MASTER YOUR PATENT ECOSYSTEM
… and get “big data” savvy in the process
“Patent Landscaping is hard – really hard.” It is basically only for those with hundreds of hours (or more) to spend, hyper-specific expertise, access to numerous data sources and tools, and high level of training in big data and specialized landscaping techniques. And of course you need advanced technical and legal training to understand it all properly, right?
Given that premise, you should be. However, the premise is mostly WRONG and doesn’t really address the various REASONS for conducting patent landscaping analysis in the first place. Further, the long-term value of these elaborate and voluminous reports is … questionable. And as an analyst, manager, or executive, the idea of embarking on one of these deep-dives is daunting (and demotivating). But fortunately, pretty much anyone can (and should) become more intimate with their technology/IP ecosystem and enhance their personal “IP KSA’s” and collective value to an enterprise.
As Ram Charan says in his book, The Attacker’s Advantage:
“Business leaders need competence in digitization – at least enough to know the right questions to ask the experts, along with the imagination to find ways for mathematics to help them redesign the consumer experience.”
Replace digitization with “IP” or “patents” and change the ending to “redesigning the innovation life cycle” and you have an important core premise of this Patent Landscaping series — get smarter, smart enough to know the questions to ask and precisely who to ask (further, we believe anyone can begin to answer some of these questions themselves). “Big data,” data analytics, and algorithmic thinking are things every enterprise has to deal with one way or another… hopefully not in the way the cartoon depicts below.
However, answering the call of “big data” and IP analytics is still an adjustment in mentality and expectation. A recent WIPO report stated:
“Concluding a comprehensive, definitive patent landscape in a major technological field such as HIV/AIDS treatments can be a massive endeavor, requiring considerable resources and expertise. It potentially entails an expert review of thousands of complex documents and fine assessments on their legal and technical content. A fully global landscape would strictly entail expert searches in over 100 patent offices worldwide. Any ‘finished’ report will be out of date within days, as further patent disclosures are published online. Keeping the landscape up-to-date for continuing reference can be just as resource intensive as its initial development. But the high cost and technical barriers are progressively declining. What once would have been a costly strategic landscape can now be prepared free of charge from a laptop with good Internet access.”
This statement from a WIPO report (found at http://www.wipo.int/wipo_magazine/en/2008/04/article_0005.html ) reveals a core reason why patent landscaping is often problematic and sometimes not terribly useful. Specifically, the massive effort and reliance on numerous experts and complex resources, both legally and technically – and the fact that the report “will be out of date within days.” The idea that these can now be done “free of charge” with any laptop isn’t really true (unless a bunch of really smart people are donating their time to these), but conceptually they are on the right track – at least if you take my approach of “start simply and build/iterate from there.”
Even in the best cases, we have heard a similar refrain from clients that “we have produced some excellent analyses and they profoundly helped us prioritize investments, identify licensing and acquisition targets, and react more quickly to emergent tasks (at that time)… But, it was incredibly costly (and worse, painful) to assemble so we haven’t done it again.”
Even if you don’t feel ready for this shift, we can help. And it IS coming – as Marc Benioff, CEO of Salesforce recently said in an interview with Fortune:
“We need a new generation of executives who understand how to manage and lead through data. And we also need a new generation of employees who are able to help us organize and structure our businesses around that data… we need more data science.”
In response to the question about his level of confidence that non-experts can consume this data, he stated very succinctly: “Very.” We agree with this premise wholeheartedly.
The McKinsey Global Institute conducted a Global Study on big data and upcoming analytical skills deficits entitled “Big data: The next frontier for innovation, competition, and productivity.” It provided a stark vision for all knowledge workers and managers to heed:
“There will be a shortage of talent necessary for organizations to take advantage of big data. By 2018, the United States alone could face a shortage of 140,000 to 190,000 people with deep analytical skills as well as 1.5 million managers and analysts with the know-how to use the analysis of big data to make effective decisions.” You can read the full report at http://www.mckinsey.com/insights/business_technology/big_data_the_next_frontier_for_innovation .
Why do I bring up the Ram Charan, Marc Benioff, and McKinsey Global Institute quotes? Because IP is not immune to these trends and in fact the gaps between those that have an analytical approach and those that don’t in their IP ecosystem/lifecycle is already wide and getting wider.
The rest of this section will give an overview of Patent Landscaping & Analysis and address some of the problems and opportunities in this evolving field. Examples here are developed using the AcclaimIP Landscape Matrix tool, but the concepts can be applied to other tools or with some data extracts and Excel work. The core concept here is simplicity and transparency – we are all used to creating and reading Excel spreadsheets: this model uses 2 axes (query intersection) and potentially a third dimension of trending/time. As you build it, your output is intuitive and transparent.
If you want to skip ahead to the examples, click the list below. I’d recommend starting with the first one:
- Company and Technology
- Problem and Solution
- Conceptual Intersection
- Technology Navigator
- Company and Product References
- Company and Non-Patent Literature Citations
More advanced use cases:
Non-traditional uses of matrix analysis:
The reality in IP is that so much of the visible press around patents revolves around law suits and what I will call “errors of commission,” but the unquantified and very real drain on corporate ROI and the economy at large is probably much more about “errors of omission” due to simply not having enough good information fast enough to take action, identify opportunities, ask the right questions in a timely way, or read the market and where it is heading. I think this is generally what both Marc Benioff and Ram Charan were also getting at, in their own way.
My goal here is to share some simple techniques to get started with patent landscaping. There is a lot of value to be gained from starting modestly and building some momentum – in fact, this is true about most difficult and complex tasks. And equally important, I will discuss developing a framework that is reusable and can be iterated.
WHAT IS A PATENT LANDSCAPE?
First of all, there are many flavors of patent landscapes, from the full-blown research reports like the WIPO cites to 3D relief maps to various matrices/grids and amalgamations of various analyses. What is really important (unless you are just trying to make pretty slides) is the reason for wanting to do landscaping in the first place. Patent Landscape analysis is typically conducted to understand one or more of the following:
- Understand the IP for products and technologies important to your enterprise’s present and future
- Identify key technology players and their relative IP strengths
- Discover areas for white space inquiry, Freedom-to-operation (FTO)/patentability
- Understand your competitors, upstream and downstream partners, and potential acquisition/acquirer targets (i.e., your IP ecosystem) and their IP holdings and trending
- Tracking the evolution of a landscape relative to active IP development and/or R&D
The last item on this list is really an important one to emphasize. Too much analysis is a one-off in the IP arena and too often the touch points in the lifecycle of products and their IP are just that: discrete points rather than embedded through the entire cycle. I’ll address this more another time, but for now I’ll simply say that this is unnecessary but stems largely from the “integration” problem (IP is a naturally cross-functional issue – most organizations don’t deal with these well).
We hope you are inspired enough to try some of this yourself. To adapt a concept coined by the famous Swede Svante Arrhenius, it is all about “activation energy” – you just need a push to get started and you can become an educated builder and consumer of patent landscapes – regardless of your role in an organization, you will be better armed in the knowledge economy.
We believe the key aspects of the techniques outlined here are:
- Easy to build simple model;
- Iterate to your heart’s content;
- Build hierarchy (AcclaimIP) if you think it enhances the extraction of knowledge;
- It is relatively easy to digest – you build your queries, it gives you numbers and trends (sparklines) – if you have ever worked with spreadsheets, it will be naturally intuitive;
- In the AcclaimIP solution, you can click through to the actual patent lists represented by each number (transparency and efficiency for drilling in);
- There is no barrier between you and the data: again, transparency!
- You can apply your own internal knowledge and expertise of products and technologies – as little or as much as you want, and integrate others into your analysis
A word about 3D “relief map” landscapes. First of all, they have their place in understanding and conceptualizing a group of patents in terms of their “technology” or “technical category.” They attempt to put a group of patents into “like groups” or clusters and map those clusters to show cluster “closeness” or similarity. The relief shows the depth/volume of patents in particular areas.
They are useful in:
- Getting a feel for the various technological/conceptual areas represented in a sample of documents, including relationships between those concepts (proximity) and the relative #’s associated with each concept (relief).
I am not a huge fan of them because:
- They put layers between me and the data (my biggest issue)
- They are often misused or misunderstood (why #1 is deadly)
- Basically underlying it is a clustering engine, which has complex rules and may be good, really good, or relatively poor (and this may depend on the areas of technology). Is the clustering engine using ad hoc (dynamically generating the clusters from ONLY the set of documents of interest) or a priori (set for the entire corpus and unchangeable, regardless of corpus in view) and does this even matter? I think it probably does, but I can’t prove it and I doubt the question is even asked 99% of the time.
- The clustering engine isn’t transparent – in other words, I am taking a leap of faith that it is good; and further, I can’t explain it to a colleague, boss, or client – at least not with high precision. How do I then apply refinements (or can I even do that) to the model if I don’t know why I get the results I am getting?
e.g., “I see that there are two areas of focus over here, “titanium widgets and buckyballs, how are those derived”?” It is either a clustering engine (OTS, custom, or hybrid; real-time or a priori) or it is using some built-in taxonomy classification system with proximity values (like an adapted patent classification system or some other custom system). Again, the tool may be great, but the analyst/manager is left not knowing exactly why and very likely has little ability to figure it out.
- They are pretty, but are not very good at conveying digestible and useful information unless something is highly visually obvious and you are extremely trusting of the data, the clustering engine, and the implementation of these various pieces together.
- At its core, these impressive diagrams show basically two dimensions, along with some graphical representation of proximity. They show you technology/concept clusters and relative #’s through the heat-map/topography.
Any other dimensions (time, assignee, etc.) are handled through the selection of documents represented in the first place, which makes this laborious to interpret if you need more multi-dimensional analysis.
Anyway, onward to the first landscape example.