned for technologies of interest on the Internet using Matheoweb software. By using targeted keyword searches and software to extract information from thousands of web pages automatically, we were able to identify a number of technologies with disruptive potential. Data from all three search domains was stored in an Excel data warehouse for further processing and clustering. 7. Transforming data into actionable intelligence In a final step, the most complex one in the technology intelligence process, data is transformed into actionable intelligence. This step is not very suitable for automation and requires a significant human effort: an understanding of the technology and the end user's or organization's requirements with respect to the information is the key. In our case, the meta-analysis from the three different data sources identified intelligence on fast-moving technology fields as evidenced by topic bursts within clusters of technologies. The search strategy also identified multiple new technologies of interest to VIB. Access to these technologies could accelerate research programs at VIB, thereby offering competitive advantage and high returns on technology investments. However, a strategy to minimize the risk of purchasing mistakes and/or investing in unknown technologies with low end user potential, leading to excess capacity, would be very helpful.
Fig. 3. Clustering of life sciences technology abstracts from the academic literature (publication years 2004–2008). A Pubmed search for life sciences technologies and methods revealed hundreds of abstracts, which were downloaded, organized in a text database and fed into the Omniviz data mining tool. Omniviz returns clustered abstracts, with keywords defining each cluster and the number of abstracts for each cluster. The figure shows the clustering of all abstracts. A cluster can group several technologies. Iterative clustering, using for example abstracts belonging to a single cluster, can be used to obtain more granularity.
340
M. Veugelers et al. / Technological Forecasting & Social Change 77 (2010) 335–343
In order to support decision-making and present the identified technologies to VIB's management, we decided to use the technology radar framework in use at Deutsche Telekom AG [23]. A technology radar offers a visual overview of the information obtained from the technology intelligence process. The technology radar screen is a powerful format for the aggregation of technology intelligence on broad technology fields, categorized in different taxonomies, and enables the management of a large number of technologies simultaneously in a single view. Compared to newsletters or technology briefings, which typically would be sent out as individual technology topics over a period of several months, thereby diluting the information over a time period, the technology radar aggregates all data at once. As we identified dozens of technologies of interest to VIB, it became necessary to apply filters that would ensure that the VIB technology radar was populated with the technologies of most interest to VIB scientists. Therefore, an internal Technology Watch team made up of VIB scientists and members of VIB's management was set up to prioritize technologies. The VIB scientists on the Tech Watch team, together with their departmental colleagues, monitor technologies closely related to their expertise. They form the nucleus of an ‘intelligence network’ distributed throughout the organization. The Technology Watch team also contains two members of VIB's management: one patent expert and one technology scout. The Technology Watch team is not full-time allocated to the technology intelligence process but regularly convenes to review and discuss the technologies identified on the VIB technology radar screen (Fig. 4). The Technology Watch team also generates a technology intelligence profile (Fig. 5) for each technology on the VIB technology radar screen. The technology intelligence profile is a review-style one-page document featuring a broad summary of the technology, including key patents, publications, and Internet intelligence data (market status, usage by early adopters, etc.). The Technology Watch team generated over 50 technology intelligence profiles. Each of these highlighted specific technologies or instruments in development at a life sciences company and/or showing accelerating technology capability. These technology intelligence profiles were made available to the entire organization on the VIB Intranet. For presentation purposes, the profiles include mostly visual information and are limited to just a half page of text. The profiles also include links to patents, company websites and literature, and are an excellent tool to alert and inform VIB scientists about disruptive technologies identified through the technology intelligence process. 8. Creating actionable intelligence: assessing technology value and fit with VIB The VIB technology radar allowed us to visualize several options for investment in disruptive technologies, making it evident that the acquired technology intelligence could be a major tool to support technology sourcing at VIB. However, the cumulative cost of investing in all technologies featured on the VIB technology radar would greatly exceed the annual budget allocated to accessing identified technologies. To prioritize investments in the technologies with the highest potential for VIB, we tried to use a technologyrating matrix, ranking technologies according to potential impact and implementation risk. After numerous discussions, this approach was abandoned as the rating of technologies, especially of high-potential/high-risk technologies, was seen as too complex. To reduce the inherent risk of investing in high-risk/high-reward disruptive technologies, we opted for a real options reasoning approach [24]. This approach entails investing in several research projects, which use the technologies of interest, to obtain
Fig. 4. Technology radar for VIB. Meta-analysis of academic literature data, patent data and Internet data identified technology clusters with temporal topic bursts. These highly active technologies were visualized on a technology radar, which depicts technologies with their risk profile (blue: low risk, green: medium risk, red: high risk) and market stage. Technologies on the market are on the inner circle, technologies nearing market launch in the next circle, and technologies further away from daily lab use in the two outer circles.
M. Veugelers et al. / Technological Forecasting & Social Change 77 (2010) 335–343
341
Fig. 5. Technology intelligence profile. The technology intelligence profile is a review-style one-page document which focuses more in detail (key patents, publications, and Internet intelligence) on a technology present on the technology radar for VIB. This example highlights the technology “gene synthesis,” in the broad technology field “Synthetic Biology.” The profile starts with a technology snapshot, an in-house made, short 10-line overview of the technology. It is followed by a table which features quantitative technology indicators (no. of Pubmed hits (hits for query “gene synthesis,” “synthetic biology”), no. of companies (identified by patent assignees and Internet search), Google alert rate (increase in Google Alerts per year, comparing 2008 to 2004), no. of patents (2004–2008) (total number of WIPO patent applications filed between 2004 and 2008), patent acceleration rate (% increase in annual WIPO patent applications, comparing 2008 to 2004)). These indicators allow quantitative comparisons between different technologies. The technology indicators are followed by a concise summary of results from academic, patent and Internet data. Key publications highlight highly cited articles on “gene synthesis” technology. Key companies in the field of gene synthesis were identified by analyzing patent databases and querying the Internet. Internet keywords were generated by generating a tagcloud from all Google Alerts. Patent landscape was generated by analyzing patent databases for “gene synthesis” patents and assignees for these patents; number of patent applications is indicated. Recent Google Alert displays recent hits for a Google Alert which scans the web for new webpage hits on “gene synthesis.”
342
M. Veugelers et al. / Technological Forecasting & Social Change 77 (2010) 335–343
information on the potential of the technologies and their fit with VIB. The information obtained from these projects would then determine follow-on investment decisions. We reasoned that such an approach would also be of value for a company developing a technology, as it would allow the company to establish market intelligence as well as validation of their technologies by VIB scientists. Companies with disruptive technologies frequently look for validation of their technology through research collaborations or services which offer companies visibility, marketing opportunities and customer feedback [25]. We therefore set up a system of annual ‘Technology Grants,’ limited in budget and number. These Technology Grants allow VIB scientists to test the potential and fit of a fixed set of technologies, all identified through the technology intelligence process and featured on the technology radar. The Technology Grants function as real options, or Stage 1 investments. In case of the successful application of the technology in Stage 1, as evaluated by our Tech Watch team, the option is exercised through a larger, follow-on Stage 2 investment in the technology. If use of the technology by VIB scientists fails to deliver promising results in Stage 1, no follow-on investment is made at Stage 2 and the option is abandoned. Each year, VIB scientists can submit proposals for Technology Grants, with a budget allowing for 5–10 real options. A decision to move to Stage 2 is made within one year. This limited timeframe allows for the collection of sufficient information on the technology's potential to facilitate decision-making about exercising or abandoning the option. Information on the importance of the Technology Grants and technology intelligence as a tool to gain access to disruptive technologies as well as a summary of the intelligence on technologies of interest to VIB scientists was presented to all VIB employees through a seminar series. Subsequently a call for projects to be financed by Technology Grants was issued to all 65 VIB research groups. By working with a fixed set of accessible technologies and limiting the number of projects a group can submit for funding, we were able to answer the question “What is needed ?,” as the number of projects submitted for a particular technology functioned as ‘votes’ for that technology. 9. Results of the first iterations of the technology intelligence cycle The response of both scientists and VIB's management to the technology intelligence presented through the technology radar framework has been very positive. The Technology Watch team continues to receive numerous requests from VIB scientists for additional information on the identified technologies. The Technology Grants program has allowed us to establish numerous collaborations with other organizations letting us plug into their technologies (open innovation in action) and leading to an increased output in scientific publications and intellectual property (patent applications). Our Technology Grants program has also worked as an internal voting system, whereby technologies that attracted interest from a large proportion of research departments justified a technology acquisition by VIB. This has reduced the risk of investing in high-potential technologies for which interest would turn out to be limited, and whose capacity would therefore not be optimally used upon their acquisition. In several cases, results from accessing the technologies through Technology Grants projects were excellent, which led VIB scientists to continue investing in these technologies with their own funds. Some of the projects did not provide a fit with VIB scientists because the market strategy of some of the companies did not match the interests of the VIB scientists. However, obtaining this information was also valuable as these projects could be ‘killed off’ quickly. 10. Discussion Setting up a technology intelligence process aimed at identifying, prioritizing and exploiting external high-potential technologies through real options reasoning facilitates open innovation and strategic decision-making, leading to competitive advantage. Investing in external, early-stage or fast-moving technologies is inherently risky as the acquirer will always have incomplete information about the potential of the external technology and its fit with the acquiring organization. Consistent with this, the reaction to the spotting of a disruptive technology is often inertia and a wait-and-see strategy. Interviews with our scientists made it clear that this was often due to risk aversion: early-stage technologies were often seen as too risky and costly. In order to reduce risk, validation or adoption of a technology by peers was often required before a scientist decided to invest in a new technology. Another roadblock turned out to be the time, cost and effort required to establish relationships with external technology providers, an example being the administrative burden associated with accessing external technology (generating CDAs, MTAs, IP-issues, collaboration agreements, etc.). Correct timing in terms of identification, prioritization and technology adoption remains critical to reducing risk exposure. The real options reasoning approach provided by our Technology Grants program partially solves this problem. By investing in several technologies via a centrally administered budget, and thus taking options on multiple technologies, VIB can increase the information about the different technologies accessed during the option period and hence reduce its risk. Taking several options by accessing multiple technologies also increases flexibility. A well-functioning technology intelligence process, which presents multiple technologies as options on a technology radar, regularly updated and presented to top management, also leads to increased opportunity awareness. Knowledge of a broad portfolio of technology opportunities probably also reduces risk exposure, as the number of entry options increases. Technology options allow for the identification of technologies that are at the right stage in terms of their technology potential and fit with VIB, so that first-mover advantage for access to these technologies can be maximized.
M. Veugelers et al. / Technological Forecasting & Social Change 77 (2010) 335–343
343
In the past, several technologies were only assessed by VIB after they were on the market for several years. With the technology intelligence process and Technology Grants program in place, this has dramatically changed: new technologies are spotted early on through our intelligence process and a significant portion of ‘innovator risk’ is removed by allowing a real options reasoning approach to probing external technologies. Since the implementation of our technology intelligence process and Technology Grants program, VIB has invested in several external technologies from companies developing early-stage technologies not yet on the market. Moreover, discussion of the technology intelligence within VIB has greatly stimulated technology information awareness as well as early spotting of technologies of interest. Open innovation requires that organizations use external knowledge, ideas, technologies and Intellectual Property to accelerate innovation.VIB identifies external technology opportunities through technology intelligence and accesses these external technologies using a real options reasoning based approach. Access to external technologies and Intellectual Property is covered by negotiated research collaboration agreements and licenses that allow VIB scientists to pursue specific, innovative applications of a partner's technology. VIB scientists' use of these technologies has led to the creation of new Intellectual Property for VIB. VIB's patent portfolio is used for generating new biotechnology start-ups or to generate revenue via out-licensing of Intellectual Property to external partners. In conclusion, this paper frames the acquired technology intelligence in the context of a technology radar, from which external technologies are selected for investment through a real options based reasoning approach. Real options reasoning have previously been connected to open innovation in the context of corporate venturing and the creation of new business models. The approach detailed in this paper extends the use of real options reasoning in open innovation to technology sourcing based on technology intelligence. Our approach, which links technology intelligence to open innovation through real options reasoning, could be of prime interest to any organization aiming to increase the value of technology intelligence for open innovation. Pharmaceutical/biotechnology research labs and life sciences institutes in particular may benefit from following approaches similar to the ones described in this paper. References [1] H. Chesbrough, The era of open innovation, MIT Sloan Management Review 44 (3) (2003) 35–41 Spring. [2] C.I.V. Kerr, L. Mortara, R. Phaal, D.R. Probert, A conceptual model for technology intelligence, Int. J. Technol. Intel. Plan. 2 (1) (2006) 73–93. [3] D. Zhu, A.L. Porter, Automated extraction and visualization of information for technological intelligence and forecasting, Technol. Forecast. Soc. Change 69 (2002) 495–506. [4] R.N. Kostoff, D.R. Toothman, H.J. Eberhart, J.A. Humenik, Text mining using database tomography and bibliometrics: a review, Technol. Forecast. Soc. Change 68 (3) (2001) 223–253. [5] J.M. Rothberg, J.H. Leamon, The development and impact of 454 sequencing, Nat. Biotechnol. 26 (10) (2008) 1117–1124. [6] Fuld&Company, Intelligence Software Report 2008, 2008 Accessible on line at http://www.fuld.com/Products/ISR2008/. [7] J. Battelle, The Search: How Google and Its Rivals Rewrote the Rules of Business and Transformed Our Culture, Portfolio Trade, New York, NY, 2005. [8] G.S. Day, P.J.H. Schoemaker, Scanning the periphery, Harvard Bus. Rev. 83 (11) (November 2005) 135–148. [9] D. Vriens, The role of information and communication technology in competitive intelligence, in: D. Vriens (Ed.), Information and Communication Technology for Competitive Intelligence, Idea Group Publishing, Hershey, PA, 2004, p. 2003. [10] S. Gordon, M. Tarafdar, R. Cook, R. Maksimoski, B. Rogowitz, Improving the front end of innovation with Information Technology, Res. Technol. Manag. 51 (3) (2008) 50–58. [11] A.L. Porter, QTIP: quick technology intelligence processes, Technol. Forecast. Soc. Change 72 (2005) 1070–1081. [12] H. Chang, J. Gausemeier, S. Ihmels, C. Wenzelmann, Innovative technology management system with bibliometrics in the context of technology intelligence, in: O. Castillo, L. Xu, S.-I. Ao (Eds.), Trends in Intelligent Systems and Computer Engineering; Lecture Notes in Electrical Engineering, Springer, New York, NY, 2008. [13] K.G. Rivette, D. Kline, Discovering new value in intellectual property, Harvard Bus. Rev. 78 (1) (January–February 2000) 54–66. [14] E. Kahn, Patent mining in a changing world of technology and product development, Intellect. Asset Mgmt (July/Aug. 2003) 7–10. [15] W. Barrett, C. Price, T. Hunt, iProperty: Profiting from Ideas in an Age of Global Innovation, Wiley, Hoboken, NJ, 2008. [16] R. Simmer, The Patent Searcher's Dilemma: Where is the Value? How to Manage Costs Conducting Patent Searching for Competitive Intelligence, 2004 Accessible on line at http://www.patex.ca/pdf/publications/Patent%20Searchers%20Dilemma%20FPTT.pdf. [17] A.J. Trippe, Patinformatics: tasks to tools, World Patent Inf. 25 (3) (2003) 211–221. [18] H. Dou, V. Leveillé, S. Manullang, J.M. Jr Dou, Benchmarking R&D and companies through patent analysis using free databases and special software: a tool to improve innovative thinking, World Patent Inf. 26 (4) (2004) 297–309. [19] J.C. Vergara, A. Comai, J.T. Millan, Software for Technological Patent Intelligence: Evaluation of Software and Technological Intelligence Needs, EMECOM Ediciones, Barcelona, Spain, 2006. [20] N. Grandjean, B. Charpiot, C.A. Pena, M.C. Peitsch, Competitive intelligence and patent analysis in drug discovery: mining the competitive knowledge bases and patents, Drug Discovery Today: Technologies 2 (3) (2005) 211–215. [21] H. Dou, V. Leveillé, S. Manullang, J.M. Jr Dou, Patent analysis for competitive technical intelligence and innovative thinking, Data Sci. J. 4 (2005) 209–237. [22] A.L. Porter, D. Schoeneck, P.R. Frey, D.M. Hicks, D.P. Libaers, Mining the internet for competitive technical intelligence, Competitive Intelligence Magazine, Society of Competitive Intelligence Professionals, vol. 10 (5), 2007, pp. 24–28. [23] R Rohrbeck, J. Heuer, H.M. Arnold, The Technology Radar — An Instrument of Technology Intelligence and Innovation Strategy, The 3rd IEEE International Conference on Management of Innovation and Technology: Singapore, IEEE Conference Publishing, Piscataway, NJ, 2006, pp. 978–983. [24] K.D. Miller, A.T. Arikan, Technology search investments: evolutionary, options reasoning, and option pricing approaches, Strateg. Manage. J. 25 (2004) 473–485. [25] J.S. MacNeil, Guide to a beta life, Genome technol. (2004) 24–27.
Mark Veugelers is Integration Manager at VIB since August 2004. In this position, he is responsible for stimulating knowledge and technology diffusion throughout VIB. He has a Ph.D. in medical sciences from the K.U. Leuven and an MBA from Vlerick Leuven Gent Management School. Jo Bury is Managing Director of VIB, the Flanders Institute for Biotechnology. VIB, established as a joint venture between 4 Flemish Universities in 1996, counts 1100 scientists dispersed over 6 Departments. He has a Ph.D. from Ghent University and an MBA from Vlerick Leuven Gent Management School. Stijn Viaene is an associate professor of management and IT at K.U. Leuven, Belgium, and a partner of the Vlerick Leuven Gent Management School, where he heads the Competence Center Operations and Technology Management. He is holder of the Vlerick Research Chair on IT and corporate performance management endowed by Deloitte, and is in charge of the Vlerick Research Center on business intelligence sponsored by SAS Institute. His research interests include managerial issues in business-IT alignment and information systems management, with a particular interest in realizing the benefits from BI systems.