Divide and Conquer: What a Physics PhD Taught Me About Growing a Local Business

People ask me why a physicist runs a marketing and technology company. The honest answer is that growing a business is a systems problem, and a physics PhD is six years of training in systems: how many small parts interact, where the hidden structure is, and what breaks first when you push.

I studied statistical mechanics and complex network dynamics at the University of Houston, then spent years building machine learning and cloud systems for energy companies. Today I use the same habits with law firms, contractors, studios and other local businesses. Here are the five lessons I use most, and how you can apply them to your own company this week.

1. Find the structure under the noise

My PhD advisor, Dr. Kevin E. Bassler, taught me to look for the fundamental structure beneath a complicated system before trying to change it. In a network, a few connections often decide how everything behaves.

A business has the same kind of structure. Most owners can describe their marketing as a list of activities: a website, some ads, a Google profile, social posts. Very few can draw the path a real customer takes from first search to signed job. That path is the structure. Until you can draw it, every new tool is a guess.

Try this: on one page, write every way a new customer finds you, how they contact you, who answers, and where that conversation is recorded. The gaps usually show up in the first ten minutes.

2. Divide the hard problem into smaller ones

Dr. Royce K. P. Zia, my mentor in complex systems, taught a method he called divide and conquer: turn a problem that looks impossible into smaller, simpler problems, solve them one at a time, then return to the whole.

"We need more customers" is an impossible problem. These are not:

• Are we found? Search, maps, AI assistants, referrals.
• Do people understand what we do within a few seconds of landing on the site?
• Can they reach us easily by phone, form or text, at the hours they actually look?
• Does someone answer quickly, and is the answer helpful?
• Do we follow up with the people who did not book the first time?
• Do we know which source brought the customers who paid?

Each of those has its own fix, its own cost and its own measurement. Solving them one by one is how a small budget beats a large, unfocused one.

3. Quality has a cost, so measure before you spend

Dr. Margaret S. Cheung, one of my research supervisors, would not let convenience set the standard for scientific work. Good results cost patience: one more check, one more run.

In marketing, the cheap shortcut is spending on ads before you can measure what they do. The patient path is to set up tracking first: which calls came from which source, which forms turned into real conversations, which conversations turned into revenue. It feels slower in the first weeks. It is the only way to know whether the money is working.

4. Push the system to its edge to see what fails

In research and in industry, I learned that normal operation hides weak points. You understand a system when you push it to its edge and watch what fails first.

You can do this to your own business in an afternoon:

• Call your main number after hours. What happens?
• Fill in your own website form on a phone. Did anyone get it? How fast?
• Search for your service and your city the way a customer would, on Google and in an AI assistant. Are you there? Is the information right?
• Ask your office manager what customers ask most often that the website does not answer.

The first failure you find is usually worth more than the next new campaign.

5. A model has to survive reality

My mentors in the energy industry taught me that an elegant model is worthless if it ignores the field: the geology, the crew, the budget, the safety rules. The same is true of a marketing plan. It has to fit the people who will run it, the season the business lives in, and the cash it can actually spend.

That is why I prefer to build the system with the client rather than hand over a strategy deck. A plan that the team cannot operate on a busy Tuesday is not a plan.

Five questions to ask this week

1. Can I draw the path from a customer's first search to a signed job?
2. Which single step on that path loses the most people?
3. Do I know which marketing source brought my last ten paying customers?
4. What happens to a call or form that arrives after hours?
5. What would I fix first if I could fix only one thing?

If you want a second pair of eyes on those answers, that is the work we do at Medai Digital. And if you want the personal story behind this way of thinking, it is in my interview series on Royal Windoor.