I have started more than twenty online businesses since 2013. Most of them are dead. A few of them paid for my life in Bali, and one of them I sold.
The single biggest factor in which ones lived and which ones died was not the traffic, the ad budget, or how hard I worked. It was whether the business model actually fit the person running it. Me.
That sounds soft. It is not. A business model is a machine with specific inputs. Some models eat money. Some eat time. Some eat your patience for six months before they pay you anything. If you pick a model whose inputs you cannot supply, you will fail at it no matter how good the tactics are, and you will blame yourself instead of blaming the fit.
So this is the article I wish somebody had handed me in 2013. Five online business models I have personally operated, what each one actually costs, what each one demands from you, what nobody mentions in the YouTube thumbnails, and a way to pick one without guessing.
No hype. No “best business to start” ranking. The honest version.
What a Business Model Actually Is
A business model is three answers: who pays you, what they get, and how the thing reaches them.
That is it. Everything else is tactics.
People confuse a model with a channel all the time. “TikTok shop” is not a business model, it is a channel. “Dropshipping” is a fulfillment method that shows up inside several different models. When you conflate the two you end up chasing whatever channel is loud this year and rebuilding your whole business every eighteen months.
Pick the model first. Then pick the channel. The model determines your economics, your daily work, and your ceiling. The channel is just how customers find you, and channels come and go.
One more thing worth saying up front: the market is not the problem. E-commerce was 16.9 percent of total US retail sales in the first quarter of 2026 according to the Census Bureau’s quarterly report. There is plenty of money moving online. The question is never whether the money exists. It is whether you built a machine you can actually keep feeding.
The Five Models I Have Actually Run
I am only writing about models I have operated with my own money. I have no opinion worth reading on Amazon FBA or SaaS because I have not run them.
1. High-Ticket Dropshipping
You build a niche store selling expensive products, usually $500 to $10,000 each, from US manufacturers who ship direct to your customer. You never touch inventory. You make 20 to 30 percent gross margin and something like 7 to 10 percent net after ads and fees.
What it demands: patience with suppliers and a willingness to pick up the phone. Getting approved as an authorized dealer takes weeks and a lot of rejection. Once you are in, you have a moat, because the next guy has to go through the same wall.
What it costs to start: realistically $2,000 to $5,000. That is a store, a business entity, and enough ad budget to get real data instead of noise.
Time to first dollar: 60 to 120 days if you are moving fast.
What nobody tells you: the sales are made on the phone. High-ticket buyers call before they spend four thousand dollars, and if you will not answer, you will not close. If you hate talking to people, this model will punish you quietly for years.
This is the model I chose, and I go deep on why in the full breakdown of how the model actually works. It runs on Shopify for me and always has, mostly because the app ecosystem and the checkout are worth more than the monthly fee.
If you want the exhaustive tactical version of this model, that is what I built Ecommerce Paradise for. The supplier side of it, which is the part most people underestimate, lives over at Supplier Paradise.
2. Low-Ticket Dropshipping and Print on Demand
Cheap products, usually $15 to $60, sold on volume, usually with paid social. Print on demand is the same shape with a design layer on top.
What it demands: creative volume and a strong stomach for ad spend. You are not really in the product business, you are in the media buying business.
What it costs to start: $500 to $1,500, but the real cost is the ad budget you burn learning. Budget more than you think.
Time to first dollar: fast. Sometimes days. That is the trap.
What nobody tells you: the fast first dollar is why people get stuck here. You make $40 profit in week one and think you cracked it. Then you scale and discover your margin cannot absorb a bad ad week. Supplier tools like Spocket make the sourcing easy and the economics still do not work for most people. I ran this model, made money, and quit it, which I wrote about in the business models I quit and why.
If you are going to do it anyway, at least do it with a real supplier feed. Inventory Source handles the sync so you are not manually updating stock counts at midnight.
3. Affiliate and Niche Content Sites
You build a site that answers questions, and you earn a commission when readers buy the tools or products you recommend. This site is partly that. So are most of the sites on my businesses page.
What it demands: the ability to keep publishing when nothing is happening. Months four through nine are the killer. You are writing into a void and the analytics look identical to the day you started.
What it costs to start: under $200. A domain from Namecheap, hosting, and a keyword tool. This is the cheapest model on this list by a wide margin.
Time to first dollar: six to twelve months. Sometimes longer. Anyone telling you ninety days is selling something.
What nobody tells you: it is not passive. It is deferred. You do the work now and get paid later, which is a completely different thing from doing no work.
I run my sites on WPX Hosting because slow hosting quietly costs you rankings and most people never trace the problem back to the host. Before I write a single word I check the term in KWFinder, because writing an article nobody is searching for is the most common way people waste a year on this model.
The full playbook for this one is at SEO Paradise. My own process, including the parts I still get wrong, is at the SEO pillar here.
4. Info Products, Courses, and Coaching
You package what you know and sell access to it. Courses, memberships, coaching calls.
What it demands: an audience, and a track record you can point at. This is the model everyone wants to skip to and almost nobody can start with, because you cannot teach a thing you have not done.
What it costs to start: almost nothing in tooling. Teachable will host a course for a small monthly fee. The real cost is the two to five years of operating experience that makes the course worth buying.
Time to first dollar: immediate if you have an audience, never if you do not.
What nobody tells you: the margin is incredible and the churn is brutal. You are always feeding the top of the funnel. Every month starts at zero. I have run Ecommerce Paradise since 2015 and it still works exactly like that.
You also need an email list, not a follower count. I use Kit for that. Social platforms rent you your audience. Email is the only list you own, and that distinction becomes very real the first time a platform changes its algorithm on you.
I broke down the whole process of turning experience into something sellable in how to turn what you already do into a paid offer.
5. Done-for-You Services and Agency Work
You do the work for other people. Store builds, ads management, SEO, content.
What it demands: the ability to sell and the ability to deliver, which are two different skills that rarely live in the same person naturally.
What it costs to start: effectively zero. You need a way to take a call and a way to take a payment.
Time to first dollar: fastest of any model on this list. Weeks, not months.
What nobody tells you: it is a job until you build a team, and building a team is its own separate skill. I hire through OnlineJobs.ph and I did not get good at it until I had written real SOPs for everything. Before that, every client was a fire I personally put out.
The upside is that services fund everything else. My agency work paid for the niche sites while they were in the dead zone. That is not a small thing.
The Four Constraints That Actually Pick For You
Here is the part that matters more than the model descriptions.
You do not really choose a business model. Your constraints choose it and you either agree with them or you spend two years arguing. Four constraints do almost all the work.
Capital
If you have under $1,000 to your name, high-ticket dropshipping is not available to you yet. Not because it is bad, but because you cannot buy enough ad data to learn anything, and you will conclude the model does not work when what actually happened is you ran out of runway. Under $1,000, you are looking at services or content. Both are close to free to start.
Time
Not hours per week. Hours per week that are uninterrupted. Ten focused hours beat thirty fragmented ones for anything involving research or writing. If your available time is chopped into twenty minute pieces around a job and kids, pick a model that tolerates that. Content tolerates it. Managing client ad accounts does not.
Temperament
This is the one people skip and it is the one that kills them.
Do you get energy from talking to strangers or does it drain you? High-ticket and services both require sales conversations. Content and low-ticket do not. Can you work on something for six months with no feedback? Content requires exactly that. Ads give you feedback daily.
Be honest here. Not aspirationally honest. Actually honest about who you are right now.
Timezone and Location
I live in Bali, which is fifteen hours ahead of Los Angeles. That single fact eliminated some models for me and made others easier.
Phone sales to US customers from this timezone means working evenings, and I decided I would rather do that than move back. Content and SEO do not care where you are at all. If you are planning to be location independent, weight that heavily, and if you want the practical side of running a US business from overseas, that is what Expat Paradise covers.
What Each Model Costs to Start, Honestly
Every model on this list needs the same three things before it needs anything model-specific: a legal entity, a way to get paid, and a place to keep the books.
For the entity, I use an LLC. Bizee is the cheapest way to file one that I still trust, and I have used them repeatedly across several businesses. If you care about privacy and want your home address kept off public filings, Northwest Registered Agent is the better call and worth the extra money.
Before you file anything, though, understand what you are actually choosing between. The SBA has a plain explanation of the structure options that takes ten minutes to read and will save you a conversation with an accountant later. I am not a lawyer or a CPA, and neither is anyone else on YouTube telling you which state to form in.
For getting paid across borders, Wise has saved me a genuinely stupid amount of money in bank fees over the years. For the books, FreshBooks is fine and cheap for a solo operator.
After that, here is the rough model-specific spend:
- High-ticket dropshipping: $2,000 to $5,000. Store platform, a decent theme like Superstore, and real ad budget.
- Low-ticket and print on demand: $500 to $1,500 in tooling, plus whatever you are willing to lose learning to buy ads.
- Affiliate and content: $150 to $400 for the first year. Domain, hosting, one keyword tool.
- Info products: under $100 in software. Years of experience not included.
- Services: effectively zero. Your time is the input.
I broke the numbers down line by line in what it really costs to start each model.
One note on tooling generally: start with less than you think. I have watched people spend $400 a month on software for a business making $0. The tool stack should grow behind the revenue, not ahead of it. The one exception I make now is AI, because Claude genuinely replaces a chunk of what I used to pay contractors for. How I actually use it day to day, and where it still falls over, is on the AI pillar.
The Mistakes I Made Picking Wrong
My first store in 2013 was a general store. Everything for everybody. It did not sell because nobody trusts a website that sells phone cases and patio furniture. I did not understand yet that specialization is what creates trust, and trust is what creates conversion. That took me almost two years to learn and it is why I now say go deep before you go wide, every single time.
The second mistake was picking a model that fought my personality. I spent a year on a low-ticket store because the barrier was low, and I hated every day of it. The daily work was creative testing and I do not enjoy creative testing. The model was fine. The fit was wrong.
The third one, and this is the expensive one, was running five things at once before any of them worked. Five half-built businesses generate the same revenue as zero businesses and roughly triple the stress. I now run one thing until it is boring and systemized, and only then start the next.
The 30-Day Test Before You Commit
Do not pick a model off an article. Not even this one. Test it cheaply first.
Here is the test I use now, compressed:
Week one: pick one model and one specific niche inside it. Not “home goods.” Something like “commercial grade saunas.” Sort competitor products by price high to low and see whether real money is actually changing hands at the top end.
Week two: do the least glamorous version of the daily work. If it is high-ticket, email fifteen suppliers and see how you feel about the rejections. If it is content, write three full articles and notice whether you dread the fourth.
Week three: try to get one signal of demand. One supplier approval. One customer inquiry. One page that ranks for anything. One paying client.
Week four: look at your own behavior, not your results. Did you avoid the work? Which part? That avoidance is data, and it is more predictive than any market research you will do.
The full version of this is in how I test a business model in 30 days before committing.
What I Would Pick If I Started Over Today
Two things, in this order.
First, a service. Something I can sell this month with skills I already have. Store builds, ads, writing, whatever. It is unglamorous and it is a job, but it produces cash immediately and cash buys you the time to build the thing that compounds.
Second, once that service covers my expenses, one high-ticket store in a niche I can stand for five years, with a content layer underneath it from day one. Not a store and then SEO later. Both from the start, because the content takes a year to work and you want that year running in the background while you sell.
That is the combination I would run. Cash flow from services, compounding from the store, and the content quietly building an asset the whole time.
What I would not do is what I actually did, which was start five things and hope one of them worked. It worked eventually. It just cost me about three extra years.
Where to Go From Here
If you got something out of this, the rest of the pillar goes deeper on every piece of it. Start with the Building Online Businesses hub if you want the whole map. If you already know which model you want, skip ahead to picking a niche you can live with for years, which is the decision that actually determines whether the next two years are pleasant or miserable.
Every tool I mentioned here, plus the rest of what I actually pay for, is listed on my resources page with notes on what each one is genuinely good and bad at.
If you have decided on high-ticket dropshipping specifically, everything tactical I know about it is at Ecommerce Paradise, including the done-for-you build if you would rather not do the first ninety days yourself.
And if you would rather just talk it through with somebody who has run all five of these models and made most of the available mistakes, I do one-on-one calls. Bring your actual situation, your actual capital, and your actual available hours, and we will figure out which model fits instead of guessing.
One last thing. Pick something and give it a real year. The model matters less than the fact that you stopped switching.
Get The Operator’s Notebook
One email when I publish something worth your time. What I am building, what the numbers look like, and what I got wrong that month. No pitch, no daily spam.
About Trevor Fenner
Founder of Ecommerce Paradise, operator of Electric Bikes Paradise, and publisher of Paradise Skate Mag. I’ve been building location-independent ecommerce businesses since 2013 — and living the life they pay for in Bali, Indonesia with my wife Widi.
Want to Build Your Own Location-Independent Business?
Get unstuck with a private 1-on-1 consulting call — or keep reading the honest lessons I’ve picked up building businesses from Bali.
Trevor Fenner is a Seattle-born entrepreneur, skateboarder, and expat who left Los Angeles in 2016 to build a location-independent life in Southeast Asia. After living in Chiang Mai and Bangkok, he settled in Bali in 2019, where he has been based ever since. He is the founder of Ecommerce Paradise, an education and services platform helping entrepreneurs build high-ticket dropshipping businesses, and operates Electric Bikes Paradise, an ecommerce store specializing in electric bikes, scooters, and mobility equipment. He also runs Paradise Skate Mag, a skate media project documenting the Bali skate scene and broader skate culture, and is building Bali Cat Paradise, a blog centered on the nearly twenty cats he and his wife care for at their home in Bali. Trevor writes about ecommerce and entrepreneurship, expat life in Southeast Asia, and the lessons skateboarding has taught him about business and life.