Understanding Private-Sector Decision Making for Early-Stage Technology Development
A “Between Invention and Innovation Project” Report

IV. Findings: Corporate ESTD Investments and Activities

Early-stage technology development (ESTD) investments are critical to sustaining long-term economic growth, and corporate funds represent the largest source of funding for the nation’s ESTD activities. The results of our interviews reveal, however, that despite recognition of the value of early stage research, ESTD investments are rarely a corporate priority, market incentives to fund ESTD are low, and ESTD budget flows are under constant pressure. The quickening pace of technological change, the increasing efficiency of capital markets, and the continual demand for profits has forced a shift from a technology-forward to a market-back paradigm within many corporations. These pressures have created a heavy bias towards product development research activities at most firms, at the expense of a more long-term inventive focus. In addition, companies earmark the vast majority of their R&D funding to support existing business lines rather than to research new technologies that could enable entry into new markets. Increasingly, it is the market that drives innovative activity, not the other way around.

ESTIMATES OF CORPORATE ESTD INVESTMENTS

Our research indicates that of the $180.4 billion invested into research & development by U.S. firms in 2000, an estimated $13.2 billion funded the kinds of ESTD activities that are targeted at bringing radical technological innovations to the marketplace. This works out to about 7.3% of total corporate R&D budgets that is dedicated to ESTD activities. As noted earlier, this may be an over estimate, since the firms interviewed were somewhat more R&D intensive than the average firm in each sector, and ESTD expenditures appear to be correlated with R&D intensity. The majority of R&D spending, 86%, is for product development; and the remaining is for concept/invention. The results of our research are summarized in Table 2 and Figure 4.

There are significant variations in ESTD expenditures across industries and between firms within specific industries. These inter- and intra-industry variations are shaped by several forces including the increasing sophistication required to develop new technological innovations, mounting pressures on corporate R&D divisions to demonstrate financial value from R&D investments, and the importance of the lifecycle position of specific industries relative to other industries and individual companies relative to their peers.

Source: BAH Analysis, Interviews with Corporations, National Science Foundation, Research and Development in Industry: 2000, Arlington VA, 2003 (NSF03318)

In 2000, the two firms we interviewed in the chemicals industry invested on average 33% of their R&D dollars on ESTD activities, the highest proportion for any group of industry firms that we interviewed. These high ESTD expenditures were driven by a common corporate emphasis on new technology and market development and by market expectations for frequent innovation. The fact that the chemicals industry seemed to have the highest ratio of ESTD to all R&D may seem counter-intuitive, given the significant amount of bulk industrial chemicals manufactured in this industry, but it is important to observe that the R&D to sales ratio in this industry is only 3.8%. Thus the industry is not nearly as R&D-intensive as biopharmaceuticals or electronics, but the R&D that they do perform is often aimed at fundamental research to spur the development of new product innovations, and thus tend to spend a larger fraction of their modest R&D investments on early stage research than do other industries.

In sheer dollar terms, the largest ESTD spender was the electronics industry. The results from our interviews suggest that the electronics industry spent $3.5 billion, or 11% of its R&D, on ESTD activities in 2000. Among the firms we spoke with in this highly competitive industry, many insisted that while the incentives to exploit and extend existing product lines are powerful, such a short-sighted strategy could be perilous. Given the rapid pace of technological change in electronics, investments into new lines of research and the pursuit of an innovation-led growth strategy are the most viable paths to long-term survival.

On the opposite end of the spectrum, the computer software industry showed no evidence of substantive ESTD activity — a result that may be surprising to some. We found that almost all software releases, creative as they may be, are built using wellestablished technologies and programming languages. They are rarely based on truly novel technological innovations—indeed, most were business model or market innovations. Numerous software industry executives provided corroboration for this finding.

Overall, we found that ESTD spending is concentrated in industries based on quickly developing technologies, like electronics, specialty chemicals and materials, and biopharmaceuticals. Mature industries based on well established technologies, like the automotive and computer software industries, typically spend less on ESTD and focus more of their resources on product development.

Within individual industries, significant firm-level variations in ESTD spending also exist. A firm’s relative lifecycle position, for example, is a key driver of intra-industry differences in ESTD investments. Companies in the early stages of their lifecycle are more likely to invest more heavily into ESTD than more mature companies who focus instead on promoting existing product lines through heavy spending on product and market development. As companies grow, their technology investments become increasingly targeted and disciplined processes are put into place to evaluate all research projects.

Another critical driver of ESTD spending is related to broader corporate strategies. Technology-centric companies for whom new technology is seen as a source of growth are more likely to invest heavily in ESTD than product-based companies for whom technology is a cost center. On the other hand, companies seeking to break out of their existing market positions or to rejuvenate their innovation resource base may make disproportionate investments into early stage R&D relative to their peers.

KEY TRENDS SHAPING CORPORATE R&D AND ESTD INVESTMENTS

1) R&D PROCESS EVOLUTION: INCREASING COMPLEXITY OF TECHNOLOGY DEVELOPMENT

Most interviewees generally agreed with the classification of R&D into the four-phase innovation framework used in our discussions (Basic, Concept/Invention, ESTD, Product Development). However, many respondents resisted the linear simplicity of our idealized framework. They noted that the actual innovation pathway is frequently much more complicated. The development of any innovation can require multiple parallel streams of research, iterative loops through any of the four stages, and linkages to developments outside the core of any single company.

Even when all the technical challenges are solved, there are still external risks that can significantly alter the development path of an innovation. The chief technology officer of a large machinery manufacturer states: “New product development is based on technologies that are largely believed to be proven, but there are still significant risks related to market, channel, and other infrastructure development. Companies are sometimes surprised when they find out that some technologies turn out to be less developed than anticipated.”

Rapid advances and the increasing breadth and depth of knowledge available across all scientific fields have also contributed to the acceleration of this complexity in recent decades. To many, the pathway from scientific invention to commercial innovation has reached the point where the process is more web-like than linear. Consequently, the ability of any one company to develop all of the technological elements required to deliver significant advances alone has rapidly diminished. According to a disk drive industry executive we interviewed, “As technology advances, it costs more to solve successive problems. At some point, solving a new problem is beyond the capabilities of any one company.” There are simply too many potential ideas and too few resources to go it alone. There was a strong sense among our interviewees that the scale of research required to create new innovations has increased as technology becomes more complex. But a firm’s ability to capture the full benefits and exploit the full potential of new research has not kept pace, making ESTD investment decisions more difficult than ever before.

2) PRESSURE FOR MEASURABLE RESULTS AND FINANCIAL RETURNS

Increased pressure on R&D to deliver measurable results was also cited as a key force that has driven corporations almost entirely away from basic R&D, making it difficult to justify many activities that do not directly support existing lines of business. One interviewee dubbed this trend the “Larry Bossidy approach,” after the famed CEO for whom he worked at Allied Signal. “Bossidy was very uneasy with our basic research work because he could not measure the expected return of his investment in financial terms.” Projects that did not have clearly demonstrable financial benefits were not funded, and the R&D portfolio shifted dramatically toward product development.

Increased pressures to deliver near term financial results and manage profits to expectations have resulted in an increased bias towards the more predictable and more immediate payoff of product development, at the expense of earlier stage investment. This increased emphasis on predictability of earnings also has created a bias toward a fast follower technology strategy. These trends were evident throughout our interviews, regardless of firm size and industry.

While general investment into earlier stages of R&D has faded, corporations will still opportunistically invest in earlier stage development in a more reactive mode, either in response to significant threats, or to meet aggressive growth objectives.

3) INDUSTRY AND COMPANY LIFECYCLE INFLUENCES

The final major influence we observed was differences in R&D investment related to industry and by company that are in part linked to lifecycle positions. Support levels for ESTD vary widely by industry, and by company within specific industries. While the average ESTD investment is about 9% of corporate R&D spending for all firms, ESTD investments in the computer software industry is essentially zero, while for the biopharmaceutical industry, the rate is 13%. Within the biopharmaceutical industry, ESTD spending ranged from 0% to 30% of R&D at the companies interviewed.

We believe that the key driver of these differences is the lifecycle position of the industry and the individual company.³¹ More mature industries such as the automotive sector tend to invest a smaller percentage of R&D into earlier stages of research than industries at an earlier stage of development such as the biotechnology sector.

However, individual companies may make disproportionate investments in early stage R&D compared to their peers as an attempt to break out of their existing positioning or to rejuvenate their innovation resource base. Several companies that we interviewed described how they reached a deliberate decision to rebalance their investments toward ESTD after recognizing that they were not positioned for growth. In some cases they have managed complete transformations out of an historical line of business and into high-tech sectors in which they did not participate a decade ago. Monsanto’s move into genetics in the 1980s is a successful example of a company temporarily moving out of a product development focus and into a strategy emphasizing basic and ESTD research.

SELECT INDUSTRY ANALYSIS: DETAILS FROM THE INTERVIEWS

COMPUTER SOFTWARE

Few industries experienced the unprecedented expansion enjoyed by the software industry throughout the 1990s. Driven by the proliferation of the Internet, whole new types of software products and services were introduced and new markets were created. Yet despite this remarkable growth, none of our software industry respondents were prepared to state that growth in the software industry was fueled by truly new technical innovations.

Based on our interviews, we found that the incentives and opportunities for invention in the software industry have been tempered in recent years. Throughout 1998 and 1999, for example, Y2K issues siphoned a significant portion of industry resources away from inventive research. More importantly, the emergence of the Internet has required the industry to respond to changing customer needs and expectations. Demand for web-enabled software has created opportunities for software vendors to generate revenue by selling web-enabled versions of existing products.

It might seem that software is early in its life cycle, not late. However, at the time of this survey, most software firms were relying on conventional software engineering tools and on Internet and web technology as if these were mature, expecting to evolve the underlying technology later if needed.

The proliferation of new web-based services, while generating important economic value, represents a new class of market innovations, relying upon unique value propositions and business models to reach customers in new ways, rather than on new technical inventions.

Software companies use existing technical tools to help expand functionality. These are not technical innovations, strictly speaking. Even today’s most creative software packages are mostly built using well-established programming languages and tools. While the configuration of programming code in a new software release may be unique and the abilities provided to the user may be novel, the fundamental technological basis of most software applications is not extremely innovative.

The introduction of Java, led by the Sun Corporation, is seen by many as one of the few true technical innovations in recent years. In contrast, XML (eXtensible Markup Language) is an extension of existing web coding standards and is not based on a new invention. In any case, Microsoft, Sun, and Oracle were not among the firms interviewed, thus suggesting that in this category the estimated ESTD in the entire software industry may have been to some extent an underestimate.

According to our respondents, the incentive to create new inventions in the software industry is small. One software executive explains, “The moment you introduce a software product to market, you need to start providing customer support. So a part of the team that developed the product has to be dedicated to support.” Another manager states, “The industry itself demands that new versions of old products be introduced into the market at least every two to three years. Customers also demand numerous minor enhancements and changes to a product after purchase.”

Within the quickly evolving enterprise resource planning (ERP) market, significant resources are still focused on developing interoperability standards between various ERP suppliers that deliver products to help companies integrate and automate business practices associated with the production or operation of a company. According to one senior ERP executive, “within the Enterprise Software space there are so many little problems that can be solved at so many different organizations, that we don’t have to worry about being in this business for the next fifteen to twenty years.”

With little incentive for invention or technical innovation, we conclude that even though the software industry has created huge economic value in recent years, there is essentially no significant ESTD activity being funded internally by the industry at the moment; virtually all R&D dollars in the software industry are dedicated to product development (see Table 3).

TELECOMMUNICATIONS

Firms engaged in relatively “new” areas of telecommunications, such as optical networking and wireless infrastructure must spend considerable amount of R&D dollars in ESTD to keep up with technological change. But firms outside these new areas focus their resources heavily on product development. Overall, only 10% of R&D dollars in the telecommunications industry is spent on ESTD and 90% on product development (see Table 4).

Deregulation of the telecommunications industry has escalated industry competition and quickened the pace of technological change. With the emergence of whole new classes of competitors to the industry, a proven capacity to innovate has become a prerequisite for any firm to remain competitive. But given tighter profit margins and shorter product development cycles, firms cannot afford to spend lavishly on unfocused R&D.

At one representative telecommunications firm, only 5 of 270 engineers were charged with researching and developing new technologies. Rather than fund expensive in-house early stage research labs, many of the firms we spoke with relied on other strategies, including company acquisition, technology licensing, and aggressive recruiting of industry experts, to acquire already proven technologies and reduce market risk.

“Most of the time, the proof of concept work has already been done by these acquired companies or hired personnel,” says the chief technology officer of a midsize switching and transmission equipment manufacturer.

The rapid pace of change in the industry requires short-term planning horizons. One mid-sized telecommunications manufacturer reported the need to develop one new marketable idea per quarter in order to stay competitive. Another mid-sized telecommunications firm stated that R&D goals are limited to a one-year time horizon, while a third noted that any product idea requiring more than five years to be commercialized is usually abandoned.

Academic collaborations also play a significant role in the telecommunications sector. Such partnerships are almost always with institutions residing close to firm offices or located in key target markets. These cooperative efforts focus on basic research and serve as idea generators for industry, as well as a talent feeder into in-house corporate labs.

AT&T funds research sites at Cambridge University and UC-Berkeley focusing on network, multimedia, and mobile communications. Other examples include the Center for Wireless Communications at the UC-San Diego, a cross-disciplinary research and education program sponsored by industry participants, and GCATT, a local R&D initiative at Georgia Tech University linking local Georgia industry, state government, and academic partners. One telecommunications executive spoke of the challenges facing joint ventures with academic partners. “We would prefer to do less ground-breaking and risky work, but most professors are not interested in partnering with us unless we are doing cutting-edge research.”

CHEMICALS

According to our interviews the chemicals industry invested on average 33% of their R&D dollars on ESTD activities, the highest proportion for any group of industry firms that we interviewed. We believe that this is a function of the chemicals industry’s position in the overall value chain. Chemicals are an input into other manufacturing industries and are rarely a final product in and of themselves. Therefore there is less onus on the chemicals industry to make the kind of engineering related or consumer related product development investments associated with industries such as manufactured goods. Basically the activity related to adapting the product to specific consumer and market requirements happens down stream. If we were to include the chemicals inputs into an analysis of the entire value chain supporting any given product we would expect that the proportion of ESTD to product development expenditures would look more like that of the other industries analyzed.

A high priority in the chemicals industry is growth through invention. Unlike other industries we interviewed, the chemicals industry has a large portion of R&D expenditure funded centrally and performed in central corporate labs. There was an explicit allocation in these companies of 3040% of research activity to markets where the companies had no current position, therefore the role of corporate “incubators” was important for these companies (see Table 5).

In addition to centrally funded invention activity, both companies cited partnerships with downstream companies as a source of invention activity. In these instances the company would work with downstream partners to develop materials that would solve a problem for the downstream company or enhance the downstream company’s product. This brings up one of the complications in the vocabulary used in discussing these issues. The development of a new to the world material by the chemical company would be considered in our vocabulary an ESTD. The application of this material to incrementally enhance an existing product would be considered product development on the part of the downstream company. However, radical innovations for use in manufacturing by established firms often encounter serious barriers due to the cost and uncertainty the customer experiences when considering a basic change in design or production processes. Thus, such projects may qualify as ESTD even though the final product differs only in cost and quality, not in product function.³²

ELECTRONIC COMPONENT MANUFACTURING

According to our interviews, early stage research activity in the electronic component manufacturing industry is relatively robust, driven by high consumer demand for innovation and intense industry competition. That being said, the vast majority of research activity in the electronic components industry is focused on improving existing electronic components or using existing technology to create components for new devices — not on groundbreaking new innovations that disrupt old markets and create new ones. Only 11% of R&D dollars in the electronic component manufacturing industry is devoted to ESTD and 5% to concept/invention, while the balance is dedicated to product development (see Table 6).

In areas where scientific knowledge and technology are well established, early stage research is low, as in the case of a producer of board-level I/O products for computers. According to its VP of engineering, “Point to point signal delivery is a very fundamental science, and since the format of the signals does not change even if the devices at the end points change, our firm does not have to do major research to stay in business.”

A leading manufacturer of graphics processors notes that it is able to rely on academic research to provide the inputs for many of the advances in its new products. Its research engineers get most of their ideas sourced from public domain papers presented at annual academic conferences and work on developing 3D rendering algorithms based on these concepts.

For most electronic component manufacturers, industry competition and diminishing gross margins place enormous pressure on R&D budgets, necessitating new strategies to distribute research costs and minimize risk while capturing the fruits of early stage research.

According to the CEO of a large company with R&D expenditures in excess of $500 million, “The traditional corporate model of R&D is dead. The centralized Edison and GE labs model was efficient when technology was less complex. But today, technology is too complex to justify development for use by just one firm. As a result, industry labs have become product development and enhancement labs, with less emphasis on developing truly innovative technologies.”

Even a large computer equipment manufacturer with revenues of $2.0 billion cannot afford to do all of its own research. The company’s senior vice president noted, “[W]e are ‘virtually vertically integrated.’ We work closely with our supplier network to identify technologies to be pursued or science to be developed. We also share costs of developing new technologies.”

An executive in the highly competitive disk drive industry said, “The products we sell are highly commoditized in today’s market and gross margins are too small to allow any one company to drive the R&D effort alone.”

AUTOMOTIVE INDUSTRIES

Because of the relatively high initial investment required to purchase an automobile, most consumers are slow to adopt new automotive technologies. As a result, automobiles have evolved slowly through waves of incremental technological improvements, punctuated by the occasional radical innovation. Over 90% of R&D dollars in the automotive industry are devoted to product development and the balance is distributed across the earlier stages of R&D (see Table 7).

The surplus of power created by the development of the high-compression engine allowed vehicles to grow in size and weight and encourage a broad new set of technologies. Some of these incremental improvements included the introduction of automatic transmission, air conditioning, electric seats and windows, dual headlamps, and wider, softer-riding tires.

The advent of the microprocessor, imported from the electronics industry, created the next major wave of technological upgrading. Some key developments that emerged were the development of sophisticated engine control modules and the ability to integrate engine control with power train and chassis electronic systems.

According to an executive at one of the Big Three automobile manufacturers, their large revenue base allows significant R&D investments across the spectrum, including at the early concept and invention as well as ESTD stages. “But we do not do nearly as much basic science research as we did when our company’s original central research lab was created in the 1950s,” he says. Motivated by a growing focus on more reliable and profitable products, the vast majority of their research is targeted at product development. Also, given the importance of design and form in the automotive industry, a significant portion of R&D efforts is driven by design-related needs. Still, he notes, the large number of patents and licensed technologies owned by the company is indicative of the company’s commitment to early stage research.

The time horizon for new R&D initiatives is also significantly longer in the automotive industry than in other industries. According to our interviewee, long range projects have timelines of about 10 years to production, while short range projects often have timelines of 3 to 5 years. Research projects are assessed against measures of risk and opportunity, with most projects being medium risk, medium opportunity. In recent years, research goals have focused on environmental regulation and fuel economy issues. Fuel cell research, for example, is a high risk, high opportunity project for the firm.

BIOPHARMACEUTICALS

The biopharmaceutical industry spends 11% of sales on research and development activity, more than any other industry in the United States except for the computer software industry which spends 18% of sales on development (the equivalent of manufacturing expense for this industry).

Biopharmaceuticals include all companies that are involved in biotechnology research, gene mapping, and genomic-database building to identify and characterize the expressed genes of the human genome as well as companies engaged in the discovery, development, and production of drug and drug-related technologies.

The process of drug discovery itself is well understood and has been fine-tuned over many years. Incremental improvements are targeted primarily at reducing the fall-off rate of drug candidates at each stage of the discovery process. Very long product development cycles, high upfront development costs, and an unpredictable rate of success limits the ability of maturing firms in the biopharmaceutical industry to spend on ESTD work to develop new products. For many young firms, bringing their one core founding idea to market is the only goal that matters.

The companies we interviewed were relatively small companies, founded around a single product idea (or a handful of closely-related products). Early on, significant resources were targeted at proof-of-principle and reduction-to-practice activities. They stated that a considerable amount of R&D money was devoted to ESTD work in the years immediately following the founding of the company. Today, these firms are focused on developing their products for clinical trials and market introduction, with almost 90% of R&D dollars devoted to product development (see Table 8).

The chief scientific officer of a young developmental stage biotech firm told us, “We spend the vast majority of our research money on product development as opposed to ESTD type work. But two to three years ago, that ratio was reversed.” After developing the initial concept for a novel vaccination treatment, the idea had to be proven at the manufacturing level. Logistics had to be worked out for complicated procedures ranging from procuring uncontaminated diseased tissue samples from around the world to developing economical manufacturing processes for the vaccinations. With many of these operational and logistical challenges worked out, the firm is now heavily focused on developing the product for clinical trials.

The firm’s R&D budget has reached 25% of their burn rate, with nearly all of that targeted at developing their vaccination products for market entry. With no profitable revenue streams yet, R&D investments must be consistent with the firm’s very sharp focus. There is little money for new “blue sky” research and no latitude for high-risk early stage research out of the firm’s core business.

____________________
32. An example would be the penetration, now underway, of advanced adhesives, replacing welding, in the assembly of door panels and other parts of automobiles.

Return to Table of Contents or go to next section.

Date created: October 7, 2005
Last updated: October 12, 2005

ATP website comments: webmaster-atp@nist.gov  / Technical ATP inquiries: InfoCoord.ATP@nist.gov.

NIST is an agency of the U.S. Commerce Department
Privacy policy / Security Notice / Accessibility Statement / Disclaimer / Freedom of Information Act (FOIA) /
No Fear Act Policy / NIST Information Quallity Standards / ExpectMore.gov (performance of federal programs)