Portfolio of Keith Elder10 Tips for Building Effective Proof of Concepts (POCs) - Lessons from a 20-Year Veteran
by Keith Elder
When it comes to building prototypes or proof of concepts (POCs), experience teaches that the process is just as important as the outcome. After two decades of trial and error, here are the top 10 lessons I’ve learned. These tips can guide anyone in getting the most out of their POC efforts, whether you're an experienced engineer or just getting started.
1. Be Prepared to Throw Away Your Best Code
Some of the best code you’ll ever write during the POC phase will likely never make it to production. The purpose of a POC is to experiment, learn, and iterate—so be ready to toss your work aside. Holding on to the idea that it must be perfect or permanent will hinder your progress.
🧹 “The best code you'll ever write is the one you're brave enough to delete.”
2. Take the Path of Least Resistance
When you're exploring new technologies, don't pick the one that takes a day to set up or is fraught with friction. For example, years ago, my team and I experimented with Riak, a distributed database. After days of troubleshooting, we pivoted to Google’s LevelDB, which we got up and running in minutes. If your chosen tool causes friction early on, it's a red flag for future scalability.
🧭 “Friction is the universe’s way of saying, ‘Try something easier.’”
3. Don’t Shave the Yak
Yak shaving refers to unnecessary tasks you find yourself doing that are tangential to your main goal. Focus on solving the immediate problem rather than creating more work. As the Pragmatic Programmer advises, avoid overcomplicating your solution during the prototyping phase.
🐃 “If you're knee-deep in yak hair, you're probably lost.”
4. Avoid Reinventing the Wheel
If you end up building something from scratch, you probably didn’t research existing solutions thoroughly enough. The odds are that someone has already solved the problem you're tackling. A POC is about learning quickly, not reinventing the wheel.
🛞 “The road to wasted time is paved with perfectly round wheels.”
5. POCs Should Stand Alone
Remember, the primary goal of a POC or prototype is tolearn, not to produce production-ready code. It’s important to ensure the code is functional, but don’t aim to have it integrated directly into your final product.
🧪 “If your prototype needs a babysitter, it’s not a prototype—it’s a problem.”
6. Embrace Writing the Worst Code of Your Life
POCs and prototypes are about rapid learning, not elegant solutions. Enjoy the freedom to code without worrying about best practices like dependency injection or error handling. This is your chance to explore without constraints.
🧠 "Show your mess, not your masterpiece."
7. Don’t Anchor Yourself to Your First Solution
One of the most common mistakes engineers make during POCs is becoming too attached to their first solution. The first attempt is rarely the best or final answer. Be open to alternative approaches and ready to pivot if necessary.
⚓ “Your first idea is just the dock. The real answers are out at sea.”
8. Leave Your Ego at the Door
If your POC isn’t selected, that’s okay. The knowledge gained is more valuable than being right. Be ready to throw away your work without taking it personally. It’s part of the process.
🚪 “POCs don’t care about your pride—they care if it works.”
9. Grow Thick Skin
It's easy to become emotionally invested in a POC, especially after pouring time and energy into it. But remember, a POC is meant to be challenged. Prepare to have your ideas tested, questioned, and sometimes rejected.
🛡️ “If your POC can’t take a hit, it’s not ready for the ring.”
10. Learn to Sell Your POC
Once your POC is working, the next step is selling the idea. Crowdsource feedback, share your progress, and demo it frequently. The more feedback you get, the better your final product will be. A personal example: Nearly 20 years ago, I had to push .NET at a company where it wasn’t used at all. Through consistent POCs, lunch-and-learn sessions, and answering endless questions, I eventually won over the team. Some POCs take time, but persistence and refining your pitch are key to success.
🎤 “A great idea dies quiet. A great demo makes noise.”
Final Thoughts
POCs are critical learning tools, and the process of building them should be just as rewarding as any final product. Remember to stay flexible, leave your ego at the door, and keep the focus on learning and refining. Happy building!