How to Run a Business Successfully: Ten Parts, One Constraint

A business runs well when it runs on a written system and not on the owner's instinct. Mine has ten parts, in a fixed order: think, decide, offer and pricing, sales, marketing, content, funnels, hiring, numbers, then tools and AI. Find the one part that caps you today and fix that first.

Instinct does not scale

Most stuck businesses do not have an effort problem or a money problem. They are run on instinct, and instinct does not scale. The owner knows the right price, the right pitch and the right hire, and nobody else does. So every decision waits for the owner, and the business grows only as fast as one person can think.

I am an engineer, so I start every business problem the same way: ask why it breaks. Today WTF runs 260+ gyms. The group has 11 revenue-generating companies and is on a ₹144 Cr annual revenue run-rate (₹12 Cr revenue in September 2026), profitable, with 0% closures since inception. I do not try to run that on instinct. I run it on the ten parts below. The longer version of how I got here is on the about page.

The ten parts, in order

For each part I give you what it has to do, the three things I teach inside it, and one question to ask yourself this week. Write your answers down. A part you cannot answer in writing is a part you are running on instinct.

1. Think. See your business like an owner, not like an employee of it. Find the one constraint that caps your growth. Work on the business, not in it. Set a 12-month number and work backwards. Question: if I doubled my leads tomorrow, what would break first?

2. Decide. Make big calls faster, and make fewer expensive ones. Write the options before you pick one. Test small before you bet big. Review every big call, good or bad. Question: what was my last big decision, and where is it written down with the options I rejected?

3. Offer and pricing. Build an offer people feel stupid saying no to. Sell the outcome, not the access. Price for profit, not for the competition. Stack value so discounts become unnecessary. Question: does my price come from my costs and my customer's result, or from the shop next door? I go deeper on this in the pricing decisions founders get wrong.

4. Sales. Turn interested leads into paying customers every time. A simple script anyone on your team can run. Follow-up that closes without chasing. A close rate you track every week. Question: what was my close rate last week, and who other than me can run the script?

5. Marketing. Get customers from channels you can measure. Know your cost to get one customer. Pick two channels and win them. Kill what does not pay and double what does. Question: what does one new customer cost me, channel by channel?

6. Content. Become the obvious choice in your market before they call you. Hooks that stop the scroll. One idea in many formats. Content that brings leads, not just likes. Question: which piece of content brought an enquiry last month?

7. Funnels and website. A path that takes a stranger to a customer without you in the room. One page, one job, one action. Lead magnets that pre-sell. Measure every step. Question: at which step do most people drop off, and do I know the number?

8. Hiring and team. Build a team that runs the business when you are not there. Hire for roles, not heroes. Write the job before you post it. Train with SOPs, not speeches. Question: which role in my business exists only in one person's head?

9. Numbers. Know exactly where the money comes from and where it leaks. Unit economics on one page. Breakeven per location and per product. A weekly dashboard you read. Question: can I say, without opening a laptop, what one location or one product earns after its own costs?

10. Tools, tech and AI. Use systems and AI to buy back your time, after the basics work. SOPs before software. Automate the boring work and keep a human check. Know where AI actually pays in a small business. Question: which task did I automate before I had written down how a person should do it?

Why the order matters

I spent ten years building AI and machine learning systems before I ran gyms. It taught me to trust data over opinions. It also taught me that tech alone never fixes a business. That is why AI is the last part, not the first. In the 15 days of the Operator Program, the first day is spent finding the constraint and the systems-and-AI day comes near the end. Software multiplies a process. If the process is wrong, you get the wrong thing faster.

There is a simpler way to look at the list. In a vlog I recorded in November 2025 I described how I break down any new line of business: split it into two parts, sales and delivery. First you sell, then you deliver, and you count how many people each side needs. Use that split to read the list: for each part, ask whether it helps you sell or helps you deliver.

The list starts with finding the constraint, and you cannot find it without numbers. In a clip about the gym business, posted in July 2024, I put it this way: if you cannot track it you cannot fix it, and if you cannot fix it you cannot make money. I used the same rule to explain why gyms fail. It applies to a clinic, a salon, a restaurant, an agency or a D2C brand in exactly the same way.

Boring is the point

In a reel I posted in June 2026 I said most businesses do not fail because of competition. They fail because the founder gets distracted. I said I got distracted once too. Every founder starts strong, and a few months later there is a new niche, a new strategy, a new business model. Success in business usually looks boring: the same product, the same process, the same execution, again and again.

So read the ten parts as a list of things to repeat, not a list of things to launch. My short version is on the about page: build boring, print money.

What I'd do

1. Take one sheet of paper. Write the ten parts down the left side.
2. Against each, answer its question in one line, with a number where the question asks for one. If you do not have the number, write that you do not have it. That is an answer.
3. Mark each part green, amber or red. Be harsh. A part that works only when you are in the room is red.
4. Find your constraint: leads, sales, delivery or profit. Then pick the first red part on the list that feeds that constraint. First means highest on the list, because the later parts depend on the earlier ones.
5. Work only on that part for the next four weeks. Write the process, give it to one person other than you, and check one number every week.
6. Review at the end of the month. If the number moved, go to the next red part. If it did not, you fixed the wrong part. Go back to step 4.

Do not start ten projects. One part, one owner, one number, one month.