The practice

The clients run from companies a few people started in a garage to some of the most valuable technology companies in the world, and the chair changes from case to case. Sometimes the client owns the patent and wants it respected. Sometimes the client built a product and is being told it belongs to somebody else.

Those two jobs are not mirror images. The side asserting a patent is trying to hold a line drawn years ago in language written by lawyers. The side defending is looking for the places that line was moved, narrowed, or drawn around something else entirely. Having done both means the argument coming back at you is rarely a surprise.

Courts have their own habits. Local rules, claim construction schedules and the appetite of a particular bench for a particular kind of motion are not details you learn from a treatise. Cases have been argued before Judges Lucy Koh, Rodney Gilstrap, Vince Chhabria, Leonie Brinkema, Jon McCalla, Catherine C. Blake, Mary Pat Thynge, Susan van Keulen, Jan DuBois, Philip Gutierrez and B. Lynn Winmill.

What that is worth to you is narrow but real. It means the first conversation about your case can be about your technology rather than about how federal litigation works.

Before any of the litigation, the job was patent prosecution: sitting with inventors, writing the application, and then arguing with an examiner for a few years until something was allowed. Hundreds of applications were drafted or supervised that way.

That is why the written record is the first thing read in a dispute rather than the last. Every patent carries a public file of the back and forth between the applicant's lawyers and the patent office, called the prosecution history. To get the patent granted, applicants very often narrow what they are claiming. Those narrowings are permanent and public, and years later they limit what the patent can be argued to mean.

Someone who has written those responses knows what the concessions in them look like, and knows which ones were made carelessly at four in the afternoon to get a case off a desk. That is unglamorous reading and a great deal of quiet leverage lives in it.

Where that shows up

• Claim construction, where the meaning of a disputed word is often settled by something the applicant already said.
• Deciding whether a patent is worth attacking at the patent office rather than only in court.
• Testing a client's own patent before it is asserted, so the weakness is found on your side of the table first.

A nine month secondment at T-Mobile, working as in-house intellectual property counsel. The work was portfolio development, managing outside counsel, sitting with inventors, and weighing offensive and defensive strategy against a budget that had a number on it.

An invoice reads differently once you have been the person approving it. So does a memo that answers a question nobody asked. The habit that comes out of that stretch is telling a client what is not worth doing, in the same tone as everything else, before the money is spent rather than after.

Some of them happen at the Patent Trial and Appeal Board, in a proceeding called an inter partes review. It does not ask whether your product infringes. It asks whether the patent should have been granted at all, and it is decided by administrative judges who read patents for a living rather than by a jury.

It runs on a different clock and a different budget than a district court case, and opening one is a strategic decision with consequences in the district court case running alongside it. Multiple inter partes reviews, five of them argued in person, plus covered business method proceedings and ex parte reexaminations.

For a company being sued, the useful question early is not only how to defend the case in front of you. It is whether the patent itself will survive a second look, and what it costs to find out.

Patent appeals go to one court, the U.S. Court of Appeals for the Federal Circuit, which hears them from every district in the country. That includes briefing and standing up for oral argument.

The argument in Buckman Laboratories against Solenis is in the court's public recordings. Anyone can listen to it, which is a fair way to hear how a technical case gets explained under questioning before deciding who should explain yours.

Cases have turned on cryptography, encryption and digital security; two-factor and multi-factor authentication; telecommunications; software; imaging and encoding; chemistry and materials; and blockchain and digital assets.

A lawyer who cannot follow the technology has to hire an expert to understand it, then hire another to explain it, and the client pays for both translations. It also means the argument is only ever as good as the translator, which is a bad position to be in the week before a hearing.

The test worth applying to anyone you interview is simple. Ask them to explain your own product back to you. You will know within a minute.

There is also trademark and copyright work, proceedings before the Trademark Trial and Appeal Board, and intellectual property transactions: the licences, assignments and diligence that decide who owns what before anybody argues about it.

Those matters arrive in the same week as everything else, usually attached to something the business was already trying to do. Keeping them with one person who knows the company saves the second and third explanation of what the company actually sells.