I ran across another article this weekend touting a region’s innovation as measured by their huge patent per capita number (the number of issued patents divided by the population in the region). I am reminded of the adage that just because something is easy to measure, that doesn’t mean it is a useful metric. Patents are used as a metric of innovation primarily because it is some of the only public innovation data we really have available. There are correlative studies showing some relationship, but there is a big difference between a rough correlation and a conclusion about a region or company’s actual innovative output.
I am reminded of a large company event where the company’s General Counsel was presenting patent filing numbers and making pronouncements about what that meant about the creative output of the company’s R&D efforts (and employee inventors). I was sitting in the back watching. The engineers were agitated, deciding that the GC was clueless or, worse, getting angry and offended about patent numbers being used to judge their contribution to the company in such a public way. The backlash from that meeting in the engineering community was not pleasant. The engineers knew what every in-house patent manager knows. Patent output in any given company is more closely tied to staffing levels, incentives for patents, budgets, and the patent culture in the company than specifically with “innovation”.
Those engineers saw the drop in patent numbers as correlating to increasing product delivery pressure from thinly staffed teams. The innovations were still in the products, but there was no way they were using their evenings and weekends to fill out patent forms. Worse, the patent team had had its budget slashed and could only afford fewer patent filings, so even if engineers submitted ideas, the patent team wasn’t filing on most of them. The invention rejection had created a downward spiral in idea submissions. No one likes rejection.
For a mature technology company, the number of patent filings is really just a choice about resources. Very few have the goal of truly maximizing patent numbers–the system has gotten too expensive with too low return on incremental patent investments. Yes, you need to have a baseline of technically creative people coming up with new stuff to file for patents, but most technology companies have that since it is the lifeblood of their products. Deciding what to actually spend on patents is about allocating money that could otherwise be spent on many other corporate priorities. Increasingly, that money is much better spent keeping dangerous patents out of the hands of patent trolls. That doesn’t make large companies less innovative, it just means a change in their budgeting priorities.
Similar rules apply in universities and other pure research organizations. Every dollar spent on patents is a dollar not spent on research (or a new building). There was a “more is better” movement for a while in some (well-funded) organizations hoping to score big in the licensing lottery while giving leaders an easy success metric to tout. But it wasn’t sustainable. Now, the question is how few patents can translate into the most commercialization deals. Once you can get a private company (licensee) on the hook for the costs, it is great to pump up the patent filings. But until then, who can afford it?
Which brings us to startups and independent inventors. Their biggest reason for filing or not filing is generally money, closely followed by personal or business philosophy on whether patents are a good idea in the first place. Venture funded companies may have the best correlation, just because they are usually encouraged to get a few patents if they can (with the resources to do it), but few investors want a small company spending too much time and attention on patents. But some of the most innovative companies out there, don’t pursue patents because they distract from execution and are viewed as too slow to be relevant. And of the patents that are filed, a lot of the “inventions” coming out of independent inventors, startups, universities, and established companies are worthless–representing no real innovation at all. Some studies put the ratio of worthless patents around 80%–pretty substantial noise for any metric.
Patent numbers are interesting. But like all metrics, you need to really understand the system behind them and be careful about your assumptions. Don’t believe the hype about patents making one region or company more innovative than another without more data.