Thursday, 26 May 2016

Competitor behaviour revealed: an exercise in software patent data

So much ink has been spilled in recent times over the demise of patents in the software sector, particularly in the United States after the Supreme Court's Alice ruling, that one might be led to believe that there were scarcely any of them left standing. The truth is however very different.  There are tens of thousands of them, in the US and beyond. Their very large number makes it difficult to gauge their potential impact in the marketplace and their role in shaping the commercial direction taken by their owners. 

In this context, it is instructive to take a look at the case study, below, which Aistemos executed through use of its Cipher tool. This study, which was first posted on the Aistemos website here, gives some idea of the benefits that can be derived through the deployment of IP analytics.


Case Study: Global Patenting Strategy


The IP team wanted to optimise their own international patenting strategy through a better understanding of the patent filing patterns of the most similar companies in this sector.


The volume of data relating to the patent portfolios to be analysed (over [25,000] families in aggregate) meant that the traditional approaches to corporate grouping, clustering and territorial analysis would be too labour- intensive, slow and expensive. It was also stipulated at the outset that portfolio-to-portfolio analysis would be too course and imprecise a measure, meaning that competitive strategy had to be analysed at a technology (patent cluster) level.


Cipher was able to respond to the challenges as follows:


The starting point was the classification supplied to the companies’ own portfolios. In this way, a cluster-to-cluster matrix (a Cipher heatmap) was produced for each company and each technology area.

Figure 1: Company/Technology matrix
For ease of visualisation, the number of patent families in each area are not included.

It was then possible to analyse a strategy for any combination. The illustration below represents the Territory Waterfall for Microsoft in the area of Virtual Machines.

Figure 2: Territory Waterfall
As illustrated above, in relation to this technology, Microsoft seeks generally seeks protection in the US, and when filing internationally starts with China, Japan and Korea. For less than 25% of its portfolio does it seek broad protection in Europe. This analysis provides insight into both patenting and market strategy.


Having clustered all the competitive portfolios, it was possible to analyse spend - the cost incurred in obtaining and maintaining each part of the portfolio. Again using Microsoft as the example:

Figure 3: Microsoft's spend per patent family
While there was a definite correlation between costs and territorial scope (more territories, more costs), this type of analysis uncovers hidden truths. There are families (albeit extended) that cost well over $500k to obtain and maintain.


As a direct result of this analysis:

the patenting budget was increased to optimise their international patenting strategy.

IP analytics were integrated as a factor into the corporate competitive intelligence reports, increasing the relevance and importance of patenting strategy.

the company was able to implement portfolio reviews as a regular, fast and inexpensive way of monitoring the competition and improving their own patenting strategy.

1 comment:

  1. "There are families (albeit extended) that cost well over $500k to obtain and maintain."

    Would love to know the facts/ assumptions behind this. Official fees are easy enough to work out worldwide through grant and maintenance, but that leaves a lot in a $500k total spend. So what figures are you using/ how are you calculating the prosecution costs..?