Large companies open facilities in new locations all the time. The reasons may be to access new markets, to expand their manufacturing and distribution footprint, or to access talent and open innovation in an industry hub.
When they do this, an early step includes contacting the economic development agencies of all the candidate jurisdictions. This is to solicit data and information to help make the final decision on where to locate the new facility.
These jurisdictions know they are in competition with other jurisdictions for this new infrastructure investment by the company. This type of private sector investment is coveted because it brings jobs to the area, and it expands the local economy and tax base. It results in a scramble by economic development agencies to provide these companies with economic information.
Of course, the company knows this too. In addition to data, it seeks incentives. For example, when Amazon planned to open a second head office in a new city, it made the incentive package an open competition among candidate cities.
The medical device company planning an R&D expansion
When I was doing economic development, there was a large global company with a medical device division that was seeking to make a major expansion in one of its four R&D facilities. These four facilities were in China, Austria, Belgium, and Canada (the cities will remain undisclosed here). The R&D performed at all four of these sites was medical imaging.
Apart from the incentive package, an important deciding factor was in which of these four R&D locations would an expansion bring an out-sized improvement in productivity? Hopefully, the answer would be the Canadian R&D site. I had to answer this question with solid data, with only the above details as background.
I will work through this example as a master class, because the data technique used here is useful to anyone working in economic development, technology scouting, technology diligence, competitive technical intelligence, and also to an entrepreneur-in-residence, and especially to professors or grad students seeking an application of their technology.
Patent intelligence is often a good starting point
The first step begins by brainstorming where to find information about the company. There are many ways. The Society of Competitive Intelligence Professionals is a good starting point to find resources on the practice of competitive intelligence performed ethically.
However, when it comes to technology, particularly where we need to deal with people actually doing the work, patents are often the first choice. It concerns assets of business value, and it gives solid data and technical details.
Patent intelligence is a sub-specialty in itself and will be the topic here. I won’t get into explanations and nuances of patents, patent applications and patent families in this post. The point is to illustrate how patterns are found through patent searches.
The first hypothesis in this case is as follows:
Can we find the patents being generated by each of these four R&D sites in China, Austria, Belgium and Canada? Will the results show that the Canadian site is most productive?
The second hypothesis is as follows:
What is the patent activity in each of these four cities, in the same type of work being done by this company?
Will it show that, of these four global cities, the Canadian city generates the most patents in this specialty area?
This helps make the case that the company’s R&D site in this Canadian city will benefit from access to available talent in the area and will also benefit from open innovation.
The world intellectual property organization (WIPO) database
There are three important patent databases, plus Google:
- The World Intellectual Property Organization (WIPO) database is usually good first choice to use.
- The US Patent Office and Trademark office (USPTO) database is useful when the U.S. context is important.
- The European Patent Office database (called ESPACENET) is also very powerful.
- Sometimes the Google patent search engine is useful for some key word searches.
The WIPO database has three search options: simple, field combination, and advanced.
A quick way to start is to use the WIPO database field search option to look up the company’s patent filings under the “APPLICANT NAME” field. This gives a first look and an idea on what needs to be done next to narrow down the search.
Figure 1 shows the results if “Apple Computer” was entered as the “applicant name.” The search returned 3527 patent filings.
Like Apple, the company in this case was a huge enterprise that had thousands of patent filings, many of which were not even in the imaging business that was relevant to us.
We need a way to filter out the hundreds of irrelevant patents first. This would be to search by International Patent Classification. Searching by IPC code is usually a very useful filtering step to get to the desired list of relevant patents.
International patent classification (IPC) codes
When scanning through the thousands of patent applications for this company, the titles allow relevant ones to be identified. How to know which ones are relevant? This comes from subject matter knowledge and from experience. My next post will address this.
Once relevant ones are found, note the IPC code(s) for the patent. A pattern should emerge.
It was found that the company’s four R&D sites were developing software for storing medical images, they were developing algorithms to analyze these images, and they were developing better interfaces for users to interpret these images.
For this company’s imaging division, their patent output was covered by five IPC codes:
- G01T – Measurement of nuclear or x-radiation…
- G06T – Image data processing or generation, in general
- G03B – Photography; cinematography…
- G01B – Measuring length, thickness or similar linear dimensions…
- H04N – Electric communication technique
These are very general IPC codes that cover a huge range of patents. The key to making them useful is to find the subsets that link to the company, the inventors, and the locations.
Filter using Boolean searches
These codes can be used to filter out the patent filings of interest. A focused search can be done on the WIPO database advanced search option using a Boolean search string, for example:
PA:(“company name”) AND IC:(G01T OR G06T OR G03B OR G01B OR H04N)
This returns a few hundred patents that can be analyzed using a further series of Boolean searches:
1) The inventors’ nationality and also the patent agents’ addresses allowed the four sites to be identified.
2) A confirmatory patent search for these inventors’ names also revealed previously unknown business subsidiaries that were filing additional patents. A search for these business subsidiaries revealed more patent filings.
3) The patent filings need to be grouped by patent families, to avoid multiple counting of the same patent invention.
The results: patent output by each of the R&D sites and by the surrounding industry hub
The results did indeed reveal that the Canadian R&D site was the most productive in patent output by a wide margin.
The next step was to perform a search for patents with these five IPC codes in each of the four global cities. This was to determine if each of these four cities are industry hubs for this type of technical capability.
The results showed that the Canadian city did indeed have a respectable hub of intellectual property (IP) activity in this area. The Austrian and Belgian cities did not. However, the Chinese city had a massively outsized level of IP activity. All the familiar global companies had R&D sites in this Chinese mega city and were filing patents. On a graph, it just went off the chart.
Telling the results through the story
The common economic development presentation is an “asset map” listing names of universities and tech companies in the area.
With this data, we can do much better.
We can provide to this company a very compelling bar chart showing the patent output of each of their four R&D sites.
One would expect they would know, but their executive management do not know.
Consider all the obfuscation by site managers promoting their site to the executive management, talking about how good each of their sites are. Patents don’t lie. Such objective data from us, an outside party, is a good reality check. An appendix of these patents can be included in the submission for good measure.
Next, we can make it colourful. We can mention the names of the inventors at each of these sites. We can cross check them on LinkedIn, on academic papers, and on thesis dissertations. Ages can be extrapolated from LinkedIn. If they have a masters or doctoral degree, a thesis dissertation database will give their exact date of birth.
Now we can construct a career story line of the company’s most prolific inventors, and of their site managers. We can overlay a geographic map of the tech social scene near their R&D site address to present a hypothetical scene where they interact with external talent. It’s always fun to tell executive management some detailed stories about their employees to catch their attention.
In the end, this medical device company did choose their Canadian R&D facility for a $200 million expansion.
Patent analysis as a strategic tool for R&D and for business
In this case study, the mystery to be solved is to find the threads that connect the relevant patents to what is happening at each R&D site.
Beyond this one example, the above case study illustrates that patent databases, combined with other data searches, can be very useful for a range of applications in the business of technology. Every project is a mystery to be solved, or an opportunity to mine for new value from seemingly incomprehensible information.
I once told an MBA graduate what I was doing for a technology scouting project. She said it sounded so cool. I was surprised to hear that it sounded glamorous. Most of the time spent on this type of work is sitting alone in a sunlight-deprived room in front of a computer.
In retrospect, this practice is becoming compelling. Who ever thought eSports would be so popular too? In searching databases, one needs to work out and be in good physical shape be able to sit up straight in front of a computer to avoid ergonomic injury, and to have the energy to avoid drowsing off reading highly technical details.
Joking aside, patents reveal what is happening at the cutting edge of any field. It is exciting to discover secrets lurking behind each technological field, to find the connections of the people working in the area, and to seek that insight that will allow new actions to be taken.
I think an analogous portrayal of this type of work is by the character Penelope Garcia in the television series Criminal Minds. She works at the FBI’s computer lab to support the Behavioral Analysis Unit, using creativity and data analysis to dig up hidden information. Just as she attempts to foreshadow the intent of criminals, the effort here is to discern the inventive mindset of people and the strategy of companies.
I’m not suggesting that I am doing glamorous or cool work.
I want to make this approach appealing to any entrepreneurial-minded scientist, engineer or graduate student or professor to plan research and development.
By junior year in undergraduate university science programs, we are taught how to search for publications in a library catalogue. Patent searching is another tool, but can yield insights that can guide work into much more valuable directions.
Here is a good reference book on how to analyze this type of information. It was written by an intelligence analyst from the Central Intelligence Agency (CIA). It describes fourteen different tools to analyze complex information to solve problems and to make decisions:
The Thinker’s Toolkit, by Morgan D. Jones, first published 1995 by Time Books, present revised edition, 2009 by Crown Business.
Despite being written in 1995, it is still an excellent reference of the foundational, must-use frameworks.
Quick take: when it comes to patent analysis, the tool of Sorting out the Chronology and Timelines is the most useful analytical method.
I have one update tool to add, and that is performing a Network Analysis. Back when this book was published, the internet (via the World Wide Web) had not yet come into shape. The power of the internet today makes network analysis one of the most powerful of tools, as shown in the above case study and in an earlier post.
A key theme of this blog concerns how globalization has a big influence on the development of national economies and on business competitiveness.
This case study is yet another example of this theme. A global company was seeking to access human talent from anywhere in the world: Canada, China, Austria and Belgium. It also sought government incentives, setting up a competitive bid among countries.
The financial incentive offered by the government was quite substantial. On a near term outlook, and on the political news cycle, the facility investment provided jobs.
However, I have worked in R&D outposts for many global companies. The policy question never asked is this:
What are the long term implications for each country’s economy, and what should be done to make sure that any government incentives are not a subsidy to a foreign company that declines in economic value over the longer term?
Foreign companies can and will shift resources as quickly as industry and market cycles change, resulting in closures and job loses. It is not up to companies to build sustainable industry hubs. It is up to the people within that hub and to government policies for that hub.
As I said before, I always follow as many of my projects as possible, to learn from them. In this case, the incentive was so appealing to this global company that the VP of that company was assigned a new role to seek government incentives for other units within the company.
As for the site itself, its productivity did increase somewhat for a few years, but it never assumed a greater role as an anchor company in this city’s industry hub. In fact, its productivity appears to be declining now, as the company is adding jobs to a new R&D site, at a new innovation hub for this technology field that is emerging in another province in Canada.