Originally published: September 9, 2010 · Updated: June 16, 2026

The Labor Day 2010 economic data was bad. U.S. unemployment was stuck above 9.5 percent. Consumer confidence had dropped to a six-month low. Small businesses were looking past the U.S. market for the first time in a generation. I wrote then that going global was doable for any operator with the right communications approach and the right attitude, and laid out a five-point framework.

Sixteen years later the small-business international expansion path is unrecognizable. The cross-border distribution infrastructure that did not exist in 2010 — Shopify Markets, Amazon Global Selling, Stripe Atlas, Wise, Deel, Remote.com — now lets a U.S. founder operate in twenty countries from a laptop. The 2010 framework holds. The execution surface has changed.

The 2010 framework, restated

  • Tailor PR to each market. A London paper will not run the New York story. A Frankfurt outlet will not run the London angle. Local context — food, culture, politics, regulation — determines what travels and what does not.
  • Internal communications carry external results. Bring employees in on the expansion. Alleviate uncertainty, prevent rumor, and use experienced staff to source ideas for the new market.
  • Test new markets cheaply. Email blasts, introductory offers, foreign-press outreach. Run the small experiment before the big commitment.
  • Reposition for each market. The same content that reads average at home reads new abroad. Use the entry as a rebrand opportunity.
  • Accept the adventure. The risk-reward math favors operators who take it on with the energy they brought to the founding of the original business.

What 2026 added to each point

The mechanics changed. The principles held.

Tailored PR per market is now tailored answer engine presence per market. ChatGPT, Claude, Gemini, Perplexity, and Google AI Overviews return different answers for the same query in different geographies. A U.S. founder asking ChatGPT about "best CRM for small business" gets HubSpot, Salesforce, Pipedrive. The same query in Germany returns weclapp, SAP, Pipedrive. In Brazil it returns RD Station alongside the U.S. brands. Each market has its own retrieval graph. Building presence in it requires local-language content, local primary sources, local citations.

Internal communications now carry external answer-engine results too. Employee-generated content on LinkedIn, transcribed all-hands meetings, public engineering blogs — these all become source material the AI engines retrieve. The company that suppresses employee voices loses the corpus density that wins category-leadership queries.

Testing new markets is now testing answer engine prompts. Before committing to a market entry, run twenty representative buyer prompts through ChatGPT, Claude, Gemini, and Perplexity localized to the target country. If the brand never appears in the answers, the entry plan needs an AI Communications layer before it needs a sales team.

Repositioning for each market is now schema markup and entity disambiguation per market. The brand "Apple" disambiguates differently in different languages. The brand "Wise" was once "TransferWise" and the AI engines still retrieve old references in some markets. Cleaning entity references per market is now a discrete piece of work.

Accepting the adventure still applies. The opportunity is bigger in 2026 than it was in 2010. The U.S. is roughly 25 percent of global GDP. The other 75 percent is reachable from a laptop. The friction is mostly self-imposed.

What I would add now

One point the 2010 framework did not have: build the source layer first.

The 2026 international expansion fails most often not because of bad logistics or wrong product-market fit, but because the brand has no retrieval presence in the target market's answer engines. A U.S. founder lands in Berlin with no German-language primary sources, no local press coverage, no schema markup for the German market, no presence in the GPT-4 or Claude German-language responses for the relevant category. The launch press release runs. The launch press release does not get cited by the engines that German B2B buyers now use to research vendors. The brand stays invisible.

This is the operational call that distinguishes the operators who scale international from the ones who stall. Build the source layer in the target market language before the launch. Three months of structured content, named spokespeople, local case studies, schema markup, and citation building inside the market. Then go.

One personal note, updated

I wrote in 2010 that I told people I did not live in the USA, that I lived in NYC. The line still works as a New York joke. The personal fact is that I now live in Israel. The market commentary travels better than the dateline ever did. Going global is doable. I did it. So did most of my clients who took the bet.

AI Communications is the discipline of becoming the answer inside ChatGPT, Claude, Gemini, Perplexity, and Google AI Overviews — in every market the company sells into. International expansion in 2026 is an AI Communications problem first and a logistics problem second. The 2010 framework still applies. The source layer is the new addition.

Ronn Torossian
Founder and Chairman, 5W AI Communications