Newton was live in two markets — Thai and English — at the same time. I was thrilled. My first SaaS was going international. Then, after running it for over a month, I shut the English side down completely and went back to focusing on just one country.

Going International Seemed Like the Obvious Next Step

When Newton started as a personal tool and turned into a real SaaS with paying customers, the thought was instant: "If it sells in Thailand, it should sell internationally too." Bigger market. Higher willingness to pay. More potential.

So I launched Newton EN — an English sales page, Facebook ads running across Vietnam, Indonesia, Malaysia, Singapore, and the Philippines. Budget was $20-30 a day.

I set up a separate control plane, built an auto-provisioning system that handled both TH and EN signups, wired up Stripe payments and a trial flow for both markets, wrote English blog content, set up an automated content pipeline for the EN pages. Everything was ready.

The Reality Was Nothing Like the Plan

The first problem was simple: I don't know the EN market.

For Thailand, I know exactly how people think. I know the language, the pain points, the copy that makes someone click. I understand the context because I live in it.

But solopreneurs in Indonesia or the Philippines? I had no insight at all. I didn't know how they buy SaaS, where they discover AI tools, or whether my copy resonated with them. I was guessing on everything.

The second problem was budget dilution. Running TH ads and EN ads simultaneously meant neither got enough spend to optimize properly. The TH side was performing — but it wasn't getting full budget because half was going to EN campaigns that weren't converting.

The third problem was support. I run this entire business solo. Thai customers message me and I respond instantly — I understand the problem fast, and I've already trained my AI to match the right tone because I know what Thai users expect. EN customers? Different timezone, different language, different context entirely.

The Moment I Decided

One day, a trial user signed up through Newton EN. Stripe charged successfully. The system auto-provisioned their server. The welcome email went out. Everything looked perfect.

Then they tried to use it. Hit an authentication issue. Tried 8 times. Rebooted the server. Still couldn't get past it. Cancelled within 3 hours.

I saw every log. I watched them try. I watched them fail. I watched them leave.

The part that hit me hardest wasn't the cancelled subscription — it was that I couldn't do anything about it in real time. I was deep in Thai-market work when it happened. If I'd been focused on just one market, I might have caught the problem before they gave up.

That night I sat with it for a while. And then I made the call — shut down EN entirely.

What I Actually Shut Down

I didn't delete anything. I hibernated it:

  • Killed all 4 EN ad campaigns — Vietnam, Indonesia, Malaysia-Singapore, Philippines. All off.
  • Stopped all EN content cron jobs — 8 automated posting slots across Facebook and Instagram. All paused.
  • Shut down the EN research pipeline — no more topic generation for EN content.

What I kept running:

  • The EN website stays up — if someone finds it organically, they can still sign up.
  • The control plane still works — existing EN customers still get support.
  • EN blog posts stay indexed — SEO traffic keeps coming. No reason to throw that away.

This was a hibernate, not a kill. If the Thai market gets strong enough and I have the bandwidth, I can flip EN back on. Everything's still there.

Why Focus Changed Everything

The moment I paused EN, the Thai side improved immediately.

Budget went full throttle. TH ads got the entire spend. I could optimize properly instead of splitting between two markets and hoping.

Time consolidated. Support, content, feature development — all of it pointed at one market. I could catch customers about to churn faster because I wasn't context-switching between two different audiences speaking two different languages.

Mental load dropped. This one's bigger than it sounds. Before, I had to think about everything in two versions — TH copy and EN copy, TH landing page and EN landing page, TH support flow and EN support flow. Now I think in one version. Everything gets deeper, better, faster.

4 Lessons From Pulling Back

1. Bigger market doesn't mean better market. The EN market is objectively larger. But if I don't have an edge there — no market insight, no cultural context, no native understanding of the buyer — then the size doesn't help me. A small market I understand deeply beats a huge market I'm blindly guessing in.

2. Solo founders have to choose. I run my business with an AI agent and 34 custom skills — but I still have limited bandwidth. No separate marketing team per country. No support agents per timezone. Resources are finite. They have to go where they'll have the most impact.

3. Pausing isn't failing. I don't see Newton EN as a failure. I see it as "not yet." The market is still there. The system is still built. The website is still live. It's just not the right time.

4. Focus beats spread thin. Every single time. I knew this intellectually. I've read it in every business book. But I didn't feel it until I lived it. Running two markets at half capacity produced worse results than running one market at full capacity. It's not even close.

What's Next — Teaching First, Then Selling

My growth strategy for Newton TH is now course-led growth. Instead of just running ads, I teach — courses, workshops, content that helps people understand what an AI agent actually is and what it can do.

Because from supporting real customers, I've learned that the best customers are the ones who understand the product before they sign up. They use it properly. They don't cancel after a week. They stick around.

I don't need a lot of customers. I need the right ones. People who are ready to pay, ready to learn, and ready to actually put AI to work — not just try it once and move on.

If you're a business owner or solopreneur thinking about getting your own AI agent on a private server, check out Newton. I'm putting everything I have into making it work for one market — and every customer gets my direct attention. See how it works →

— Pond