Patent valuation is a hot topic that has been addressed in numerous discussions, publications and conferences over the past few years. Whether the purpose is to include patents in the balance sheet or to measure their technical relevance, to date the only consensus about patent valuation seems to be its importance and necessity.

Patent valuation is often misunderstood and confused with the principle of patent evaluation. Further, patent valuation seems difficult for several reasons. However, patent evaluation has come a long way and some methods have proved to be surprisingly efficient – provided that the context is established correctly.

Valuation or evaluation?

Wikipedia defines ‘valuation’ as the “determination of the economic value of an asset or liability”.

It defines ‘evaluation’ as “a systematic determination of a subject’s merit, worth and significance, using criteria governed by a set of standards… to help in decision-making; … any particular definition of evaluation would have to be tailored to its context”.

In short, valuation allows for an assessment of economic worth (how much?), while evaluation offers a systematic assessment of worth to help the decision-making process in a given context (should we?).

Although the list of metrics used for valuing patents is widely recognised by IP industry experts, the main challenge lies in the weighting of these metrics in order to suit the specific requirements of each industry. There are various factors of influence, such as the competitive environment, patent life expectancy, patent office citing habits (ie, the average number of references cited), geographical scope, industry litigiousness, covered markets, legal protection, licensing opportunities and best practice. These are undeniably factors to be integrated into the valuation algorithms. This is precisely where the difficulty lies and where experts seem to have trouble reaching a consensus.

More than 120 patent portfolio valuation metrics have been submitted, discussed and challenged by IP experts and business decision makers over the past 20 years. Some of those metrics that particularly stand out and are considered meaningful are:

  • patent litigation count;
  • patent re-examination count;
  • patent opposition count;
  • uncited prior art count;
  • co-pending prior art count;
  • average number of backward citations;
  • average patent portfolio age;
  • co-assigned art;
  • average number of inventors;
  • patent citation velocity;
  • average number of forward citations;
  • average forward citations by others;
  • predator presence;
  • company patent fences;
  • self-citations versus others’ patent citations;
  • international patent classification dispersity;
  • generality index;
  • originality index;
  • radicalness index;
  • claim length;
  • patent portfolio reassignment frequency;
  • average family size; and
  • total count of Tier 1 and BRIC counting patents.

Any evaluation process is contextual and should always take industry specificities and environment into account. For example, for the first metric listed, it is obvious that litigating habits and trends vary significantly from one company to another, and legal protection also differs from one country to another. Such differences may have a major impact on the way this metric is used and weighted.

Contrast uses

A global valuation model provides a single number per patent, representing its ‘economic worth’. It may be aggregated for any large corpus of patents and provide a single value for a large portfolio.

On the other hand, a goal-driven evaluation aims to answer a specific question, providing rationales for enhancing and accelerating the decision-making process, which leads to quicker and higher-quality documented decisions on specific issues.

Thus, choosing the right approach requires the identification of a clear goal. Here are some sample questions:

  • How much has the value of US companies’ patent portfolios grown in 2013?
  • Are my patents worth more than those of my competitors?
  • Are my 10 targeted patents worth licensing out? And if so, to whom ?
  • Which patent should I stop renewing this year?

Figure 1. Building a global valuation model

Click here to view image.

Global valuation models

Three main steps are required to build a global valuation model:

  • Estimate the value V for historical patents (based on renewal behaviour for the set of patents considered).
  • Compute predictor metrics P for the same corpus (numeric patent data such as citations, number of patent family members and claim length can be used).
  • Build the model.

Machine learning techniques can be used to estimate V = f (P).

The value V can now be estimated for any patent for which the metrics can be computed, as well as for any set of patents (eg, patent pools, patent portfolios).

Goal-driven evaluation

Defining the IP-based business goal drives how to make the best decision on an IP-related issue. Thus, asking the right question is essential:

  • Can we license out some of our patents? To whom?
  • Is this patent cluster offered to us for licensing a good match for our company?
  • For which patents should we stop paying annuities?

To answer these questions, licensing professionals use specific metrics to compare the patents at hand with comparable patents. There is a general consensus in the industry on what the most relevant metrics are. Among a list of more than 120 identified metrics, some are considered more relevant than others. For example, high-level, business-driven metrics include legal strength, originality, generality, radicalness, predator presence, citation velocity, citation frequency, number of years left until expiration and geographical coverage.

Such metrics contribute to the decision-making process and must be understood by decision-making stakeholders. The question being addressed defines the nature of metrics to be used.

Nothing is absolute

The core of the evaluation process consists of comparing metric value with the distribution of metric values for a comparable set (similar art). Values such as ‘originality score of 0.7’ have no intrinsic meaning for business stakeholders. Instead, benchmarking a portfolio with similar art gives a much better and clearer indication to stakeholders: “How do the selected patents compare to similar patents in the industry?”

What is a comparable set?

Evaluating a patent portfolio means comparing it to similar ones from a technical standpoint. This comparison sits on the principle of benchmarking, which requires the building of a comparable set that is as good as possible.

Building a relevant set of comparable patents requires expert patent searchers to explore all available techniques, applying best practices and tips, such as combining patent classifications (eg, international patent classifications, cooperative patent classifications, Japanese FI-terms and F-terms) with keywords, value-added content (eg, special indexing), competitors and citations.

Some systems offer optimised algorithms combining experts’ experience and best practices with computer value-added content, which allows the instant retrieval of comparable sets. Semantic search engines that combine citations and classification codes offer unprecedented precision and power for retrieving similar patents.

Business needs drive evaluation methods

Each phase of the innovation lifecycle raises specific business needs and an IP-related challenge.

Figure 2. Innovation lifecycle

Click here to view image.

Technology development: speed R&D

In the first phase of the innovation lifecycle, technology development, business decision makers must stress strategic challenges to drive their organisations.

Developing new technologies and penetrating a new market require key decision makers to understand industry ecosystems and make the best investment decisions to accelerate the R&D process. The key questions at this level address the need to identify essential patents and major players in order to assess acquisition opportunities:

  • What patents should we acquire in order to accelerate the development of our technology? Technology scouting can help to answer this question.
  • Who are the key players and can we license in their technology? A licensing-in assessment is then required in order to enable effective negotiations with such players.

Figure 3. Licensing-in methodology

Click here to view image.

Product development: mitigate risks

Launching a new technology requires absolute freedom to operate. It also takes balanced patenting activity to build a patent pool protecting such technology.

Go-to-market: sustain competitive advantage

When a technology or product is reaching maturity, decision makers ought to audit their patent portfolios regularly to assess their ability to exclude other players and keep them at a distance. Figure 4 explains the process for assessing the ability to exclude others.

Managing a patent portfolio also involves pruning weak and meaningless patents to reduce high renewal fees. It optimises patent spending and enables full control over the portfolio. Figure 5 shows the patent pruning methodology.

Figure 4. Assessing ability to exclude others

Click here to view image.

Asset monetisation: maximise profit

The final phase of the innovation lifecycle focuses on revenue generation – in other words, identifying, within a patent portfolio, which patents are good candidates for out-licensing.

Classifying a portfolio by good and poor candidates for licensing and identifying the most probable licensees is a first step towards effective negotiation. This process enables companies to maximise their intellectual assets’ return on investment.

The out-licensing assessment methodology is set out in Figure 6.

Figure 5. Patent pruning methodology

Click here to view image.

Figure 6. Patent pruning methodology

Click here to view image.

Negotiations can then start based on a patent portfolio’s strengths and weaknesses, compared to the industry average (as defined by the comparable set). Radar charts can illustrate the positioning of a portfolio by highlighting below-average, above-average and outstanding zones. Figures 7, 8 and 9 show how an evaluation can be made based on three key aspects: legal, technical and geographical. The red line represents the portfolio being assessed for out-licensing opportunities. Where the red line falls within the orange zone, the portfolio is below average for such metric. Where it falls within the yellow zone, the portfolio is above average. Finally, where the red line falls within the white zone, the assessed portfolio indicates an outstanding performance, compared to the industry’s comparable art.

Figure 7. Technical metrics

Click here to view image.

Figure 8. Legal metrics

Click here to view image.

Figure 9. Geographical metrics

Click here to view image.

Technical metrics: For the purpose of licensing out, a set of 19 technical metrics has been retained in order to assess the portfolio’s strengths and weaknesses compared to the industry’s comparable patents.

Legal metrics: For the purpose of licensing out, a set of 11 legal metrics is proposed in order to assess the portfolio’s strengths and weaknesses compared to the industry’s comparable patents.

Geographical metrics: For the purpose of licensing out, a set of 14 geographical metrics is proposed in order to assess the portfolio’s strengths and weaknesses compared with the industry’s comparable patents.

Ranked patent candidates: Each patent in the portfolio assessed for out-licensing is ranked based on a set of the top seven metrics. A rank score is obtained for each individual patent.

Prerequisites for patent evaluation

The evaluation of a patent portfolio is based on the principle of benchmarking and comparison with comparable art on criteria such as legal, technical and geographical measures (metrics). Computing such metrics involves gathering data from many different sources and providers, then aggregating it in spreadsheets and extracting results in a visual form that is clear and intelligible for decision-making stakeholders.

An alternative is to rely on computer-based systems to automate and compute all metrics in just a few minutes.

Whatever the means, the following items are prerequisites in order to carry out an efficient evaluation process:

  • worldwide invention-based family-grouped patent database;
  • up-to-date normalised legal status information;
  • reassignments;
  • oppositions;
  • litigations;
  • licensing;
  • alive/dead state;
  • comprehensive citation coverage;
  • clean and normalised patent assignee names;
  • up-to-date patent classifications;
  • deep indexing (semantic technical concepts);
  • sophisticated similarity algorithms;
  • citations (forward and backward);
  • patent classifications (international patent classifications, cooperative patent classifications);
  • semantic indexing;
  • broad choice of pre-defined valuation metrics (120-plus);
  • full user customisation on metrics (weighting); and
  • fast on-the-fly metrics computing capacity.

Figure 10. Ranked patent candidates

Click here to view table.

Patent evaluation: bringing intellectual property to the board

While patent valuation serves primarily financial needs, evaluation serves a different but more strategic purpose. While patent valuation is extremely difficult to carry out and is still much discussed and challenged by IP and financial experts, patent evaluation has become an essential activity and asset addressing the needs of C-suite decision makers. Patent evaluation methods allow for the extraction and highlighting of key business indicators hidden in patents, which in turn provides the board with a better understanding of the business environment and trends, thus supporting and speeding up the decision-making process throughout the innovation lifecycle.

Renaud Garat

This article first appeared in IAM magazine. For further information please visit www.iam-magazine.com.