**4. Results**

To understand how firms are sourcing innovation in the Digital Age, we first examine the broad trends in how innovation sourcing is evolving within large firms. Then, in the next section, we connect these trends to characteristics of digital


**Table 2.** *Sample summary statistics.*

*Innovation – Research and Development for Human, Economic and Institutional Growth*

innovation, e.g., the challenge that businesses have in building advanced capabilities in these areas that seem to be driving them.

#### **4.1 Which innovation sources are corporations using?**

The question of which innovation sources firms use and why has been a major topic in the academic literature. One area of particular focus is the usage of external innovation sources ([6, 7, 16, 30, 31]; etc.). Our research provides new, detailed evidence on this question.

Of the 11 innovation sources listed in **Table 1**, we find that five are currently used by more than half of firms: Suppliers (86%), Central R&D (79%), Universities (64%), Business Unit with staff dedicated to innovation (63%, hereafter "BU Dedicated") and Third-Party (54%), as shown in **Table 3**. Here, BU Dedicated refers to the company's innovators that are managed by their business units (not centrally) and whose principal job is innovation – in contrast to those who do so part-time who we call Business Unit staff with operational responsibilities (hereafter "BU Operational").

Of the sources used most often, it is notable that suppliers and universities, the first and third most-used, are external to the firm.<sup>1</sup> At the other extreme, we find much-touted sources that are used by fewer firms, including Crowd (18%), BU Operational (30%), Startups (32%), and Innovation Labs (37%). Of these, the share for BU Operational is notably low given the widespread attention given to industry manifestations of this, including the 3 m 15%-time and Google 20%-time models. The small share using innovation labs is perhaps less surprising since the costs of this type of innovation are high and much of the growth in usage has been very recent [32]. We also find that customer/user-innovation is used by 41% of companies and that 22% source innovation from competitors (e.g., through open-source software).

In addition to indicating whether their company used a given innovation source, respondents indicated how long they had been using it.2 This provided insights on how innovation sourcing has shifted over time. For some external innovation sources, the changes have been rapid: only 2% of current users of start-ups and crowd have been doing so for more than five years, with the remaining 98% adopting since then. Other innovation sources are more stable, with most of the companies that use them having done so for more than five years, including BU Dedicated (90%), Customers (89%), BU Operational (84%), and Central R&D (67%). One, perhaps surprising, result is that one-third of the companies working with universities, a long-discussed source of innovation [33], only started doing so in the past two years. These results strongly support the argument in the open innovation literature about the growth of external sourcing of innovation (e.g., [6, 7]), but they suggest that much of this adoption has happened quite recently, despite it having been widely discussed for nearly two decades.

By multiplying the share of firms using an innovation source with the share that have adopted in the last two years, we can calculate the share of all firms that have

<sup>1</sup> Later, we provide data justifying our assertions about which sources are internal and external.

<sup>2</sup> This was asked via "how long have you been using" type questions. In early drafts of the questionnaire, we considered asking directly about which sources were used two and five years ago, which would have had the benefit of allowing us to detect which sources were being dropped, but test-respondents indicated that they could not answer this reliably (because personnel had left, etc.).


**Table 3.**

*The usage and importance of innovation sources.*

### *Sourcing Innovation in the Digital Age DOI: http://dx.doi.org/10.5772/intechopen.111707*

adopted in the last two years. **Table 4** shows this result, revealing that despite the substantial hype around crowd and startups, it is actually universities that are the most-adopted source in the past two years, with 21% of firms starting to innovate with them. This finding is in line with Fey and Birkinshaw's [34] recommendation that "all else equal, to maximize R&D effectiveness, firms should promote university partnering" because "universities, which are not potential direct competitors, are preferable R&D partners" as well as Giannopoulou's et al. [35] view of the importance of collaboration with universities.

These results also suggest a categorization that we will use later in our analysis between fast-growing sources (Universities, Third-Party, Startups, Crowd, and Innovation Labs), and slow-growing traditional sources (Central R&D, Suppliers, BU Dedicated, BU Operational, Customers and Competitors).

With so much adoption of external innovation sources, the question arises whether these external sources are substituting for internal ones. In particular, there is a discussion in the practitioner literature that innovation may be becoming "virtual," with internal sources being jettisoned as external ones take their place [36, 37]. In contrast, academic research has argued that internal and external sources are complementary (Wuyts and Dutta [38], Thompson, Bonnet and Ye [39]), with Wuyts and Dutta saying, "the experiences gained through internal knowledge creation help firms benefit from the opportunities of portfolio diversity." Our data allows us to interrogate this empirically (**Figure 1**).

If firms were making their innovation "virtual", we would expect **Figure 1** to show a negative slope between the number of internal and external sources that firms use, as firms substitute away from internal sources. Instead, we observe a positive slope. This suggests that firms are not going to virtual innovation, but that they are instead choosing to broaden their innovation portfolio to use more sources, both internal and external. An OLS regression confirms this statistical relationship, showing that an increase of one additional internal innovation source is associated with a 0.41 increase in the number of external innovation sources used (p-value = 0.000), as shown in **Appendix B**.


#### **Table 4.**

*Most adopted innovation sources in the past two years (share of firms adopting).*


**Figure 1.** *Number of internal and external innovation sources used (% of companies).*
