I’m based in the UAE, and I’ve spent more than a decade helping products grow across borders, in tech, crypto and fintech, selling to businesses and to consumers. The mistake I see most often is treating a new market as a translation job. Translate the site, run the same ads and wait. When it doesn’t work, the team decides the market wasn’t ready. Usually the market was fine. The product was talking to people it had never listened to. Here’s what actually changes when you take a product somewhere new, what shouldn’t and where AI helps without replacing the part that matters.
Language before translation
Customers in each market describe the same problem differently. They use different words, compare you with different alternatives and worry about different things.
“Translation carries your words into their language. Research finds their words.”
This holds even when the language is the same. English in one country isn’t quite English in another, and the same goes for Arabic, French or Spanish across the countries that speak them. The product might be identical everywhere. The reason someone buys it often isn’t.
So research each market on its own. Collect local reviews of your product and of the local alternatives. Read the forums and groups people there use. Look at what local competitors lead with. Check what people search for there, in their own words, rather than translating your keyword list.
Then talk to people, because conversations are where the real language shows up. Write the headline from what you hear, not from your home market’s homepage.
Trust travels slowly
Your product travels faster than your proof. A wall of well-known logos from your home market means less than you’d hope to a buyer who can’t see anyone like them using you. Local proof beats global logos.
Local proof looks like:
- customers in that market, ideally ones your buyer would recognise
- reviews on the platforms people there check
- prices in the local currency, and the payment methods people expect
- support in their language, during their working hours
- for B2B, a local partner or reference where buyers expect one
My rule of thumb is to get three local customers before you spend on ads. Sell to them directly, founder to customer if you can. Give them more attention than they’d get anywhere else, and ask if you can tell their story in return. Their stories become your first local proof, and their words become your first local headlines. Ads can scale trust once you have some. They rarely create it from nothing.
Keep one engine, change the inputs
The opposite mistake is rebuilding everything for every market: a new agency, new tools, new reports and a new process each time. It feels thorough. In practice it makes markets impossible to compare, and it multiplies the work for a team that’s already stretched.
Run the same engine everywhere: research, message, channels, follow-up and one number, reviewed the same way every week. Change what goes in, meaning the customer words, the proof, the channel mix, the offer and the price. Keep how it runs. Give each market an owner who knows it well, but one shared playbook.
This is what lets a small team run several markets at once. When the engine is the same, you can see which market is working and why, and a lesson from one market can reach the next in days rather than months.
B2B and B2C change in different ways
In B2B, trust runs through people. Buyers want references they can call, ideally someone local. Sales cycles stretch while procurement, legal and compliance catch up, and partners or resellers can matter more than they did at home. Local paperwork matters too: contracts, invoices and tax details may need to meet local requirements before anyone can sign. In regulated sectors like fintech and crypto, both of which I’ve worked in, the regulator can decide what you’re allowed to say before marketing gets a vote.
In B2C, the biggest change is usually habit. People use different apps, shop on different marketplaces, pay in different ways and follow different creators. The calendar changes too: the moments when people buy, give gifts and save aren’t the same everywhere. What counts as a fair price or a good offer changes with it.
In both, the first question is the same. Who does this buyer already trust, and how do you earn a place next to them?
Where AI helps, and where it doesn’t
AI has made the desk research for a new market much faster. Use it to:
- read local reviews, forums and competitor sites in languages you don’t speak, and cluster what people say
- compare how local competitors position themselves, what they charge and what they promise
- prepare interview questions for each market, then transcribe and translate the calls
- draft local versions of your message for a native speaker to rewrite
- get a first read of the rules that apply, before you check them with someone qualified
Run the same prompts in every market so the answers can be compared. Same engine, again. And ask the model to show where each claim came from and to mark anything it’s guessing. Local detail is exactly where models fill the gaps with confident guesses.
What AI doesn’t replace is being in the conversation. It can’t tell you how a sentence lands with someone who grew up speaking the language, which local partner people actually trust or what a buyer left unsaid. Fluent machine translation is the risky kind, because it sounds right to you. Have a native speaker approve anything that gets published.
Use AI to walk into those conversations better prepared, not to skip them. Build with the people you’re building for, wherever they are.
A checklist for the next market
- Research the market on its own, starting from local customer words.
- Talk to at least five people there before you write the message.
- Write the headline from what they said. Don’t translate the old one.
- Win three local customers before you buy ads.
- Localise the proof, the price, the payment and the support hours.
- Have a native speaker approve anything that gets published.
- Keep the engine, change the inputs and watch the same one number.
Where to start
If you’re planning a move into a new market, start with the free Scaling worldwide guide on this site. It’s a short version of these lessons that you can share with your team before anyone books a flight.