The EPO’s Vision (III) – Quality

Kluwer Patent Blog
March 5, 2018

Please refer to this post as: , ‘The EPO’s Vision (III) – Quality’, Kluwer Patent Blog, March 5 2018,


This will be a very long post (sorry, dear readers!), yet it cannot and will not be comprehensive enough to cover all aspects of this topic. In summary, my take on EPO quality is that not all is bad and a lot of people are still doing a good job. However, it is presently not at all the case that quality is outstanding or “setting worldwide standards”. Even worse, the trend of quality is downwards, which is most likely caused by the current EPO policy focusing too much on “production”. Honesty and a sober and realistic approach by EPO management would be needed to really improve quality and to bring the EPO closer to its vision.


The EPO strives to set worldwide standards in quality. Official communications from the EPO emphasize this time and again. President Battistelli writes: “But no matter the project or initiative scheduled for 2018, one issue at the EPO will take precedence above all others – quality. 2017 was a significant year for quality at the EPO and has given us a strong momentum to take forward into the next twelve months.” Vice President Casado writes “While we have kept pace with handling the rising number of applications by increasing productivity, quality takes precedence above all other considerations at the EPO.” Very good, so at least the priorities are set right.

Where does the EPO now stand with regard to quality? Looking again at the official communications and reports first, it seems that quality at the EPO is nothing but an incredible success story. And what is more, you can simultaneously rise both quality and productivity! In the EPO Gazette of 2017, the President wrote:

“So far, we’ve done a good job of ensuring that quality and productivity are rising simultaneously and we’ve shown that by the results in recent years – both in terms of internal quality targets and also user satisfaction. Once again we were ranked first among the IP5 offices and there has been increasing satisfaction in our User Satisfaction Surveys. But what will separate us from the pack (sic!) in the future is the ability to provide quality patents in a timely manner.”

To attain this goal, the President wrote in the same Gazette that “internal quality objectives continue to be set and achieved.” and in the very next sentence he continues: “Each year our staff attain more ambitious goals”. Hmm, so is there still something to improve? Not much it seems, since according to the EPO’s Quality Report 2016 as many as 80% of “Users” are “satisfied or very satisfied” with search and examination services, 17% find it average and only 4% seem to be “not satisfied”. Likewise, regarding patent administration services, even 87% of “Users” in 2016 were “satisfied” or “very satisfied” and only 4% “not satisfied”.

So everything is in perfect order and the quality of EPO services is fantastic.


I find “quality” an extremely difficult subject to cover objectively. Firstly, how to measure it? User satisfaction surveys seem to strongly depend on the participants and the questions that were asked. JuVe, for example, published the following survey results on their website at about the same time as the EPO, i.e. in 2016:

“Zunehmend Sorge bereiten der Industrie dagegen mögliche Qualitätsprobleme als Folge interner Querelen, die das Münchner Amt nun schon seit mehreren Jahren beschäftigen. Nur noch 46 Prozent der Umfrageteilnehmer sind mit der Qualität der Patenterteilungsverfahren zufrieden, 54 Prozent sind es nicht. Auch in Bezug auf die Beschwerdeverfahren sieht eine knappe Mehrheit von 50,2 Prozent ein Qualitätsproblem.”

“Industry is increasingly worried about possible quality problems as a consequence of internal struggles that keep the Munich Office busy for several years by now. Only 46% of the survey participants were satisfied with the quality of the patent examination proceedings, 54 were not. Also in regard to appeal proceedings, a small majority of 50.2% sees a quality problem.”

The EPO figures in the Intellectual Asset Management (IAM) survey seem to be better, though not nearly as good as the EPO’s own figures.

According to Wikipedia,

Quality is a perceptual, conditional, and somewhat subjective attribute and may be understood differently by different people. Consumers may focus on the specification quality of a product/service, or how it compares to competitors in the marketplace. Producers might measure the conformance quality, or degree to which the product/service was produced correctly.

So for what it’s worth, let me give you my own take on what I perceive the quality of EPO products to be in these days:

1. I think that the overall quality is still mostly satisfactory, but not as good as it could or should be. I would rate it perhaps with a 3 on a scale from 1 (very good) to 6 (unsatisfactory), on average. I will provide a few examples below that I hope will justify my rating. I am perfectly happy to concede that other sensible observers may come to a slightly better or worse rating.
2. In normal times, the quality of EPO products only changes relatively slowly over time.
3. Having said that, I do think that quality has decreased over the last two years.
4. My greatest concern is the long-term impact of the current policy of EPO management, which I think is very much at odds with the stated objective of “quality takes precedence over all other considerations”.

In this contribution, I will try to strictly separate quality (in content/accuracy) from efficiency (speed), which will be subject of a different blog. Note, however, that many observers/surveys may confuse the two for understandable reasons. In the end, no patent office will be rated with “good” if its office actions or decisions take forever.

1. Quality – Current Ratings

Most, but certainly not all, office actions that I get to read are reasonably well substantiated and provide an acceptable basis for proceeding further. The same is true for decisions by the opposition divisions. I do not have my own well-substantiated view on the quality of searches – so, dear readers, you may wish to chime in here.

I also do not have the time to systematically study the quality of the patents granted, but my general experience-based impression is that if and when the EPO errs, it mostly errs in favour of applicants. Note that this will not necessarily generate complaints or negative ratings in user surveys. If anything, it may rather have a positive impact. Thus, the dramatic increase in patent grants over the last two years will not necessarily be accompanied by great user dissatisfaction. In the end, every applicant gets what he/she wants, and “only a fraction” of patents granted have a scope that seriously disturbs and annoys competitors to an extent that they will file an opposition. Moreover, the budgets of most companies are pretty tightly constrained, with the result that there is not necessarily more money for filing oppositions now than in the past. I would therefore caution against taking too much comfort from positive ratings in undefined “user surveys” and from the number of oppositions filed per year, which, as I would have expected, has not increased with the strongly increasing number of grants. It has even decreased somewhat from 2014 (5%) to 4% in the years 2015 and 2016. In my view, none of this is a reliable quality indicator.

The EPO’s own figures in the Quality Report 2016 seem to suggest that its Directorate Quality Audit (DQA) is happy with about 85% of the intentions to grant (R 71(3) Communications), whereas the number of search reports found “compliant” by quality audits is in the order of as much as 95%. This in turn means that 15% of the (internal) decisions to grant are not found compliant with the EPO’s own standards.

The EPO’s representation of these results is interesting. Here is the relevant paragraph from the EPO’s Quality Report 2016 2016, p. 18, with my emphasis added.

The EPO’s Directorate Quality Audit (DQA), which is placed under the direct control of the President, audits the compliance of products delivered by patent examiners and patent administration with legal requirements. DQA performs annual audits on European and international search reports and on applications proposed for grant. Furthermore it audits opposition and refusal decisions bi-annually. Patent administration products and processes are audited based on risks identified in this area. A detailed analysis of approximately 925 search and examination procedures per year takes place. The 2016 results of the quality checks were positive: the objective for compliance in search was exceeded and the objective for classification was met. The audits produced recommendations for improvement that are being addressed by specific actions.

It seems to me a bit of an attempt to embellish the situation that the rate of compliance in grants is not mentioned or critically commented in this summary. Perhaps it is hidden behind the cloudy language “recommendations for improvement”, which the audits produced and which are (of course) being addressed, albeit not with any measurable success at least between 2015 and 2016. But I would personally not call the fact that about 15% of all grants are non-compliant with the EPO’s DQA’s own standards a “positive” result.

Let me be clear: I like the fact that the EPO has produced a comprehensive Quality Report for 2016 and I would encourage it to compile such reports in the future. However, when reading and trying to understand the 2016 Report, I could not resist feeling that it uses a too bright colour palette to depict the present situation. Let me give you one more example:

Patent invalidity in Europe is very low. This is illustrated by the example of Germany, the main validation country for European patents, and one of the main European jurisdictions. (…) These invalidated patents represent less than 0.01% of the total number of granted EP patents valid in Germany. 

The figure of less than 0.01% may be correct as such, but it is almost certainly misleading. Only a minute fraction of EP patents is attacked in German nullity proceedings. Of these patents, about 70-75% are found wholly or partially invalid by the German Federal Court every year. Note that I am not arguing here that this figure would mean that 70-75% of EP patents are completely or partly invalid, since most nullity plaintiffs obviously select the patents they want to attack under the aspects of (a) whether and how the patent disturbs their economic activities or plans and (b) the chances of success. I would, however, posit that the figure of 70-75% of invalid or at least partially invalid patents tells you more about patent quality than the absolute figure of 0.01%.

The EPO’s second observation about Germany is that

“In 80% of infringement cases in Germany the defendant does not even try to challenge validity. Indeed, there are approximately 1200 infringement cases in Germany each year, compared to only 250 nullity cases (Bundespatentgericht, Annual Report 2014). This suggests that in the vast majority of cases the potential infringer does not see any chance of challenging the validity of the patent concerned.”

Again, and with all due respect, I think this is wrong and misleading. Firstly, the counting of infringement cases by the infringement courts is different and more generous than the counting by the Federal Patent Court. For example, if a patent is enforced in parallel by an infringement action and a request for provisional injunction, this counts as two infringement cases, but the invalidity counter-attack only counts as one. Secondly, a German nullity action may only be filed once no opposition proceedings are pending (or can be made pending) anymore, which means that the only available counter-measure in many cases is the filing of an opposition or an intervention in pending opposition proceedings. Thirdly, the defendant in an infringement action may obviously be of the view that he has sufficiently good chances to win by pleading non-infringement only. Therefore, the discrepancy between the figures of the infringement courts and the figures of the Bundespatentgericht is certainly not due to “the infringer not seeing any chance of challenging the validity of the patent concerned”. Some reality check may be appropriate here: In my entire professional life of now more than 25 years, I have indeed reviewed and attacked many patents, but I have never come to the professional conclusion that there is “not any chance of challenging the validity of the patent concerned”.

I must admit that statements like these in a Quality Report of the EPO, which are obviously made with an intent to assure readers that the quality of EPO products is fantastic, make me a bit suspicious. Maybe the German Patent and Trademark Office or the UKIPO might wish to carry out their own and independent validity check of a random sample of EP patents – it would certainly be interesting to compare these results with the EPO’s. Moreover and more importantly, it would provide the EPO’s Administrative Council with a truly independent quality control audit. I am sure that the EPO’s DQA’s are doing a good job, but I am also certain that they would not be allowed to publish any result that would put the EPO in a negative light. Readers be reminded of Article 20(2) of the Service Regulations according to which:

A permanent employee shall not, whether alone or together with others, publish or cause to be published, without the permission of the President of the Office, any matter dealing with the work of the Organisation. Permission shall be refused only where the proposed publication is liable to prejudice the interests of the Organisation.

So we can be 100% certain that if the President wants the quality of the EPO to be “excellent” and “setting world standards”, and if he thinks, for whatever reason, that it is in the interest of the “Organisation” that the current quality of the EPO products is praised as much as possible in order to justify his agenda, then the quality will be praised exactly like that. Any more sober or realistic rating by an EPO employee runs the risk of “prejudicing the interests of the Organisation” and may have personal consequences for him or her. It is bitter that I have to write this so clearly, but I would urge the EPO’s supervisors to bear that firmly in mind when looking at the EPO’s official figures.

2 Quality – “Stickiness”

My second thesis about quality at the EPO is that it does not dramatically change overnight. The reason is quite simple: Well-trained and experienced examiners and formalities officers will generally continue to provide good products, at least unless they are pressurized to an extent that this ceases to be possible. As I argued before, you cannot turn the conveyor belt faster every year, thus I find the EPO management’s stated objective for 2018 to increase the number of “products” per employee by another 7% both ridiculous and dangerous. We may live in Modern Times, but even there turning the conveyor belt faster has not really worked and, on a more serious note, the time necessary for a careful review of a case or a thorough prior art search cannot simply be shortened ad libitum. This is the difference between a car manufacturer and a Patent Office, and I specifically dedicate this comment to the AC’s Chairman Dr. Ernst, who gave me the impression that he does not (want to) comprehend this connection.

But coming back to my second thesis, I think it is true that “quality” will only slowly and incrementally change under “normal” circumstances. The EPO’s quality report 2016, for what it’s worth, seems to confirm this by and large, see particularly figures 9, 10, and 12, whereas I do not believe that the sudden jump shown in figure 14 between 2013 and 2014 has to do with real-world quality metrics. Furthermore, I find it hard to make sense of figure 15, where some bars or parts thereof seem to be missing.

3 Quality – Direction of Change

Why am I of the opinion that quality has somewhat decreased over the last few years? A few observations:

My firm files more than 3000 EP applications every year. This means we have certain standards and routine procedures that we follow in their prosecution. One of them was to make use, as a standard procedure, of the possibility to waive a second communication according to Rule 71(3) after we have drawn the EPO’s attention to errors in the first communication (which are quite frequent!). Before this backdrop, here is an internal email of one of my partners:

About 10 % of 71(3) texts that I receive include a Druckexemplar including errors in compiling the text within the EPO. These errors can be corrected when responding. If the amendments are trivial, I have tended to file the response with a waiver. No longer.
Sadly the EPO’s incompetence has extended itself on 2 of my files to the Druckexemplar attached to the “Information” Communication following such a waiver by introducing new random errors. These are much less easy to correct as the file irrevocably enters the EPO’s granting machinery about 3 days or so after the date of the “Information” Communication. In my latest case our Office did not receive the “Information” until 7 days after the date which it bears. Too late to halt grant.
I am swiftly coming to the view that filing a waiver to the first 71(3) Communication is no longer appropriate in view of the current abysmal levels of quality control within the EPO.

And my partner was not alone with his complaint. Meanwhile we have changed our standard procedure and no longer recommend our clients such a waiver, since we cannot rely on the EPO any more.

When preparing for this blog, I thought I might as well do a short poll among my colleagues to see whether my own rating is an outlier or well within the mainstream. So I asked them two quick questions: (i) How do you rate the overall quality of EPO work products on a scale from 1 (very good) to 6 (unsatisfactory)? (ii) In your opinion, has this quality improved, deteriorated or stayed about the same in the last 2-3 years? I received more than 50 answers. The average rating on this scale was 3.2, with a few “2”s, many “3”s, a couple of “4”s and one “5”. On question (ii), two thirds responded that quality has decreased over the last 2-3 years, one third responded that quality has stayed about the same, and exactly none (zero) responded that there was an improvement in quality. This result alone should give the EPO management and its supervisors, the Administrative Council, something to think about.

Some of my colleagues provided further comments, so I thought I might share them with you and, hopefully, the EPO in order to give them an idea where a bunch of people are happy and where there might still be room for improvement:

– Some of the “products” have clearly unjustified objections but most are good.

– I would say the decrease in quality was slight. Where I have seen changes: Examiners are granting praps sooner, but are making more amendments themselves in the R71(3). I also have the impression that an Exam div will issue a summons, but then avoid oral procs if they can. In terms of formalities, recently I have had a couple of slip ups.

– Quality has (very evidently) decreased: some patent applications are granted, although I myself believe that the claims on file do not merit a grant at all …, more and more simple but serious mistakes, both in examination and opposition proceedings

– Clearly decreased (4-)! Number of R 71(3) communications is steadily increasing, at the same time EESR and Rule 70 as well as Art. 94(3) communications only deal with few specific issues (like inventive step), other issues that should be discussed are simply left for a later point in time! Clear impression of more time-pressure that is being compensated by a decrease of quality!

– I see the full breadth of quality variations, would tend towards 3-4 (for Opposition Divisions rather 4; perhaps this is because I have more hearings in The Hague than in Munich:-)

– Stayed about the same? Yes – the average is pretty good, but still quite a lot of variation (both examination and opposition). Some examiners are very reasonable, others are unwilling to engage with our submissions.

– Obsession with “clarity”, Art 84, is getting more and more irritating. They object under Art 84 just for the sake of it. Any excuse will do. Deferring to their objections does not improve “clarity”. Often, it prejudices it. Often, I have to ask them to withdraw their silly insertions in their Druckexemplar.

– One has the feeling that, for Examiners, the days of making an adult, intelligent, overall assessment of patentability takes second place to virtue signalling, to the EPO’s in-house Quality Police, using Art 84 EPC, that I’m meeting the quality metrics imposed on me.

– The examiners seem to be under a lot of pressure and do strange things. Example: Examiner calls me indicating that the claims are grantable, but the adaptation of the description is missing. He then sends me a draft adapted description by email, asking for a quick reply. The whole thing had to be done twice since his/her adaptation was quite sloppy.

– The worst point is the 71(3) quality. Nearly all texts are faulty. I have complained to several examiners about this. They know about this problem but allegedly can’t change it (it’s allegedly a software problem) and encouraged me to complain officially.

– Also noticed more errors in decisions, such as missing pages etc.

– Generally it seems to me that more and more tasks are pushed onto the applicant (or rather the representative) while the EPO still levies the same fees. For example the EPO online submission of oppositions requires a detailed submission of information for each cited document, so that it can readily be moved onto the register without any input from the EPO. Another example, due to the errors in the 71(3) I have now been asked by examiners to not only file the amended pages but the entire spec, so that they do not have to do any work anymore putting it together or checking correctness.

– The examiners are forced to summon very early in prosecution (to a degree where even they consider it “ridiculous” as one examiner put it). Various examiners are helpful to telephone instead of issuing an examination report, but this cannot be the solution to this inflexibility.

– The quality of first instance decisions are extremely variable (but I think there are plans to create a pool of examiners specialised on opposition proceedings).

– There are of course still good examiners, ODs and BoAs, but it seems to me that they are put under more and more time-pressure to deliver, and this (at least in my view) clearly takes its toll on quality.

– A positive point: the ticket system for enquiries. Generally you get an answer within 1 or 2 days!!

– Decreased – primarily caused by over reliance on IT systems which cannot “think”.

– Everything is relative – compared to the US, I think they are better, but this doesn’t help your survey. (3 – a bit better than satisfactory)

– As to direction, I say it has decreased: I see less and less situation-specific reasoning and engagement with written arguments and more “standard sentences” to raise objections / maintain objections.

– I reserve particular criticism for the current quality of the Annex to the Summons in Opposition which has in my view overall degraded, the workload to prepare for OPs, the number of auxiliary requests which are prudent and thus the costs for the party having increased because one simply has no idea of the direction in which the OD is thinking, so one must assume that even crazy points run by the other side will gain traction!

Several of these views were also earlier held by “Anonymous Attorney” in response to one of my earlier posts. Let me take this opportunity to thank all responders for your thoughtful (or at least thought-provoking) opinionated comments.

If I may summarize in one word the main problem that I and many others are perceiving at present, it would be superficiality; sometimes even sloppiness. My perception is that examiners are meanwhile being exposed to so much “production pressure” that at least a part of them is no longer examining applications thoroughly. The processes revolving around the communication pursuant Rule 71(3) EPC may deserve special attention in this regard.

4 Quality – Policy Recommendations

In view of the above, I would argue that both the EPO’s and the Applicants’ long-term interests would be served better if the EPO management were to reverse its misplaced “production policy” and allow examiners sufficient time for a thorough search and examination, as well as for professional training. I also think that many of the recent HR measures are not going in the right direction. For example, the EPO used to have small directorates headed by one director who oversaw about 20-30 examiners. One of the jobs of this director was to review the products of his/her examiners and provide feedback on quality. This was a manageable task. However, this structure has changed recently. I was told that, right now, one director oversees more than 100 examiners. Instead, smaller groups of examiners are now led by a temporary “team leader” who must spend 80% of his/her time to examine his/her own cases and thus may feel little inclination to thoroughly look after quality in his/her team’s work.

The EPO’s latest idea to generate “more flexibility” by employing more examiners on the basis of 5-year contracts rather than permanently is also completely counter-productive to quality and should be firmly rejected by the Administrative Council. There may be certain positions for which temporary contracts may make sense, e.g. in the case of a well-defined, time-limited project where little time is needed for staff to get up to speed. But the job of a patent examiner is the exact opposite. It takes a good while to get familiar with the EPC and its by-laws (for European Patent Attorneys a three-year practical education is thought indispensable), and the project of examination is never finished. Moreover, what should an examiner do after his/her five-year period? If you are an HR senior manager or an IT specialist, you may perhaps find other employers easily, but there is no real market for patent examiners. And I will refrain from commenting on the proposal, from whomever it may have originated, to change Art 53 of the Service Regulations so that employees could be sacked at any time if “the Organisation” no longer needs them. Fortunately, the EPO has a social and magnanimous President who takes the well-being of his staff to heart and has prevented such a nonsense, as he himself helpfully explained in his internal Communiqué 3/2018:

Moreover, during the December AC meeting, one delegation, supported by others, proposed to introduce in the Service Regulations the possibility of separation with an employee (appointed on fixed- term or permanent basis) at any-time and on the grounds that the post would be suppressed or the staffing level decreased (new Art 53(1) f) ServRegs). It was inspired by the model of open-ended contracts of the private sector and also implemented in some national patent offices or International Organisations, like WIPO.
One could question the real need to implement such a measure for an office that performs as we do. In any event, I understood that it caused of great deal of concerns among the staff and this is why I have convinced the B28 during its second meeting on Wednesday this week to abandon it. Therefore it will not appear anymore in the revised version of CA/3/18 which will be presented for the GCC consultation at end of February and to the Administrative Council in March for decision.

In regard to professional training, I understand that a lot of that is now done by webinars. A wonderful invention if you have staff that has the time to watch and digest them carefully. But if staff is under immense production (increase) pressure, it is predictable that the first thing they will do is to simply click on the webinar slides from time to time while using 90% of their brain capacity for simultaneously working on their cases.

Another myth is the quality allegedly provided by the fact that patents are granted by a three-membered Examining Division, rather than by one examiner. It is true that a decision to grant must be signed by three examiners, but if time is of the essence, why should the fellow examiners spend much time in correcting the first examiner who must dispose of his/her case quickly in order to make the points necessary to reach his/her annual target? One of my readers even went as far to write that

There are clear oral instructions given by some directors: if the first member decides to grant, the two other have to shut up and sign. With the premium system introduced, the examiners will not annoy each other. Another stupidity.

Thus, in summary, it seems to me that a sustainable long-term policy to safeguard quality includes allowing motivated staff sufficient time to do their job properly. Increasing production pressure year-by-year does exactly the opposite and thus will not be a sustainable and sensible policy in view of the EPO’s own mission and vision.