When does a company need research and development?

A company needs research and development when a job is stuck and the answer is not already on a ticket. A process that takes hours when it should take minutes. Manual work everyone repeats. Systems that do not talk, so someone retypes the truth. Everyone can see it. Nobody questions it, because that is how the work has always been done.

You do not need R&D because a competitor opened a lab, a vendor said “AI,” or the about page felt thin. Those are marketing problems. Calling them research flatters the spend.

You probably do not need it if

You can specify the work and be unsurprised by the result. You are buying a tool already proven in a company like yours. You want a brand refresh, a brochure site, or a chatbot on empty FAQs. Hire delivery. Do not build a department around a purchase.

You probably do if

The workaround has become the real process, and you still cannot say why the official one fails. Physics, a regulated step, or a model might apply — and you might be wrong. You can protect a small block of time next to the work. You will write a kill rule on day one.

Size is not the test

A twenty-person company can run R&D with two or three people and one bet. A thousand-person company can fail it with a campus and no unknown. How to build an R&D team is the next page if the job is real.

This week

1. 01

   Write the stuck job

   Hours, workaround, who does it. If you cannot, you do not need a lab. You need a walk.

2. 02

   Say whether the fix is known

   Known fix: hire engineers. Unknown: one bet, a kill rule, two or three people.

3. 03

   Ignore the competitor’s campus

   Their press release is not your unknown.

Questions

When does a company need research and development?

When the same job keeps costing hours, the workaround has become normal, and you cannot honestly say you already know the fix. If a competent team could specify the work on Monday and be unsurprised by Friday, hire engineers. Do not call it R&D.