Why the most valuable trust signals in B2B are still made in hallways, and why AI is now reading them.
On the second day in San Diego, I got to talk with a project manager at a US healthcare services provider. He told me his team can now translate with LLMs faster than ever. Then he told me the catch: quality assurance has become their biggest bottleneck. Whatever time they gain on translation, they now lose checking it.
I’ve thought about that conversation more than any session I sat in. It was a problem no dashboard would have shown me, stated in the buyer’s own words by someone with every reason to be candid. And it only happened because we were standing in the same hallway. That, more than the badge scans, is what conferences are for.
Why a fractional CMO spends so much time in localization rooms
Full disclosure, because it matters to the argument: I’m CMO of LanguageCheck.ai, an AI-powered translation quality assurance platform built by AQRATE in Bologna. That’s why I was in San Diego, and it’s why I’ll be on stage at LocWorld56 in Vancouver this month.
This isn’t a product post. It’s about what the localization industry has taught me about trust, because it is one of the hardest rooms in B2B to sell AI into. Its buyers have heard every promise about machine translation for twenty years. They grade quality for a living. When a vendor claims something, they check.
That makes it the perfect laboratory for a question every founder I advise is facing: how do you become believable when your buyer is skeptical and their first research step is now a chat window?
What I heard in San Diego: the problem isn’t AI, it’s proof
The conversation on the floor has shifted. A year ago the question was should we use AI? Now it’s how do we prove the output is good enough, to our clients, and to ourselves?
Underneath every version of that question is the same gap. Claims are cheap; evidence is scarce. Everyone has a demo. Very few have a structured, third-party-backed story that a cautious buyer can repeat to their own boss.
This isn’t unique to localization. I see the same gap in medical aesthetics, in B2B services, in every technical category where the buyer is an expert. Localization just makes it impossible to ignore.
Conferences are the evidence layer, in physical form
In my last piece I argued that B2B brands need an evidence layer: structured, third-party proof that buyers and AI systems can both rely on. Conferences are where much of that proof is manufactured.
A speaker slot is a public endorsement by an independent organizer. The agenda page, the speaker bio, the session recap, the association newsletter, the photo a peer posts on LinkedIn — each one is a third-party document that says this person and this company are credible here. And those documents are exactly what answer engines read when someone asks who the serious players are.
This is how I map event work onto the Growth Visibility Framework:
| Stage | The question | What an event contributes |
|---|---|---|
| Understand | Can the market tell what you are? | You hear the category’s real language and test your one-line definition on fifty people in two days |
| Trust | Is there credible evidence? | An independent stage, peer introductions, and visible repeat presence |
| Retrieve | Can people and machines find you? | Speaker pages, agendas, recaps and press that get indexed and cited |
| Choose | Is the next step obvious? | A face-to-face conversation with a clear follow-up already agreed |
Most companies treat a conference as a three-day cost centre that ends at the airport. I treat it as the raw material for six months of retrievable authority.
The playbook: before, during, after
Here’s what I actually do so an event compounds instead of evaporating. Steal any of it.
Before
- Lock the entity definition. Your speaker bio, booth copy, LinkedIn headline, and website should describe the company in the same words. Organizers publish your bio verbatim, so write it as the sentence you want an AI to repeat.
- Publish a point of view first. A piece like this one, live before the event, gives every hallway conversation somewhere to land afterwards.
- Book conversations, not just attendance. The best meetings are set two weeks out, not stumbled into at the coffee station.
During
- Collect the market’s exact words. I keep a running note of the questions buyers ask, phrased the way they phrase them. Those become FAQ entries, headings, and future talks.
- Capture proof in the moment. Photos on stage, a short clip, a peer’s quote with their permission. Evidence fades fast if you don’t record it.
After
- Recap within 72 hours. A structured post that links to the official session page, names the organizer, and states one clear takeaway. It ties you to the event in the record that AI systems read.
- Feed the trade press. Associations and industry publications need content. An insight from your session is often exactly what they’re looking for.
- Follow up with a reason. Not “great to meet you” — a resource, a draft, an answer to the question they actually asked.
None of it is glamorous. All of it is the difference between being seen at an event and being remembered after it, by people and by machines.
Next stop: Vancouver
From 19 to 21 October, I’ll be at LocWorld56 in Vancouver, giving a microtalk called From Fluency to Risk: Rethinking Translation Quality in the Age of AI on October 20th at 4:30 PM.
I won’t spoil it here — it’s built to land in the room. But I’ll say this much: I designed it the way I’d advise any founder to design a talk. One idea people can retell at dinner, and one line they can quote the next morning. A repeatable idea is what travels from the room into recaps, LinkedIn posts, and, eventually, AI answers. Design a talk to be cited, and the event keeps working long after the badge comes off.
I’ll run this exact playbook there, and I’ll publish the recap here the week after.
If you’ll be in Vancouver, whether you build language technology, buy it, or are simply trying to make a technical product believable to a skeptical market, send me a note and let’s find twenty minutes.
If you’re selling something hard to believe
The localization industry taught me something I now bring to every engagement. When buyers are experts and skeptics, attention is not the scarce resource. Credible evidence is. And the brands that build it deliberately: on stages, in the trade press, in structured content that machines can read, become the default answer.
That’s the work I do as a fractional CMO for VC-backed founders and scaleup marketing teams, usually from a desk in Calabria and occasionally from a conference stage on another continent.
If your product is better than your market believes, let’s talk about closing that gap.