Last month, I came across an article from BT Group’s Peter Shaw, titled, “How the cloud is clearing the headaches of drug development” and was truly marveled by something. The similarities between the development of drugs and software are remarkable.
This isn’t the first time the software development process has been compared to another industry, but in the two most widely known instances (see Eli Goldratt’s The Goal, and Gene Kim’s The Phoenix Project) it was the manufacturing industry that seemed the most similar. The delivery of code between developers and testers, eventually being deployed to production really does resemble the movement of any widget down an industrial assembly line.
Unfortunately, the uncanny and overwhelming similarities between software development and that of pharmaceuticals lie within the ROI-killing constraints around data that these teams each face in getting their products to market.
Let’s use direct quotes from Shaw’s article to prove this point. Try switching out each “drug” and “researcher” for “application” and “tester” and watch what happens.
“Every pharmaceutical company wants to speed up the process of developing a drug and bringing it to market. That way, they have more time to get return on their investment before the patent runs out.”
Okay, that one’s a little easy. All businesses want to get their product to market quickly, but speed is absolutely crucial, often priority number 1, in the software industry.
“There’s confidential information involved, and as a drug gets closer to market there’s intellectual property to protect, and the regulations get tougher. So sharing information at the end of the development chain has always been a big headache for CIOs.”
I love this one, and it’s interesting that in software that private or sensitive information, or a masked/virtualized version of it, isn’t just needed at the end of the SDLC—it’s needed far sooner, so that continuous testing can be performed in order to ensure not just turbo speed, but also outstanding quality.
“In the past, a corporate IT department might have taken months to meet scientists’ requests for multiple servers with the right power, memory, apps and security…Now researchers can test more compounds, and quicker. If they’re going to fail, they want to fail early.”
Failing fast, early, often, cheap, forward, etc., has its fans—and perhaps just as many naysayers—but these mantras inarguably apply best to the software and pharmaceutical worlds. As Shaw points out, by providing scientists with the complete servers they need, they’re given the time they need to find problems, early enough in a drug’s development, so there’s also time to resolve them without impacting release deadlines.
“The collaborative nature of drug development means companies…have to share research and data with external research organisations and academia.”
I have to admit, I had no idea that drug development was the collaborative process that it is. This is software development to a T, and easy and secure collaboration is one of Skytap’s primary and most popular benefits. Agile dev/test teams, especially those that are globally distributed, must have access to the environments they need in order to deliver business needs.
“…Researchers can use virtual machines in places like New Jersey or Greater London and harness huge amounts of computing power, all from a simple device. Like an iPad. And because it’s virtual we can make it secure and share it between organisations.”
This is just great, and Shaw is 100% correct. Drug researchers, just like software developers and testers—and I’m sure many others out there—have to be able to access the data, tools, and environments they need to do their jobs. They won’t always be at their desk, and they’ll often need to be able to share their work with someone who’s not at theirs.
No matter what you’re building, resource constraints better not prevent your teams from delivering your product to your customers before they can get it from your competition.