“Honestly, there were hardly any companies offering the same service. Cost, wanting to stop the opportunity loss from leads slipping through, and the timing of needing to explore something new all came together, and we decided on Meeton ai. It has led to 2 deals in 3 months, and our number of proposals has grown. Even when it does not become a meeting, we can instantly see which companies are viewing our content. Even the quality of our market and competitor analysis has gone up.”
— Shota Ueda, Univis Group
Univis Group is a full-service professional firm that provides, end to end and under one roof, areas that are normally split across separate firms: accounting, M&A, management consulting, tax, legal and HR. It has built up more than 1,000 advisory clients, more than 500 M&A engagements and more than 50 IPO support engagements, with offices in Tokyo, Osaka, Fukuoka, Sapporo and Nagoya as well as New York and Malaysia.
M&A consultants, management consultants, certified public accountants, tax accountants, lawyers, labor and social security attorneys — the ability to staff specialists who would normally sit in silos across a single engagement is Univis's strength. For that same reason, customer inquiries range from accounting to M&A to organizational issues, and both the level of interest at the door and the expertise required differ widely.
There was a structural problem particular to a multi-service business. Because the service categories are so far apart, when an inquiry comes in through the website it is hard to sort on the front end, in the moment, whether it is about accounting, about M&A, or in the organizational consulting area.
An inquiry could be about any of the services we provide, and the service categories are quite far apart. It is hard even for us to judge that properly.
Visitors who use chat or an inquiry form come in even at a stage when their own problem is not yet defined. It might be the early stage of considering M&A, or it might be a question about an organizational issue. Continuing to sort, by hand alone, questions that the person asking has not even organized was not realistic.
In the sales flow, the team first sent a follow-up email and then filtered: only customers who replied got a next action such as a phone call. The reason was simple. A certain number of customers download a document and "honestly are probably not that interested," so calling right away wastes resources.
Under this approach, however, the enthusiasm of customers who really were keen cools while waiting for an email reply. There was a risk of a delayed first response, stuck in a state of "not knowing how far to push."
As sales channels, a combination of list-based outreach and automated form-submission tools was producing a certain volume. But on its own, the ceiling on growth was in sight. From the standpoint of business responsibility at Univis, the expectation to keep testing new acquisition channels regularly coincided with the evaluation of Meeton ai.
Unlike a typical purchase evaluation, no real competitive comparison took place when Meeton ai was adopted. The reason is that, at the time, there were almost no players in the "AI SDR" category, which interprets visitor intent with AI and completes conversation, booking and follow-up in real time. There were several AI chatbot and single-function vendors, but no product in sight that could handle inquiries across a multi-service business like Univis.
Honestly, there were hardly any companies offering exactly the same service. As for whether we evaluated anything else in the same category, we did not.
The deciding factors Ueda named overlapped.
Cost was part of it, of course, but simply put, I figured we probably were letting leads slip through, and I wanted to prevent that opportunity loss. In my position it was also a time when we had to keep exploring new things, so we went ahead with the decision.
In the three months since adoption, Univis's chat-originated leads have accumulated steadily, continuing to capture a segment that could not be picked up when the firm operated with forms alone. In terms of deals won, 2 closings were made in 3 months, centered on the M&A consulting area. The number of proposals (proposals made by the firm to prospects) has also grown to about 2x compared with before adoption.
The numbers have definitely gone up by a certain amount. Some have turned into won deals on the M&A consulting side, and the number of proposals has actually increased thanks to this. Deals won in three months: about two.
For Univis, the value of Meeton ai is not limited to the number of leads acquired. If anything, the qualitative impact is greater. Even for visits that do not lead to an appointment, it is instantly clear "which company is viewing which of our content." This is becoming the primary data for market and competitor analysis.
Even when it has not led to an appointment, we can instantly see that these are the kinds of companies actually looking. Based on that, it helps to a certain degree with anticipating deals, checking industry trends as a sales team, and market and competitor analysis.
For a professional firm, grasping industry trends early ties directly to win probability. Being able to detect, at the stage of website visits, "which companies in which industries are now starting to think seriously about reviewing their accounting, M&A, or reorganization" has a strategic value greater than the numbers.
Another qualitative change is lead source visibility. In the Meeton ai admin screen, "form-sourced (reached a form after engaging with Meeton ai)" and "Meeton ai-sourced (converted inside the chat)" can be displayed separately. Contacts who came through the website without engaging with Meeton ai are also shown on the same screen, so the contribution of each initiative can be compared.
Since adopting Meeton ai, there is a phrase Ueda has started to use: "even when it has not led to an appointment." In the business of M&A and management consulting, a customer who is only at the very start of considering can turn into a qualified opportunity six months or a year later. That is exactly why the meta-information itself — "who is visiting our site right now, and with what area of interest" — raises the precision of lists, proposal materials and sales hypotheses.
Meeton ai is neither merely an acquisition tool nor an SDR replacement; for Univis it has started to function as a "radar on the market."
If a visitor is viewing an accounting page, respond in the context of accounting; if they are viewing M&A-related pages, in the context of M&A. Meeton ai automates this by having AI read the website's content. Even a multi-service business like Univis does not need to write exhaustive scenarios in advance.
As long as the materials we present are properly lined up with the landing page or document the visitor is looking at, everything else is fine for now. A feature where the AI reads the code and adjusts its behavior to the page is just about to be released for us.
Univis's existing operation was the filter model of "send an email first → phone only the customers who reply." Meeton Email takes this over. The AI interprets the pages the lead viewed and the questions they asked in chat, and automatically generates an email fully personalized to each lead. Up to five emails go out at 72-hour intervals, and if one goes unopened, the next is sent with updated context.
For the initial operation, Univis chose "approval mode," in which a person reviews emails written by the AI before they are sent. Through integration with Univis's SMTP server, they reach leads naturally, straight from Ueda's own email address.
What Ueda expects of Meeton ai going forward is simple.
Naturally, what I'd appreciate most is for the number of deals won from here to grow. I hope we can consistently get those qualified leads and have them lead to won deals. Beyond that, if the quality of the extracted intent list keeps improving, it will also improve the quality of our sales lists.
For Univis, Meeton ai has begun to take on the role not of a one-off acquisition tactic, but of "an always-on sales foundation that keeps capturing intent."
In a 30-minute demo, you can concretely confirm how it works for your company.
