Reputation for technological innovation: Does it actually cohere with innovative activity?

Höflinger, Patrick J.; Nagel, Christian Siegfried Anton; Sandner, Philipp Article Reputation for technological innovation: Does it actually cohere with innovative activity? Journal of Innovation & Knowledge (JIK) Provided in Cooperation with: Elsevier Suggested Citation: Höflinger, Patrick J.; Nagel, Christian Siegfried Anton; Sandner, Philipp (2018) : Reputation for technological innovation: Does it actually cohere with innovative activity?, Journal of Innovation & Knowledge (JIK), ISSN 2444-569X, Elsevier, Amsterdam, Vol. 3, Iss. 1, pp. 26-39, https://doi.org/10.1016/j.jik.2017.08.002 This Version is available at: https://hdl.handle.net/10419/190727 Standard-Nutzungsbedingungen: Die Dokumente auf EconStor dürfen zu eigenen wissenschaftlichen Zwecken und zum Privatgebrauch gespeichert und kopiert werden. Sie dürfen die Dokumente nicht für öffentliche oder kommerzielle Zwecke vervielfältigen, öffentlich ausstellen, öffentlich zugänglich machen, vertreiben oder anderweitig nutzen. Sofern die Verfasser die Dokumente unter Open-Content-Lizenzen (insbesondere CC-Lizenzen) zur Verfügung gestellt haben sollten, gelten abweichend von diesen Nutzungsbedingungen die in der dort genannten Lizenz gewährten Nutzungsrechte. Terms of use: Documents in EconStor may be saved and copied for your personal and scholarly purposes. You are not to copy documents for public or commercial purposes, to exhibit the documents publicly, to make them publicly available on the internet, or to distribute or otherwise use the documents in public. If the documents have been made available under an Open Content Licence (especially Creative Commons Licences), you may exercise further usage rights as specified in the indicated licence. https://creativecommons.org/licenses/by-nc-nd/4.0/ Journal of Innovation & Knowledge 3 (2018) 26–39 Journal of Innovation & Knowledge https://www.journals.elsevier.com/journal-of-innovation-and-knowledge Empirical paper Reputation for technological innovation: Does it actually cohere with innovative activity? Patrick J. Höflingera,∗, Christian Nagelb, Philipp Sandnerc aTUM School of Management, Technische Universität München (TUM), Munich, Germany bChair of Marketing and Innovation; University of Passau, Passau, Germany cFrankfurt School of Finance & Management, Frankfurt am Main, Germany a r t i c l e i n f o Article history: Received 3 January 2017 Accepted 23 August 2017 Available online 27 October 2017 JEL classification: M01 M31 O30 O31 O32 Keywords: Technological innovation Innovative performance Reputation Patents a b s t r a c t Firms strategically promote, foster and pursue a reputation for technological innovation. Yet, present research did not examine whether such perception actually coheres with innovative activity itself. Previous studies in this field often relied solely on tangible products/product introductions, hence we apply multiple proxies based on a firm’s intangible innovative performance to examine the influences on reputation for technological innovation. Using patent, financial and consumer data our Poisson regression analyzes 65 international firms which have been nominated by 231 consumers. We apply time-series and likewise cross-sectional data for our interdisciplinary analysis. Our findings demonstrate that innovative performance (citation intensity) is linked to reputation for technological innovation. Counter-intuitively, our results provide evidence that marketing intensity negatively influences a reputation for technological innovation. From the results, we conclude that innovative performance may be associated with a reputation for technological innovation. Actual technological advancement attracts attention from consumers that cannot be purchased with greater marketing investments. This implies that consumers appreciate the costly and uncertain R&D efforts and value those firms that constantly offer innovation. As a theoretical implication, the consideration of intangible inputs for reputation research is an important contribution for a holistic understanding. The results represent essential strategic information for innovation and marketing functions, where both divisions need to align their activities and investments. © 2017 Journal of Innovation & Knowledge. Published by Elsevier Espa˜ na, S.L.U. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). “You can’t buy a good reputation, you must earn it” (Harvey Mackay, Businessmen and Author) Introduction Can consumers assess the innovative performance of firms? There are strong signs suggesting that the subjective perception of innovative performance differs from objectively measurable performance criteria: The study “Navigating the Digital Future: The Global Innovation 1000” by Jaruzelski, Loehr, and Holman (2013) analyzed the top R&D spenders worldwide. In the same study, senior managers also ranked the most innovative companies according to their own perceptions. Exemplarily, Apple Inc. was perceived as most innovative company, however ranked on position 43 based on their R&D spending. Studying whether innovative performance and reputation cohere is legitimate and urgent, ∗Corresponding author. E-mail address: p.hoefl[email protected] (P.J. Höflinger). particularly when considering the benefits of a good reputation (Boyd, Bergh, & Ketchen, 2010; Henard & Dacin, 2010). Most papers in this field investigated reputation as a resource of competitive advantage with effects on financial results (Roberts & Dowling, 2002) or shareholder value (Raithel & Schwaiger, 2014). Fewer studies researched how firms gain reputational status and how reputation as a signal contributes to defending a strategic, competitive position (Shamsie, 2003). Basdeo, Smith, Grimm, Rindova, and Derfus (2006) found that reputation is substantially affected by both a firm’s own and its rivals’ activities. Henard and Dacin (2010) specifically investigated the development of reputation for product innovation, finding that consumers increase their loyalty, their excitement and their tolerance for failure. Recently, Stock and Zacharias (2013) found positive effects of product meaningfulness and negative effects of product newness on customer loyalty. Although the topic is relevant for research and practice, prior papers have been limited. First, previous research on the reputation for technological innovation addressed “innovative performance in the narrow sense” (Freeman & Soete, 1997; Hagedoorn & Cloodt, https://doi.org/10.1016/j.jik.2017.08.002 2444-569X/© 2017 Journal of Innovation & Knowledge. Published by Elsevier Espa˜ na, S.L.U. This is an open access article under the CC BY-NC-ND license (http:// creativecommons.org/licenses/by-nc-nd/4.0/). P.J. Höflinger et al. / Journal of Innovation & Knowledge 3 (2018) 26–39 27 2003), i.e. product innovations. Henard and Dacin (2010) developed a specific “reputation for product innovation”, showing that firms yield a specific reputation for introducing product innovations to the market. A shortcoming is the focus on “innovative performance in the narrow sense” (Hagedoorn & Cloodt, 2003), by focusing on products/product introductions which are tangible and visible aspects (Freeman & Soete, 1997; Henard & Szymanski, 2001; Szymanski, Kroff, & Troy, 2007). Henard and Dacin (2010, p. 322) in particular state that “one cannot simply rely on a proxy (e.g. number of patents filed, R&D dollars spent) to adequately capture the true measure of an organization’s reputation for product innovation”. We extend this by using multiple proxies based on patents, its citations and R&D spending. Patents are often considered to be the ultimate and accepted measure of an organization’s innovative output (Griliches, 1990; Hirshleifer, Hsu, & Li, 2013). In many high-tech industries, innovative products new to markets are often based on patents (Battisti, Gallego, Rubalcaba, & Windrum, 2015) as they protect innovation which is often economically relevant and thus actively commercialized (Eusebi & Silberglitt, 2014). Therefore, patents are established intellectual property rights used by firms to sustain competitive advantage (Hall, 1993). Today, it remains unclear whether these two factors for competitive advantage, innovative performance and reputation for technological innovation, are empirically associated. Considering its market benefits and its influence on long-term competitiveness (Hörner, 2002; Raithel & Schwaiger, 2014), the question of whether innovative performance can yield effects on reputation needs to be researched. Second, we examine the influences of marketing investments in this regard. Scholars find that companies, which deploy more resources in marketing are thus able to better preserve innovation (Chandy & Tellis, 2000). The possibility of a firm to strategically position itself in a favorable light reduces the constituents’ search for information (Fombrun & Shanley, 1990). For example, Samsung Electronics spent 14 billion US-Dollars on advertising in 2013 (Gilbert, 2013). However, truly innovative firms, often younger and new to the market thus have scare resources to invest in marketing. The establishment of their technological reputation and its strategic positioning may rely rather on their ability to develop and introduce cutting edge research and development than on marketing efforts. Small firms can be efficient in coordination, being able to answer and adapt rapidly to the market with new and innovative products (Artz, Norman, Hatfield, & Cardinal, 2010). Such technology lead “is like a high tree in a small forest; since it gets more sun, it will grow faster and stay taller” (Wernerfelt, 1984, p. 174). Since previous research did not concentrate on patents and R&D in this setting, we aim to find whether placing marketing cues, e.g. to demonstrate technological superiority for consumers, may be advantageous to achieve gains for a technological reputation. Third, there is a methodological gap because research in innovation management usually applies variables such as patents or R&D spending, whereas consumer research uses psychological measures (Archibugi & Planta, 1996). Our interdisciplinary approach applies specific algorithms to connect consumers’ perceptions of innovative performance to the corresponding innovative performance of firms (cf. Nagel, Höflinger, & Welpe, 2014). We apply patent information from the worldwide patent statistical database (PATSTAT1) provided by the European Patent Office. This study relies on a fact-based evaluation rather than survey-based constructs for the assessment of innovative performance. This enables and justifies the pressing research quest for actual innovative activ1Available from the OECD-EPO Task Force on Patent Statistics: http://www.epo.org/searching/subscription/raw/product-14-24.html. ity. With the combination of both approaches, it is possible to derive a comprehensive perspective on how perceptions and reality cohere. In addressing the aforementioned gaps, this paper yields the following contributions. We add to current reputation research by showing that reputation for technological innovation is actually associated to the innovative performance of companies. We find significant support that inventive performance as measured by citation intensity is related to reputation for technological innovation. The higher the citation intensity, usually employed to identify patent quality and value (Hall, Jaffe, & Trajtenberg, 2005), the better the corresponding reputation. Thus, we are able to show that innovative performance can be positively associated with the consumer’s perception of innovative entities. This insight is new to current reputation and innovation research because prior studies mainly investigated “innovative performance in the narrow sense”, by exclusively using products/product introductions (Freeman & Soete, 1997; Hagedoorn & Cloodt, 2003; Henard & Dacin, 2010). Above, we have significant evidence that marketing intensity is negatively related to reputation for technological innovation. In the case for a technological reputation, the result suggest that marketing investments should be rather spent on “true” technological advancements rather than on advertising to achieve gains in technological reputation. Further, the incorporation of intellectual assets as patents and citations is an important step toward a better understanding of knowledge and a technological reputation. The remainder of this paper is divided as follows. The next chapter provides a theoretical background to develop the model and its hypotheses. Our description of methods introduces data and measures. Then, we provide additional statistics and a test of our hypotheses in the results section to critically discuss the findings. After outlining the implications of this research, we suggest opportunities for further research. Theoretical background Reputation Reputation can be viewed as the sole most appraised organizational resource (Gibson, Gonzales, & Castanon, 2006; Walker, 2010). Similarly, Hall (1993) discovered that chief executive officer’s classify reputation as the most relevant intangible asset. It is also widely accepted that reputation and competitive advantage cohere (Fombrun & Shanley, 1990; Hall, 1993; Roberts & Dowling, 2002), and multiple scholars found that reputation and firm performance are linked (Brown & Perry, 1994; Deephouse, 2000; Fombrun & Shanley, 1990; Walker, 2010). In today’s competitive environments, the role of reputation is remarkably essential (Abimbola & Vallaster, 2007). Generally, companies compete for reputational status as they compete for customers (Fombrun & Shanley, 1990), because reputation represents an important intangible resource for competitive advantage (Hall, 1993; Roberts & Dowling, 2002; Walker, 2010). A good reputation has multiple strategic benefits, because it helps firms to distinguish themselves from rivals (Peteraf, 1993), and it decreases information asymmetry and uncertainties of the consumer (Boyd et al., 2010; Weigelt & Camerer, 1988). Moreover, a good reputation increases overall firm image, the loyalty to the firm, the tolerance for failure and the propensity to pay premium prices (Henard & Dacin, 2010). Reputation is key in understanding why certain firms outperform other organizations (Boyd et al., 2010), which is viewed by many as the fundamental question of the strategic management field (Hitt, Boyd, & Li, 2004; Rumelt, Schendel, & Teece, 1994). 28 P.J. Höflinger et al. / Journal of Innovation & Knowledge 3 (2018) 26–39 Reputation for technological innovation Reputation is manifold (Rindova, Williamson, Petkova, & Sever, 2005), meaning that organizations operate in diverse settings and address multiple groups that evaluate signals differently. Lange, Lee, and Dai (2011) developed a general conceptualization that organizations are either “being known”, “being known for something” or are regarded as “generally favorable”. For instance, accounting signals can be used to assess corporate performance, whereas strategic signals evaluate the relative position of companies in competitive environments. Similarly, there are multiple ways a company can signal quality to consumers. Potential signals of product quality include e.g. price, brand name, warranty and advertising expenditures (Kirmani & Rao, 2000). Within this study, we concentrate on organizations that are “known for something” (Lange et al., 2011), more specifically for reputation for technological innovation. Such reputation entails perceptions regarding a very specific and interesting characteristic for the observer (Lange et al., 2011). “Innovation” is generally associated with positive ideas such as change, newness, improvement, value and benefits (Courtright & Smudde, 2009). Schumpeter (1934) finds that innovative organizations are regarded as advanced, active and willing to take risks. In other words, innovation covers a set of characteristics that are valuable to possess within competitive markets. Reputation for technological innovation is one of these multifaceted constructs of reputation (Rindova et al., 2005). Every year firms spend a billions in R&D in order to compete, succeed and survive (Wood & Hoeffler, 2013). Firms aggressively foster their perception of being innovative entities (Henard & Dacin, 2010). We extend the research of Henard and Dacin (2010) to research the less tangible facets of an innovate reputation (e.g. patents). We do so because products are often only consequences of patents (Malewicki & Sivakumar, 2004) and patents have regularly been a measure of technological innovation (e.g. Abraham & Moitra, 2001; Johnstone, Haˇ sˇ ciˇ c, & Popp, 2010). Our study adapts the definition of Henard and Dacin (2010) and defines (cf. p. 1594 in Shams, Alpert, & Brown, 2015): [. . .] consumer’s perception of a firm’s track record of technological innovation, degree of creativity, and potential for continued innovative activity. Reputation among customers, consumers and further constituents is hence a subjective impression of firms (Schwaiger, 2004) and of course its offerings (Shams et al., 2015). How firms can achieve gains, specifically for a technological reputation is hence a difficult task but offers many different opportunities as e.g. leaps in science, new products, design innovations, process innovations and so forth. An innovation is viewed to be “an idea, practice, or object which is perceived as new by the individuals or other units of adoption” (p. 11 in Rogers, 2003). Scholars view that the subjective assessment of innovation is often related to two conceptualizations: basically either product-based or organizationally related (Shams et al., 2015). Kunz, Schmitt, and Meyer (2011) suggest that both concepts are thus interrelated and consumers capture a more comprehensive perspective. They rely on a range of activities and elements to derive a judgment whether a company is innovative as a whole. Further, the success of an innovative firm, meaning its reputation, also depends upon their capability of commercializing their innovative and creative knowledge into marketable solutions (Kunz et al., 2011). Theories of reputation Walker (2010) divided the current research on reputation from a theoretical perspective into institutional theory, signaling theory and finally the resource-based view. Institutional theory (Meyer & Rowan, 1977) is focused on building reputation, more specifically, on how organizations obtain support as institutions and receive legitimacy (Deephouse & Carter, 2005; Walker, 2010). Conversely, the resource-based view (RBV) aims at investigating outcomes of reputation (Barney, 1991), viewing reputation as an intangible resource. The RBV for instance explains why a severe reputation enables a competitive advantage (Hall, 1992). Competitive advantage exists when rivals are unable to reproduce or implement the benefits of a certain behavior (see Barney, 1986 for more information). Applying the RBV for reputation fulfills all the requirements to explain a competitive advantage (Lin, Yang, & Arya, 2009; cf. Nagel et al., 2014); a proper reputation is valuable because it helps to overcome information asymmetries to achieve a positive public perception (Fombrun & Shanley, 1990). Reputation is rare, because it is inconsistently spread across industry participants. Reputation is inimitable because it is causally ambiguous and socially complex (Lin et al., 2009). Based on the fact that reputation is hierarchically allocated and information is incomplete, there is no substitute available to it (Lin et al., 2009). Finally, signaling theory is focused on how to gain, maintain and defend reputation (Basdeo et al., 2006; Shamsie, 2003). Several studies employed signaling theory to investigate the development of reputation and how reputation as a signal contributes to defending a strategic position in competition (Basdeo et al., 2006; Shamsie, 2003). Signaling theory is based on situations that are characterized by incomplete information (Akerlof, 1970; Spence, 1973). These settings feature ubiquitous uncertainty between interacting parties. The inherent risk that is taken by one party, while engaging in any type of exchange relationship with another, can be lowered via searching and processing information regarding the counterpart. This action is performed by interpreting observable characteristics. Those attributes are fixed to the entities, and others can be actively influenced. In the terminology of signaling theory, the former are called indices and the latter are termed signals (Spence, 1973). In practice, incomplete information is an obstacle, hindering risk averse consumers from buying from companies that are new to the market and yet unknown (Nayyar, 1990). This initial competitive disadvantage must be overcome by effectively signaling quality to consumers (Kirmani & Rao, 2000; Rao, Qu, & Ruekert, 1999); this commonly means attesting to the firm’s capability to meet customer needs. Together with signaling theory, cue utilization theory is also often used to explain how firms shape their image across consumers. A signal is basically a cue which a seller can use “to convey information credibly about unobservable product quality to the buyer” (Rao et al., 1999, p. 259). The theory proposes that e.g. a product does exhibit a bundle of cues according to its diagnosticity and the existence of other cues (Purohit & Srivastava, 2001; Wang, Cui, Huang, & Dai, 2016). Gidron, Koehler, and Tversky (1993) categorized cues into either “high scope” or “low scope”. The former are usually substantiated over a longer time and are thus rather believable and reliable (e.g. reputation) whereas the latter can be manipulated as product warranty or price (Wang et al., 2016). When individuals are encounter multiple cues during decisions, they tend to enact more often based on high scope cues and the effect of other cues is mitigated (Utz, Kerkhof, & van den Bos, 2012; Wang et al., 2016). Innovative performance For many firms, intangible assets increasingly “surpass physical assets [. . .] both in value and contribution to growth” (p. 7 in Lev, 2000). Hence, abnormal profits and dominant competitive positioning are achieved by the right deployment of intangibles (e.g. patents, trademarks, copyrights) and with other types of assets such as raw materials, financials or productions facilities (Lev, 2000). Innovative performance is the outcome of these complex processes. The ability of firms to develop, produce and successfully introduce innovations to the market is related to competitive advantage (Lengnick-Hall, 1992). Hall (1993) and Coyne (1986) point at the use and strategic value of intellectual assets as patents in this P.J. Höflinger et al. / Journal of Innovation & Knowledge 3 (2018) 26–39 29 regard, providing regulatory capabilities in strategic positioning. To measure a firm’s performance in regard to innovation dimensions, Hagedoorn and Cloodt (2003) introduced the concept of “innovative performance”. Thus, innovative performance is useful to evaluate various dimensions of innovative efforts and can be narrowly subdivided into the following: (1) “inventive performance”, (2) “technological performance” and (3) “innovative performance in the narrow sense” (p. 1366 in Hagedoorn & Cloodt, 2003). “Inventive performance” can be defined “as the achievements of companies in terms of their ideas, sketches, models of new devices, products, processes and systems” (Ernst, 2001; Hagedoorn & Cloodt, 2003). The concept represents the output of the innovative efforts of firms and thus deliberates the intellectual creativity and the available knowledge within a firm. It is usually applied by using patent counts or patent citations (Hagedoorn & Cloodt, 2003). “Technological performance” is determined as “the accomplishment of companies with regard to the combination of their R&D input, as an indicator of their research capabilities, and their R&D output in terms of patents” (p. 1367 in Hagedoorn & Cloodt, 2003). Technological performance is comparable to Hirshleifer et al. (2013) understanding of “innovative efficiency”, which is “a firm’s ability to generate patents and patent citations per dollar of research and development (R&D) investment” (p. 2). “Innovative performance in the narrow sense” is regarded as an organization’s capability to regularly introduce products to the marketplace (Freeman & Soete, 1997; Hagedoorn & Cloodt, 2003) and makes use of new product introductions. All of the constructs introduced in this section describe different perspectives and measurements of the output of innovation. Many studies employ this concept as a prevalent source in different contexts. In a recent study, Zwick, Frosch, Hoisl, and Harhoff (2017) research individual-level drivers of fostering inventive performance in firms and find that ability and problem solving style are significantly related. Brem, Nylund, and Schuster (2016) examine the how the existence of a dominant design affects innovative performance in a sample of 2.5 million patents among radical innovation and process innovation. Fang, Wang, and Chen (2016) investigate network structures in order to evaluate the chances of increasing the firm’s innovative performance by making use of recognizing new knowledge sources. Recent research applies similar measures of innovative performance and many of its sub segments. Hypotheses In competitive high-tech environments, firms actively promote their perception of being innovative entities (Henard & Dacin, 2010). Although innovative performance is an essential part of competitive advantage, its relation to reputation remains to be researched more extensively. Rindova et al. (2005) explicitly find that further research needs to determine how organizational resources attributes affect reputation. In the following hypothesis development, we will explain that this interdisciplinary setting requires elements of both theories, namely the resource-based view and signaling theory. Anything that is made by companies for the sale on competitive markets, obviously with the help of intangible assets, is logically required to be somehow perceived by its constituents. The resource based view is suited to explain particular capability of firms to develop and commercialize innovative solutions (Barney, 1991). Signaling theory and cue utilization theory are used to explain how firms shape their image across consumers (Rao et al., 1999; Spence, 1973). Examining important studies in reputation research, the common finding is that consumers in different settings have difficulties in predicting and evaluating future firm outputs and their behavior (Lange et al., 2011). The internal workings of a firm are rather opaque with limited availability of information regarding organizational abilities and intentions (Rindova et al., 2005). Therefore, there are reasons to doubt the signaling value of less tangible facets, such as patents and R&D (Henard & Dacin, 2010). Rindova et al. (2005) explicitly highlighted the role of knowledge as an intangible asset in this context. The authors argue that “the difference [to inputs] may also be because of the fact that the quality of productive assets, especially knowledge assets, is complex and difficult to observe, therefore reducing their value as a signal that stakeholders can readily use to form expectations about quality” (p. 1045 in Rindova et al., 2005). This is most likely the case, even though the external validation of the basic technology (by the respective patent office with the patent grant) can often be seen as an external positive evaluation of the companies’ ability to develop innovation. Strictly in accordance with these assumptions, we take away that intangible resources as patents, citations and R&D spending are very complex and difficult to detect by consumers. From a practical perspective, the above mentioned argumentation may not fully incorporate certain preconditions that characterize industries where technically innovative firms operate. High-tech firms usually act on markets with high barriers to enter (Carpenter & Petersen, 2002), shortening product cycles (Ali, Krapfel, & LaBahn, 1995) and large R&D investments (Neelankavil & Alaganar, 2003). The intellectual property protection in these industries is shaped by a trend of increased patenting activity (Hirshleifer et al., 2013), also often referred to as a “patent explosion” (Hall, 2004). The annual report “The most innovative companies” from The Boston Consulting Group (Wagner, Taylor, Zablit, & Foo, 2014) highlights that a key strategy of the firms on the report is “to leverage their IP both to exclude rivals and to build markets” (p. 11). That strategy has been found to be repeatedly successful for “breakthrough innovators” for several years (Wagner et al., 2014). In the long run, firms do likely secure its competitive advantage via patents (Cohen, Nelson, & Walsh, 2000), including established firms as well as ventures for whom intellectual property is their precious asset (Gans & Stern, 2003). The consequence is that innovative products new to markets are likely to be protected by patents (Battisti et al., 2015). In that sense, Wernerfelt (1984) finds that “most products require the services of several resources and most resources can be used in several products” (p. 171). Even tough consumers may have difficulties to accurately observe intangible assets as described above (e.g. Rindova et al., 2005), the products are still likely to be embodied by them (Archibugi & Planta, 1996). Many products are even composed of “numerous separately patentable elements” (p. 14 in Cohen et al., 2000). The fact that these patented technologies have to be novel in order to get granted by the patent office (USPTO, 2016) however logically increases the introduction of “new” technology. By definition, patents have to be the first of its kind (“novelty”), a nontrivial extension of what is known and finally useful thus having commercial potential (Hall, Jaffe, & Trajtenberg, 2000; Merges, 1988). Securitizing proprietary innovation is usually fostered when technologies are likely to become economically valuable (Eusebi & Silberglitt, 2014) and are thus commercialized. Otherwise, the economic payoff for engineering, time, capital, legal and human resources into filing patents would not be given (Archibugi & Planta, 1996). The higher the ability to produce leading patented innovation increases the chance that technologically innovative products are finally brought to the market. This is also confirmed by Artz et al. (2010) who find that patents are positively associated to product introductions in a longitudinal study in 35 industries over 19 years. Above, the temporal monopoly for that certain invention secures the ability to be the only firm offering that technology (USPTO, 2016). Creating something inimitably that makes it difficult for rivals to copy with certain technology-based features enhances the value for the consumer (Hanaysha, Hilman, & Abdul-Ghani, 2014). Research also discovered that the main factor of product failure is 30 P.J. Höflinger et al. / Journal of Innovation & Knowledge 3 (2018) 26–39 the absence of innovation with a novel and unique aspect (Sethi, Smith, & Park, 2001). As “newness” usually largely shapes consumers perception of innovation (Garcia & Calantone, 2002; Shams et al., 2015; Tomczak, Vogt, & Frischeisen, 2016), the perception should therefore be positively influenced based on fundamentally new technology. We therefore argue that the higher the technologial performance and inventive performance of firms, the higher should be its reputation among consumers to be also technologically innovative. Weighing the arguments from above, relying on signaling theory and cue utilization theory alone would logically result in a negation of these relations (cf. Popper, 1959), meaning that consumers can most likely not evaluate and assess intangibles as patents and R&D. However, looking at the predominant circumstances in competitive markets and the inclination toward proprietary protection it is hypothesized as follows: Hypothesis H1. Technological performance (patent stock/R&D stock) is positively associated with reputation for technological innovation. Hypothesis H2. Inventive performance (citation intensity) is positively associated with reputation for technological innovation. Furthermore, we hypothesize about influences that may shape the perception of a technologically innovative firm. In almost any market, marketing is a central part of an organization’s collective strategy (Neelankavil & Alaganar, 2003). Scholars found that firms advertise to build favorable and strong images in consumers’ minds (Aaker, 2012; Collins & Han, 2004). Furthermore, it has been shown that these investments do result in positive outcomes in terms of firm perception (Cobb-Walgren, Ruble, & Donthu, 1995). Firms set cues over time to establish their reputation (e.g. Wang et al., 2016) by making use of marketing investments (Srinivasan, Pauwels, Silva-Risso, & Hanssens, 2009). Advertising functions ideal when a company does offer something new to the consumer (Lodish et al., 1995), which can help to shape strategic positions of firms to differentiate then from competitors (Fombrun & Shanley, 1990; Rumelt, 1987). In a high-tech environment, firms actively promote and shape their perception of being innovative entities (Henard & Dacin, 2010), as technologically innovative firms need to introduce their products successfully (Neelankavil & Alaganar, 2003). Scholars find that firms which deploy more resources in marketing are thus able to better preserve innovation (Chandy & Tellis, 2000). In this vein, marketing cues can include technical novelty in comparison to its predecessors (Pauwels, Silva-Risso, Srinivasan, & Hanssens, 2004), new attributes of an innovation providing added value (Mukherjee & Hoyer, 2001) or a novel design for the perception of technological innovation (Rindova & Petkova, 2007). The possibility of a firm to strategically position itself in a favorable light reduces the constituents’ search for information (Fombrun & Shanley, 1990). As any allocation of capital, marketing spending is likewise a strategic investment that should most effectively foster the desired firm image to withstand competition. In accordance with the logic described above, the hypothesis is as follows: Hypothesis H3. Marketing intensity is positively associated with reputation for technological innovation. Data and methodology Data The data for this study consist of consumer data, patent data and financial data (cf. Nagel et al., 2014). Consumer data We obtained the consumer data via an online survey among a consumer sample from February through March 2013. We invited 2367 consumers by email to complete our survey in an online community. The topic was the technological performance of firms. We did not provide any incentives. In this approach, 7.09 percent responded to the invitation. Second, we approached another online community in which 474 consumers were invited to complete the identical survey with a response rate of 16.24 percent. In total, we gathered 244 questionnaires in which 231 (8.13%) provided the needed information. In the survey, we requested a recall task; consumers needed to freely recall the names of the five technologically most innovative companies they can think of (for more information see measures). Additionally, we collected demographic data (age, gender, level of education, field of study, and field of work experience) from the participants. With this list, we generated and searched for the required patent and financial information. Patent data This study’s patent data were generated from the EPO worldwide patent statistical database (PATSTAT). EPO’s PATSTAT collects bibliographic patent information for over a hundred patent offices in a transparent manner, with data that often date to the nineteenth century (De Rassenfosse, Dernis, & Boedt, 2014). The data provide information on the firm’s patent applications and the portfolio. Because consumers occasionally named the parent company (e.g. “Facebook”) or a subsidiary (e.g. “Whatsapp”), we collected the data for both entities. In order to assign the appropriate patents to a company we applied a semi-manual approach (Boeing, Mueller, & Sandner, 2016). For each company we defined a set of search patterns including the corporate name, its main name changes and variations (Nagel et al., 2014). Exemplarily, a company named StemCells, Inc. holds patents under the patent applicant names “StemCells, Inc.” and “Stem Cells, Inc.”; to obtain all relevant patents we exemplarily defined the two search patterns “STEMCELLS” and “STEM CELLS”. With these two search patterns we were able to identify all relevant patents even though different variations of the corporate name exist. This way, we were able to compensate different variations of name spellings in multiple legislations and also spelling errors on behalf of the patent offices. Hence, for the firms we are not able to gather patent data within the PATSTAT database, due to inconsistencies in large datasets, we applied dummy variable to account for these firms (see measures). In total, we gathered patent data for 46 of 65 companies. Financial data We collected most firms’ financials from Thomson Reuters. For the non-publicly traded firms, we manually collected the information from income statements and balance sheets from e.g. data providers as Orbis. For firms for which there was no information collected in the prior steps, we contacted the public relations department by email to provide the necessary financials. Not all firms provided us with information because of their restricted financial policies. Some firms explicitly do not publish or any financial information. In total, we were able to obtain complete financial data for 48 of 65 firms. Next, we merged these three datasets for further analysis. The dataset for regression offers information for 231 consumers, 65 firms and in total 1046 observations (Nagel et al., 2014). Measures Dependent variable Reputation for technological innovation: We invited the participants to freely recall the five most innovative companies in their perception. Consumers needed to provide their answers in ranked order starting with the technologically most innovative firms. Furthermore, we did ask the consumers to name only technologically innovative companies (Nagel et al., 2014). We are aware that there P.J. Höflinger et al. / Journal of Innovation & Knowledge 3 (2018) 26–39 31 are various paths to innovation (Chandy & Tellis, 1998), so asking strictly for technologically innovative firm should ensure that respondents are not confused with other types of non-technology innovations (e.g. new organizational processes such as selling directly to consumers or a new website). Our procedure is similar to the annual report “The most innovative companies” from The Boston Consulting Group (Wagner et al., 2014). We specifically did not provide a list of predefined firms to choose from. Instead, we used an unaided recall task as e.g. in Nedungadi and Hutchinson (1985), Kim and Kim (2005) or Agarwal and Rao (1996). Predefined firm lists, such as the Fortune ranking used by Roberts and Dowling (2002), preselect a firm by its revenue in a given year. In addition, this would result solely in large firms, but technically innovative firms are often relatively new to a market, such as Tesla or Dropbox. For the further understanding of the data, it is important to note a lower number is actually a better place in the corresponding ranking (position “1” is the lowest number for the statistical analysis, but the technologically most innovative firm). Exemplarily, negative correlations indicate that higher values are associated with higher ranks. We finally deleted those firms that occurred once, leading to a list that covered 90.56 percent of all mentioned firms. Operationally, we ranked the remaining firms based on the frequency of their nominations by the consumers (see Table 1). The company that was recalled by most consumers is ranked number 1; the firm that was least nominated by the sample is ranked last. For firms in which the frequency of nominations is identical, the equivalent position in the ranking is provided. Furthermore, the subsequent ranking positions remain empty if two firms share the identical ranking position (cf. Nagel et al., 2014). Independent variables Inventive performance (citation intensity): Citation intensity is a relative measure for the overall quality of the patent portfolio and serves as a measure of inventive performance (Hagedoorn & Cloodt, 2003). Inventive performance represents the output of the R&D process; it is regarded as a measure of the value of a firm’s science and technology (Hand & Lev, 2003). To calculate the citation intensity, we divide the citation stock by the patent stock. The patent stock is based on patent counts, measuring the absolute number of patent applications filed by a firm, representing a proxy for new technologies (Acs & Audretsch, 1989; Hagedoorn & Cloodt, 2003). The stock variables aggregate the patent counts; here, it is calculated as in Sandner and Block (2011) with a 15% depreciation rate. Likewise, the citation stock is based on patent citations to incorporate a quality factor in the analysis. This paper uses forward citations, basically citations from successive future patents, which are regarded to be a better value indicator than backward citation (Lanjouw & Schankermann, 2004). Likewise, we apply citations that have been obtained at least five years after its release (Bloom & Van Reenen, 2002). The citation stock is calculated as in Hall et al. (2005). Technological performance (patent stock/R&D stock): Hagedoorn and Cloodt (2003) find that technological performance is defined by the relation of R&D input and R&D output as measured by patents. We calculate this ratio using the firms patent stock divided by the R&D stock. Marketing intensity: Advertising can help to shape strategic positions of firms to differentiate themselves from competitors (Fombrun & Shanley, 1990; Rumelt, 1987). Because advertising is usually used to influence the public’s attention, we employ marketing intensity using the average selling, general and administrative expense for 2012 to explain short-term influences of marketing activities. Furthermore, these expenses are divided by the corresponding revenue (Krishnan, Tadepalli, & Park, 2009). It is used by a variety of studies as a valid measure to operationalize marketing investments (Krishnan et al., 2009; Lin, Lee, & Hung, 2006). Both information can be taken right from income statements. Exemplarily, such information is also available from Google Finance or other open access financial providers (e.g. Google Finance – Daimler AG Income Statement). The ratio is calculated as follows: Marketing Intensityt=SG&At Revenuet R&D stock: R&D inputs are usually applied as an indicator for the efforts a firm undertakes to generate innovative output (Hagedoorn & Cloodt, 2003). However, R&D spending is also associated with the innovative competence, because it is a part of broader routines a firm follows to be innovative (Hagedoorn & Cloodt, 2003). Successful R&D input in previous years results in the increasing commitment of the allocation of future R&D resources (Hagedoorn & Cloodt, 2003). To measure the amount of R&D, we use the declining-balance formula as in Hall (2007a); Sandner and Block (2011), with a stock variable. The variable is log transformed. The variable, is calculated as follows with a usual 15 per cent depreciation rate for comparison to earlier work: R&Dstock t= R&Dflow t+ (1 − ∂)R&Dstock t−1 Revenue: Larger firms tend to receive much public attention (Fombrun & Shanley, 1990), and reputation is positively related to a firm’s size (Hutton, Goodman, Alexander, & Genest, 2001). The availability of more information biases the judgment of individuals, which, in turn, affects the familiarity of these organizations (Tversky & Kahneman, 1974). Consequently, we use annual revenue to explain size effects. The variable is log transformed. Demographics: We also obtained supplemental information concerning the individual background of the participants. Thereafter, we received the age, the gender, the level of education, the time of work experience and the type of industry the experience was obtained in. Control variables Type of industry: The importance of patents and R&D varies across industries, particularly high technology industries facing decisive allocations of resources to innovative efforts (Neelankavil & Alaganar, 2003). We apply the standard industrial classification (SIC) system to incorporate the effects of these differences. More specifically, we use the 2 digit major group code for further differentiation. Country: Because consumers were able to freely recall firms, different origins were noted. As for the type of industry, a country variable explains the influence of country variations (Hall, 2007b; Sandner & Block, 2011). Patent dummy: If it was not possible to collect patent data for firms, we employ a patent dummy to explain these influences (Suits, 1957). R&D dummy: Similarly, when firms do not disclose available information on their R&D spending, we employ a dummy variable to account for this (Suits, 1957). Estimation In this study, we employ a Poisson regression to research the influence of intangible assets on the reputation for technological innovation. We do so because the dependent variable of this study, the ranking for the reputation for technological innovation conducted in 2013 has count data of the most innovative firms for the dependent variable (Fahrmeir, Kneib, & Lang, 2009). Similarly, all our consumer variables are obtained in 2013. In contrast, the independent variables are bound to time lags. The reasoning originates from the fact that it takes several years until ideas emerge into 32 P.J. Höflinger et al. / Journal of Innovation & Knowledge 3 (2018) 26–39 Table 1 Firm nominations. Ranking Firm Number of employees Country Count of nominations 1 Apple 110000 USA 179 2 Google 61814 USA 130 3 BMW 122244 Germany 85 4 Samsung 489000 South Korea 84 5 Siemens 348000 Germany 64 6 Microsoft 112388 USA 61 7 Amazon 230800 USA 43 8 Audi 82838 Germany 29 9 Facebook 12691 USA 27 10 Robert Bosch 375000 Germany 25 10 IBM 377757 USA 25 12 Volkswagen 610076 Germany 20 13 Daimler 284015 Germany 19 13 General Electric 301000 USA 19 15 Tesla Motors 13058 USA 15 15 Toyota 338875 Japan 15 17 3 M 89446 USA 14 18 BASF 112435 Germany 13 19 SAP 76986 Germany 12 20 Sony 131700 Japan 11 20 Intel 107300 USA 11 22 Bayer 116800 Germany 9 23 Roche 28517 Switzerland 7 24 Nike 62600 USA 6 Festo 17800 Germany 6 26 EADS 138622 Germany 5 Beiersdorf 17659 Germany 5 Porsche 22401 Germany 5 Procter & Gamble 110000 USA 5 Airbus 138622 France 5 Unilever 172000 England 5 32 Infineon 35424 Germany 4 SpaceX 5000 USA 4 Boeing 165500 USA 4 Linde 64538 Germany 4 Gore 10000 USA 4 Dropbox – USA 4 38 Salesforce 13300 USA 3 Ideo 550 USA 3 Ikea 147000 Netherlands 3 Virgin 50000 England 3 Lenovo 33000 Hongkong 3 Phillips 112959 Netherlands 3 Novartis 135696 Switzerland 3 Frauenhofer Society 23786 Germany 3 46 Twitter 3638 USA 2 Nvidia 6384 USA 2 Red Bull 10997 Austria 2 Cisco Systems 75000 USA 2 MTU Aero Engines 8334 Germany 2 Coca Cola 123200 USA 2 Brainlab 1129 Germany 2 Kuka 12300 Germany 2 Toshiba 199000 Japan 2 Oracle 122458 USA 2 HTC 17280 Taiwan 2 Continental 208000 Germany 2 Hewlett Packard 252000 USA 2 Dell 111300 USA 2 Nintendo 5095 Japan 2 Bang & Olufsen – Denmark 2 Nokia 110000 Finland 2 Qualcomm 26000 USA 2 Bose 10700 USA 2 Thyssen Krupp 154906 Germany 2 Spotify 1500 Sweden 2 Notes: Firm nominations by the consumers. Number of employees are from 2015. technology and finally patents. Therefore, studies usually employ time lags, e.g. Nooteboom, Van Haverbeke, Duysters, Gilsing, and Van den Oord (2007) find that a 5 year time frame to measure technological impact (for patents) is appropriate. We extend this time frame and employ a maximum time lag of 7 years until 4 years (2006–2009) prior to the dependent variable in 2013. Results Descriptive statistics For our study, we were able to gather information from 231 consumers, each noting five firms to be among the most innovative P.J. Höflinger et al. / Journal of Innovation & Knowledge 3 (2018) 26–39 33 Table 2 Descriptive statistics. Variables N Mean SD Min. Max. Dependent variable Reputation for tech. innovation 1046 9791 11.322 1000 46.000 Patent variables Technological performance 1046 0.734 0.634 0.000 2444 Citation intensity 1046 0.896 0.305 0.000 1608 Company variables Marketing intensity 1046 0.143 0.084 0.000 0.583 R&D stock (in bn) 1046 15.599 12.736 0.000 41.985 R&D intensity 1046 0.285 0.221 0.000 1607 Revenue (in bn) 1046 59.655 44.207 0.000 207.852 Consumer variables Gender 1046 0.328 0.470 0.000 1000 Age 1046 24.984 6168 19.000 81.000 Work experience (time in years) 1046 1234 2873 0.000 37.000 Work experience (type of industry) 1046 2156 1187 1000 4000 Education (6 Categories) 1046 1389 1396 0.000 6000 Field of study (6 Categories) 1046 0.739 1360 0.000 5000 Controls Country (13 Categories) 1046 1823 1464 1000 13.000 Industry (11 Categories) 1046 44.293 16.843 20.000 73.000 No R&D (dummy) 1046 0.118 0.322 0.000 1000 No Patents (dummy) 1046 0.157 0.364 0.000 1000 Notes: SD = standard deviation, 65 firms and 231 consumers; patent and company variables for 2009, consumer variables for 2013, Industry categories are based upon the SIC classification. firms in their perception. This procedure resulted in 1155 nominations for our ranking (cf. Nagel et al., 2014). From these initial company names, 1046 were determined to be mentioned at least twice. The top five nominations are Apple (17.11%), Google, Inc. (12.43%), BMW AG (8.12%), Samsung Corp. (8.03%) and Siemens AG (6.11%). In sum, the previously mentioned firms explain approximately 51.82 percent of all recalled firms. The top fifteen combined explain 80.31 percent of all ranking nominations. Interestingly, our respondents’ nominations are very similar to the annual report “The most innovative companies” from The Boston Consulting Group (Wagner et al., 2014). To provide further information, the complete list is attached in Table 1. Furthermore, we gathered important information regarding consumers in our sample. The gender ratio in our sample of respondents is split into 67.50 percent male and 32.50 percent female. Furthermore, the age ranges from 19 to 81 years, averaging 24.98 years. Our participants have mean work experience of 1.22 years, with a minimum of none to a maximum of 37 years. Within this time frame, 42.10 percent have had experience in high tech industries, whereas 22.80 percent have experience in low-tech industries. From an education perspective, 44.60 percent of the sample has an educational background with a university entrance qualification. Of the respondents, 31.20 percent completed a bachelor’s degree, and 19.50 percent finalized a master’s program. During their studies, 68.40 percent focused on “business & economics” and 12.10 percent on engineering related subjects. Table 2 provides information for the variables. The patent data were gathered for the firms mentioned in the ranking. In detail, the largest patent stock originates from Samsung with 47,711.79 patents. The average patent stock size is 12,563.67 with a standard deviation of SD = 15,979.52. From the patent citation stock, we can highlight that the highest number of citations again originate from Samsung, which has 10,813.83. On average, the firms in the sample have 14,328.89 patent citations. The citation intensity ratio (citation stock/patent stock) is, on average, 0.896. The highest ratio originates from Cisco Systems with 1.608. The technological performance (patent stock/R&D stock) has a mean of 0.734. Furthermore, company variables show that the firms achieved annual revenues with a mean of 59.65 billion US-dollars. The maximum annual revenue originates from Toyota Motor Corporation, which had approximately 207.85 billion US-dollars. The firms nominated in this study emerge from an international sample. Most perceived innovative companies originate from the USA (54.68%) and from Germany (29.90%). The largest industry type of our study, measured using the SIC classification, resulted in the following split: “Transportation equipment” (22.40%), “Machinery and computer equipment” (26.70%) and “Business Services” (23.80%). Information on the correlations among this study’s variables can be found in Table 3. To additionally check for multicollinearity, the variance inflation factors (VIF) appear to be uncritical as none of the variables has a value above ten (Kennedy, 2003; Neter, Wasserman, & Kutner, 1989). Table 3 indicates that VIF values range from 1.07 to 4.27. We have also calculated an average VIF of 2.10, which indicates no apparent multicollinearity. Because some correlation coefficients are relatively high for R&D stock and revenues, the results should be still interpreted with care. Regression results In our regression, we estimate four different models for the different time spans to explain the fact that knowledge needs time to be incorporated in products and services. Therefore, we employ regression models beginning in 2006 (M1) until 2009 (M4). We have clustered our models by the consumer because the consumers named companies, which resulted in the ranking for the dependent variable reputation for technological innovation. With this method, we find robust standard errors. To highlight our results, we start with largest time lag (M1) to the smallest time lag (M4). The complete results can be found in Table 4. We find significant support for the influences of technological performance (H1) on reputation for technological innovation in M1. Again, one must bear in mind that negative correlations indicate that higher values are associated with higher ranks. A lower number is actually a better place in the corresponding ranking. In M1, technological performance is positively and significantly related to reputation for technological innovation, (0.092, p < 0.05). The other models M2–M4 do not show significant results and are negative. Furthermore, we find significant and