No, my site did not get overrun by the pharmaceutical pushers that like to submit "advice" in the comments section. The title is referring to entity size (i.e., undiscounted, small, or micro). This posts will compare the three entity types on three high-level prosecution metrics: allowance rate, actions per allowance, actions per abandonment.

The Inspiration

During beta testing of the BigPatentData Portfolio Visualizer I have generally been analyzing portfolios of large entities (as they are typically the ones with portfolios too large for a simple manual assessment). In looking at a number of these, one thing jumped out at me: allowance rates almost universally near 100%. Here's one example (clicking the image will bring up the live visualization; (free) registration required):

See those allowance rates? The lowest is 97% (861 allowances vs. 30 abandonments). So, since I have been primarily analyzing portfolios of large (i.e., "undiscounted") entities, I started to wonder if this was a function of money that the applicant is willing to throw at an application. That would agree with my own experience which is that, if the applicant just wants a patent -- any patent -- its usually possible with enough bites at the apple. So, I decided to compare allowance rates, actions per allowance, and abandonments per allowance for the three entity types.

Allowance Rate by Applicant Size

So first I plotted allowance rate and was a bit shocked to see this:

Woah. It appears micro entities got absolutely crushed by starting in 2014. Correlation is not causation, but what could it be other than Alice?¹ Now, here again is where I am calling on my readers (there may be three of you now?) to fact check² and help me. There are surely many factors at play here. For example, I suspect micro entities are much more likely to be pro se, so that might account for a lot of this (I intend to explore that in my next post -- subscribe!). But what other factors should I look into?

Mo' Money Mo' Patents?

Now, I don't think we needed this blog post to know that ability to spend more money improves your odds at the patent office, but to try and quantify it, I used actions per allowance and actions per abandonment as a rough proxy for dollars spent per application. In the first chart below we see that micro entities are almost half an action faster to allowance than undiscounted entities. There are a lot of factors at play here, but it seems unlikely that smaller entities have better inventions and/or counsel. It seems more likely that bigger entities just fight longer and harder (i.e., spend more).

I think actions per abandonment is a more direct measurement of willingness/ability to spend more. And there the disparity among the entities is even greater -- in 2018, undiscounted entities pursued doomed applications for an additional ~1.5 office actions.

Takeaways

Again, there could be a lot of other factors at play here and so there isn't really any concrete take away other than the mere facts (unless and until disproven by a reader) that

micro entities have the substantially lower allowance rates than small entities, who in turn have lower allowance rates than undiscounted entities.
Micro entities are less willing to suffer through many rounds of prosecution, than small entities who in turn bow out sooner than undiscounted entities.

Wait...what's up with those micro entity dispositions pre AIA!?

The astute reader will have noticed some micro entity allowances and abandoments in years predating the AIA. I was curious about that too. For the allowances, most of them subsequently certified as micro entities for the purposes of maintenance fees. The handful of abandonments have petitions to revive pending (so arguably I should change those from abandoned to pending). 1One possible reason, mentioned in this older article from Kate Gaudry (follow her, if you are into prosecution statistics), is that smaller entities are more likely to file on business methods 2I spot checked a bunch of applications and they all agreed with my categorization of allowed or abandonment (again, using the BigPatentData definition of allowance and abandonment, discussed here). 3I suppose its possible larger entities have negotiated enough of a discount with their firms to offset the cost of an additional 0.5 office actions per allowance.