ZEALS.

Somewhere, a phone buzzes. That is ZEALS now.

Awoman in Osaka clicks a Shiseido ad on Instagram, fully expecting to land on the usual stale product page with a stock photo and a 9% off banner. Instead, a chat opens. The brand asks two questions. She answers. Ninety seconds later she has a recommended serum, a delivery date, and a discount that feels suspiciously specific to her skin type. She does not know it, but a ZEALS conversation flow just did the work of a website, a sales rep, and a CRM in less time than it takes to reheat tea.

This is not a tech demo. This is a Tuesday.

ZEALS, founded in Tokyo in 2014 and now operating out of both Shibuya and San Francisco's Mission Street, has built its business on a single bet that, in 2026, no longer looks eccentric: people would rather chat than click. The company calls the result Chat Commerce. Roughly 400 enterprises - including Toyota, Dyson, JINS, Zenni Optical, and Shiseido - have decided to test that theory with real budgets.

The brand becomes a person you DM. The funnel collapses into a conversation. - The ZEALS pitch, in one breath

The web form was always a small humiliation.

Every marketer alive has stared at the same problem. You spend money to drive someone to your landing page. They scroll. They half-read. They abandon. The industry's response, for fifteen years, has been to optimize the landing page - rounder buttons, bigger hero images, slightly less obnoxious popups. The conversion rate moves a tenth of a percent. Everyone agrees this is progress.

Masa Shimizu, ZEALS's founder, looked at the same problem and decided the landing page was the bug, not the canvas. Shimizu - born in Okayama in 1992, briefly a factory worker, later a Meiji University freshman - started ZEALS in April 2014 with what amounts to a contrarian observation: in Japan, people already spent their day inside one app. It was called LINE. They did not want to leave it. So why was every brand still building experiences somewhere else?

It is the kind of observation that sounds obvious in retrospect, which is usually a sign that almost no one acted on it at the time. ZEALS did, in 2016, by becoming one of the first companies in Japan to build marketing chatbots on LINE at industrial scale.

"Not another chatbot." - The first thing the ZEALS website tells you, slightly defensively. - zeals.ai, homepage

A freshman, a factory floor, and a hunch about hospitality.

Most foundation myths in tech involve a dorm room and a hoodie. The ZEALS one is stranger. Shimizu had worked on a factory line before college. He had seen what real customer service looked like in Japan - the cultural concept called omotenashi, roughly translated as wholehearted hospitality. He wanted to know if you could ship that through software.

ZEALS gave the answer a brand name: Omotenashi DX. The premise is that a digital interaction can still feel attentive if you build it with the same care a tea house would. That sounds like a brochure. The thing is - it shows up in the product. ZEALS does not just sell software. It pairs every enterprise customer with conversation designers - actual humans who write the dialog, study how customers reply, and tune the flows. AI handles the volume. Humans hold the standard.

It is an awkward sales pitch for a SaaS company. It is also why ZEALS works on accounts that pure-software chatbots tend to lose.

Software at scale. Hospitality at retail. Most companies pick one. - Editor's note

What it actually does.

The ZEALS platform sits on top of LINE, Instagram DMs, and Facebook Messenger. A brand runs an ad. Instead of pointing it at a landing page, the ad points into a chat - the company calls these click-to-experience ads. The conversation that follows is partly scripted, partly generative AI, and entirely instrumented. ZEALS captures zero-party data (the stuff customers volunteer), routes it into the brand's Salesforce or Shopify stack, and pushes follow-ups when context warrants it.

Layered on top is Omakase.ai, ZEALS's generative-AI agent product, which turns the chat from a flowchart into something closer to a salesperson with infinite patience and no commission incentive.

What you can do with it

If you sell cars, ZEALS qualifies a test drive. If you sell skincare, it picks a serum. If you sell glasses, it walks customers through frames - Zenni Optical and JINS both use it. If you are a telco, it onboards a customer without a phone call. The pattern is consistent: replace whatever form, page, or call-center step used to come next with a conversation that already knows the answer to its own first question.

The receipts, charted.

It would be tidier if a profile like this could close with a single, satisfying revenue chart. ZEALS, like most private companies, does not hand those out. What it does disclose - and what investors can verify - is heft. The fundraising and the customer roster are not subtle.

Customers are the harder currency. Toyota using ZEALS to qualify leads is a useful sentence. Shiseido using it to recommend serums is useful. Dyson is the kind of brand whose marketing team is constitutionally allergic to looking ordinary - their participation matters. So does JINS, the Japanese eyewear chain with thousands of stores, which has built ZEALS chat flows that quietly absorb work that used to belong to the floor associate.

400 enterprises is not a logo wall. It is an admission that the message bubble is finally infrastructure. - A point ZEALS would prefer to make less directly

Hospitality, productized. For better and worse.

There is a small philosophical hazard in what ZEALS does. If you take a culture of attentive in-person service and automate it, you risk getting a smiling chatbot that has read the etiquette manual and learned nothing. ZEALS is aware of this trap - hence the conversation-designer-as-headcount approach, hence Shimizu's frequent insistence that the software is a delivery mechanism for human craft rather than a replacement for it.

Whether that distinction holds at the scale of 400 enterprises is a fair thing for any skeptic to test. The mission, at least, is unusually specific for a SaaS company: not "win the messaging channel", but "make the messaging channel feel like good service did, before software." It is the kind of mission that produces interesting engineering choices. ZEALS's stack reportedly includes Elixir, Go, Scala, and Spark - a polyglot setup that suggests they treat conversation routing as a real-time data problem, not a CMS problem.

The next storefront is the one already in your pocket.

Look at the trend lines and the bet ZEALS made in 2014 - that messaging would eat the rest of the funnel - looks less like a hunch and more like an actuarial certainty. Instagram now lets brands transact in DMs. WhatsApp has commerce APIs. Apple's Business Messages exists, even if no one quite remembers it does. The infrastructure has caught up to the original premise.

That puts ZEALS in a peculiar position. The category it helped invent - chat commerce - is now noisy, with every CRM vendor announcing a chat feature on a quarterly basis. ZEALS's defense is the boring, unfashionable thing: ten years of conversation design data, a customer list that includes brands too risk-averse to bet on a startup, and a U.S. office that did not exist three years ago. None of that is moat in the textbook sense. All of it is hard to copy in a quarter.

Back to the woman in Osaka. She finishes the chat. She buys the serum. She does not tell her friends about ZEALS, because she did not notice ZEALS - she noticed Shiseido. That, in a category obsessed with brand visibility, is the highest compliment a software vendor can be paid. The infrastructure worked. The brand looked good. The funnel collapsed into a conversation and nobody had to apologize for it.

The best chat commerce is the kind you don't notice you're in. - The case for ZEALS, in one line

A phone buzzes. Somewhere. That is ZEALS, still.

Read, watch, follow