Plenty of ryokan and hotels run guest surveys. Far fewer manage to get those responses in front of the entire team and turn them into better service the very next day.
The responses come in, but tallying them takes time. By the time they become a report, the information is already stale. And improvement work ends up confined to a handful of managers.
To solve exactly that, Hotel Yagi in Awara Onsen, Fukui Prefecture, introduced an AI system for survey analysis and improvement briefings. Every morning, AI analyzes guest feedback and automatically distributes that day's improvement points to every member of staff. The system has since come to shape not only frontline service quality, but management decisions, capital investment, and even community development across the entire hot spring town.
We spoke with Mr. Tsukasa Yagi, President of Hotel Yagi, about the background to the project and everything that has changed since.
Chapter 1 — Where the Guest's Voice Used to Stop
Before AI, how did Hotel Yagi collect guest feedback and share it with the floor? Paper surveys, tallied and entered by the president himself, then analyzed — a workflow that is hardly unusual among ryokan and hotels. But inside that workflow sat a structural problem that made real improvement difficult.
"The daily tallying was my own job"
Before, who received guest feedback, how often, and in what form?
We collected it on paper surveys. The daily tallying and manual data entry were my own job. I would print out a week's worth at a time and hand it around to the staff. That was how information flowed in this hotel.
"By the time it arrived, it was slightly old news"
Did that feedback actually reach the staff on the floor?
I did distribute it, but honestly, not many staff read it properly. On top of the one-week lag, my own work would sometimes fall behind — so by the time it reached the floor, it was slightly old news.
The frustration of "only management knows"
What was frustrating about the state where only management had the information?
The gap in understanding was the biggest problem. Because I was entering the data every single day, I naturally developed a feel for what guests were experiencing. But the staff on the floor never got the context — why a particular improvement mattered. If even one staff member truly understood, the pace of improvement would rise dramatically. That was a period where I felt both that potential and that frustration at the same time. I did base my morning briefings on guest feedback. But the evidence I was speaking from was a week old.
Collecting guest feedback was never the issue. Whether that feedback became shared knowledge across the floor and translated into daily action — there, Mr. Yagi says, the answer was simply "no."
An owner reading surveys every day, and staff who encountered the information once a week. That information gap had a direct effect on the speed of improvement. What finally changed the situation was AI.
Chapter 2 — The Decision: "Every Day, by AI, to Everyone"
Hotel Yagi's decision to adopt the system wasn't only about AI enabling sophisticated analysis. What mattered more was that the information would arrive every day, to everyone, without a human in the middle. That mechanism, he says, felt capable of changing the organization itself.
"It sounded interesting. That was my first impression."
What was your honest first impression of the proposal, and what made you decide to go ahead?
Quite simply, I thought it sounded interesting. The deciding factor was that it would be distributed every day in my place. Having the president drone on every morning is no fun for the staff either. I had almost no hesitation.
Nobody said "I don't want AI telling me how to serve guests"
Was there any anxiety, or resistance from the floor, about everyone reading the hard-to-hear feedback the next morning?
There was almost no anxiety. And no particular resistance from the floor either. Nobody reacted with "I don't want AI telling me how to serve guests." If anything, my instinct was that coming from AI, it would land more easily.
Precisely because it comes from AI, it lands flat
Was there something about AI being the author that made it easier to accept?
In a small organization like ours, when instructions pass from person to person, personal feelings inevitably come along with them. Once it becomes "the president said so" or "that person said so," the content picks up the color of the relationship before anyone even considers what was said, and it gets harder to take at face value. Instructions that arrive from AI don't carry that color. So people can read them flat. The smaller the organization, the bigger this effect is.
Adoption went more smoothly than expected
Still, was there nothing that gave you trouble before the operation settled in?
There was no major disruption. Early on, though, the content of the briefs leaned toward generic advice, and it took a little time before they sounded like us. Collecting the surveys was something we were already doing, so it never felt like extra work for the floor. If anything, the load got lighter, because my own data entry disappeared.
Adopting AI did not add work for people. The job the president alone had been carrying — tally, interpret, communicate — was handed to AI, and the result was an environment where everyone shares the same information every day. That change began to show on the floor the very next day.
Chapter 3 — The Front Line Changed
The improvement brief AI distributes each morning isn't just a report. It organizes the guest feedback received the previous day and shares "what we should improve today" with the whole team. Not just knowing the information, but being able to act on it immediately. At Hotel Yagi, that accumulation dramatically shifted how the floor thinks.
An actual improvement brief ①
Based on the guest feedback received the previous day, AI organizes "the improvement points to address today."
What stands out is that even the framing — how to receive the feedback — is put into words. "Rather than major complaints, these were forward-looking suggestions based on the expectation of 'making it even better.'" That single line makes hard-to-hear content much easier for the floor to accept.
The instructions themselves are concrete — "Today, please focus on the floors (including 7th floor Japanese-style rooms), tatami mats, and grass-fiber clothing" — written at a level of detail that lets people act the same day they read it.
Click the image to enlarge
Restaurant service changed
How did the brief actually function on the floor the following day?
Service from our restaurant staff improved noticeably. As for in-room defects, they now come through linked to the room number, so we can act on them immediately.
The morning meeting went from "reporting" to "acting"
How did your morning meetings change?
Since that day's guest information is shared through AI, the wasted time is gone. As for the content of the surveys, it comes up naturally in casual conversation among the staff. Conversation happening in that natural way is exactly my ideal.
From waiting for instructions to moving on their own
Have you seen staff start acting before being told?
Staff started observing guests much more closely. When you observe, you get praised by name more often. When you're praised, you gain confidence. I think that's the cycle. Praised staff become self-directed, and that turns into "next time, let's do it this way."
The brief AI distributes isn't a document that scolds anyone. It is a shared language for reflecting yesterday's guest feedback in today's service. That is exactly why guests come up more often in conversation among staff, and why a culture of asking "how could this be better?" has taken root.
Even harsh feedback becomes material for improvement
What is the mood like on a morning when tough criticism appears?
For feedback the floor can absorb and fix, it doesn't weigh heavily. The difficult ones are the items that cost money. Staff never raise this kind of issue themselves; it requires a decision from management. Honestly, sometimes I think, that's easy to say but I can't do it right away. Whether we can be clear about that is what matters. That said, as the same voices pile up day after day, the priorities become obvious — and it has proven extremely useful as input for new investment decisions.
Never let people think "saying something changes nothing"
How do you actually handle that feedback that costs money?
Some feedback simply can't be moved by effort on the floor — replacing equipment, for example, where a financial decision is needed. That kind of voice never comes up from the floor, because people know there's no point raising it. So management has to go and pick it up. Here, we look at how many instances of the same feedback have accumulated and use that to set investment priorities. For things we can't start on right away, we make a point of telling the floor when the decision will be made.
Leaving it alone is the worst thing you can do. Once people learn that saying something changes nothing, I think they'll stop reading whatever brief you hand them after that.
That may be a more important point than the AI adoption itself. AI can find the problems. But facing them, deciding, investing, and improving is the job of the owner. At Hotel Yagi, the essence of using AI is not "reading" guest feedback but connecting it to management decisions.
Relationships between departments changed too
Has coordination between departments — restaurant, front desk, housekeeping — changed?
More than the speed of resolution, I feel the thought process for getting to the root of a problem has changed. It used to easily turn into a hunt for whose department made the mistake. Now we start from why it happened.
Not "who is at fault," but "why did this happen." That shift in perspective may be exactly the organizational culture that grows from facing guest feedback every single day.
Chapter 4 — Changes in Management and Ways of Working
AI didn't only change the front line. How the owner spends his time, how he thinks, and how he manages have all changed substantially.
Task time became thinking time
Having handed the work of reading feedback, summarizing it, and relaying it to the floor over to AI, what have you been able to use the freed-up time for?
Dialogue with AI. I throw whatever question comes to mind at it on the spot. Operational improvements, struggles with staff development, reading the weekly report to analyze where our own hotel stands. And not just our hotel — I think about the area as a whole. Time spent on tasks became time spent thinking.
What AI eliminated wasn't merely working hours. The real value was the creation of the "time to think" that an owner ought to have.
Act first, meet later
Has the nature of meetings and reporting changed?
We barely hold meetings to begin with. We look at the report and act immediately. Rate control, Instagram posts — action comes before analysis.
Nobody holds the information anymore
Has the manager's role changed?
Rather than the role changing, it's closer to saying that because managers and staff now read the same brief, the hierarchy of information disappeared. It isn't someone holding information and passing it down — everyone reads the same thing at the same time. That's the state we're in.
From an organization with people who share information, to an organization where everyone holds the same information. At Hotel Yagi, AI functions as something that erases the information gap.
Changes that showed up in the numbers
Have you seen changes reflected in the numbers?
It's directly tied to revenue and occupancy. Our occupancy rate and RevPAR are both tracking above the Awara Onsen area average. We're withholding the specific figures because our own tallies are on a business-day basis while the area average is on a calendar basis, so the comparison isn't like-for-like. But for metrics that are fully internal to us, here is the change we've seen.
Survey Responses
~36 per month (average)
↓~150 per month (average)
| Metric | Before | Now |
|---|---|---|
| Guest survey responses | ~36 / month (avg.) | ~150 / month (avg.) |
| Days from reported in-room defect to resolution | ~7 days | 1–2 days |
| President's own time spent tallying and entering data | ~8 hours / week | ~1 hour / week |
| Responses containing praise for a named staff member | — | ~5 / month |
* Response counts are monthly averages calculated from actual data for the old form (February 2023 – May 2026) and the new form (April – July 2026). Now that the switchover has settled, the figure runs at around 150 per month.
* "Recurrence count for identical feedback" is not shown, as no aggregated data exists for it at this time.
Chapter 5 — AI Learned "the Yagi Way" and Began to Reflect the Culture
AI isn't complete the moment you install it. At Hotel Yagi too, the early AI often returned generic improvement suggestions that didn't sound like Hotel Yagi. But over months of operation, AI shifted from being a mere analysis tool into something that reflects the hotel's philosophy and values.
How do you teach AI "the Yagi way"?
What was the process of teaching AI what makes Hotel Yagi, Hotel Yagi?
We handed over our credo and our mission, as they were. Which also means the reverse: at a property where the philosophy hasn't been put into words, this mechanism won't work well. Here, understanding of the credo ultimately feeds directly into evaluations. Because the daily briefs are distributed tied to our action rules, they also function as a shared yardstick in evaluation interviews.
Even AI's writing came to sound like Hotel Yagi
What changed between the early briefs and today's?
The tone became ours. After handing over the credo and mission and running it for a while, I feel our vocabulary and our values have started to seep into the briefs.
An actual improvement brief ②
Under "Today's Action," the instructions are specific right down to the room number — "please prioritize 7th floor rooms (especially Room 701) and 3rd floor Japanese-style rooms (especially Room 303)." This is exactly what Mr. Yagi meant by feedback "coming through linked to the room number."
And every action ends with a link back to one of the action rules: "→ Action Rule 2 'Act without being told, for the sake of the heart.'" Handing over the credo connected the daily instructions to the philosophy.
The brief closes with the day's message and the mission — "Pursue satisfaction, transform the industry, and shape a future of regional tourism." It reads not as generic AI output, but as something written in Hotel Yagi's own voice.
Click the image to enlarge
AI is not omnipotent. At Hotel Yagi, the posture isn't "leave it to AI" but "raise it until it can make judgments the Hotel Yagi way." As a result, the brief that arrives each morning has evolved from a simple improvement report into a communication tool that carries the hotel's philosophy to the floor.
Are there times when AI's analysis feels wrong?
Have you ever looked at AI's analysis and thought, that's not right?
Yes. When that happens, I tell the floor, "you can ignore this one." The one making the judgment is, and should be, the human side.
How often does that happen, and what kind of errors are they?
Not often. Two or three times a month, I'd say. The common case is a single, unusual comment being written up as though it were an overall trend. The other is a list of measures that aren't realistic given our particular circumstances. In those cases I tell the floor to disregard it.
Don't believe AI — master it
AI organizes vast amounts of information and surfaces new insight. Whether to act on its suggestions, however, is something a human must decide. At Hotel Yagi, that division of labor was clear from the outset. AI is there to provide material for thinking, not to substitute for management judgment itself.
How would you sum up your division of labor with AI in one line?
AI gives you the noticing and the spark. The human question is how well you can catch it and act on it. That's how I think about it.
The greatest outcome of Hotel Yagi's AI adoption was not becoming "a ryokan that thinks with AI." AI surfaces the insight, people decide, and people execute. Because that division of labor became explicit, AI settled naturally into both the floor and the management, functioning as part of the organization. And this mechanism did not stay within one property — it went on to spread across the entire hot spring town.
Chapter 6 — From a Single Ryokan to an Entire Hot Spring Town
The initiative that succeeded at Hotel Yagi eventually outgrew a single property and developed into an Awara Onsen–wide effort. Competing ryokan sharing guest feedback with one another — how was something generally assumed to be difficult actually achieved?
The first reaction was "we don't want our data shown"
What hopes and concerns did individual properties raise during the discussions at the association?
We heard the concern that they would rather not have each property's satisfaction levels and complaints laid bare.
Not aiming for perfect — starting from a form everyone could join
Was there really resistance to sharing data among competitors?
Personally, I think it's guest satisfaction, so there's no problem in being open about it. But we prioritized building consensus, so we began by producing and sharing a report with the satisfaction portion removed. Starting from a form everyone could get on board with turned out to be the condition for continuing. Today the survey is shared with four parties: members of the ryokan association, the city's tourism promotion division, the tourism association, and the community development vision committee. The first three receive it with satisfaction figures withheld, and our policy — agreed by the association — is to disclose all data to the community development vision committee. That's because we believe this survey will become the foundation for community development.
Conversations between properties began to change
Have association meetings and conversations between properties changed?
I feel we talk about Awara Onsen as a whole much more often. Less in formal meetings than in everyday conversation — the area as a whole now comes up. Seeing other properties' ratings and guest feedback is stimulating and instructive too. At the very least, I now look forward to analyzing the monthly report.
Success at one property moved the whole area
Did the results at your own property help push the area-wide rollout forward?
They did. We were the first adopter, and because we had solid results there, I was able to propose an area-wide shared survey at the association and get it passed. Without the practice at one property, I don't think we would have had the credibility to expand it across the area.
The AI initiative that began at Hotel Yagi was never confined to operational improvement at a single facility. It has grown into an effort to share guest feedback across the region and, in cooperation with local government and the tourism association, raise the value of the hot spring town as a whole.
AI as Infrastructure That Changes Not Just a Ryokan, but a Town
The mechanism for accumulating, analyzing, and acting on guest feedback is now being used as a foundation for thinking about the challenges facing the whole hot spring town. Mr. Yagi argues that the system's real value lies in being "a mechanism for continuously searching for problems" rather than "a survey for measuring satisfaction."
It isn't as simple as "the problems are outside the ryokan"
Has the pattern of "high ratings inside the property, problems outside it" changed the area's mindset?
Honestly, I don't feel that it has. The reason is that I suspect every property's real sense is that we should be looking at our guests more closely. However much you develop the town, forcing guests who came to relax out onto the street feels wrong to me. The more useful question is: at what time of day, and with what on offer, would people naturally want to take a walk? I feel this survey comes into its own in unpacking exactly that.
"Let's be like Kinosaki Onsen" isn't a sustainable idea
How do you respond to voices saying "if only we had a townscape like Kinosaki Onsen"?
For Awara Onsen, I think "townscape" is a slightly difficult word, because the large properties are scattered away from Yu-no-machi Plaza. Even if we developed the central district, I don't think that opinion would disappear. We already have Ashiyu. What I'd rather discuss is how to design and develop what is genuinely good about Awara. First of all, that's the hot spring water itself — the surveys rate it highly. Then it's a question of how we convey the context of that water. Beyond that: our enormous variety in price points and guest segments, and the fact that the ryokan here get along well with each other. Like the Sanyakushi guide I've been experimenting with, or the Kinzu Sosaku no Mori experience we built into our all-inclusive plan, I'd like us to connect many of the places each property recommends. The value to protect is loving Awara. The ryokan association's tagline is "OUR LOVE."
It isn't a choice between "bustle" and "quiet"
On balancing liveliness with tranquility.
Guests come here to relax. Not undermining that premise is what matters. It isn't a binary of whether or not to create bustle — it's at which time of day, and with what, people would naturally want to walk. That's what the survey should be helping us solve.
The dialogue with local government changed too
Has your dialogue with local government changed?
Clearly. Specifically, the tourism division and the tourism association now attend the junior chamber's regular meetings, and we discuss concrete measures to attract visitors. Out of those discussions came a decision to run an Awara Onsen feature in a travel magazine. I think that's a very good thing.
"Continuing to search for problems" is what makes a region strong
Can this mechanism become decision-making infrastructure for community development? And how would you like Awara Onsen to have changed in five years?
I'm convinced it will become infrastructure. In five years, I hope this system will have gone through further upgrades and will still be searching for Awara's problems. There is no end to problem-solving. The posture of always continuing to look for problems is precisely what should become our strength.
Chapter 7 — A "Mechanism for Finding Problems" Changes Management
When people hear the word AI, many think of automation or efficiency. The value Mr. Yagi describes is a little different: AI exists not to produce answers, but to find the problems you haven't noticed yet.
It became infrastructure for management decisions
Daily improvement at one facility, and monthly strategy and community development for the area — what does it mean that both come out of the same data foundation?
It's indispensable. I see it as infrastructure for management decisions.
Not a satisfaction survey, but a problem-finding mechanism
Do the unmet-demand estimates match your instincts? And if you had to express the value of this system in one line?
The order of magnitude in the estimate doesn't strike me as off. That there is that much latent demand sleeping there matches my own sense of it. That said, I don't believe we should go after all of that number as lost revenue. Guests come here to relax. Out of the unmet needs, we identify the ones we can meet without breaking the quiet. It's in that prioritization that the estimate really earns its keep.
If I had to put the value in one line, it's a mechanism for finding problems. It is not a survey for measuring satisfaction. I'm also grateful that we can change the question design flexibly — I look at the results and think, what would happen if we added this question? Ideas like that keep coming.
At Hotel Yagi, AI is not an efficiency tool. Frontline improvement, capital investment, staff development, and regional strategy — it is used as the "mechanism for finding problems" that serves as the starting point for every decision.
Final Chapter — What You Really Need Before Adopting AI
What stood out across this conversation was that Mr. Yagi never talked about AI as a magic tool. AI organizes guest feedback, finds problems, and offers hints for improvement. But the judgment and the execution that follow have to be carried by people. And the standard for that judgment is the philosophy — what the ryokan holds dear.
So we closed by asking: what kind of ryokan do you think can make the most of this system?
"I'd recommend it to properties that can put their philosophy into words"
What kind of property, and what kind of hot spring town, would you recommend this to? What separates the organizations where it works from those where it doesn't?
I'd recommend it to a property that can put its philosophy into words. Size doesn't matter. For a hot spring town, it's a region where the ryokan can join hands even while competing.
First, on the property. The dividing line isn't scale or budget — it's whether the philosophy has been articulated. What we handed to AI were our credo and our mission. If an organization has something to hand over, the brief becomes a vehicle for the philosophy. If it has nothing to hand over, it's just an internal memo.
The other factor is speed of execution. We barely hold meetings. We look at the report and act immediately. In an organization where analysis becomes the goal, I think the morning brief ends up as just something to read.
And then there's management's resolve. Feedback the floor can fix through effort and feedback that requires a financial decision are two different things. The latter never comes up from the floor. If management doesn't take it on and just leaves it, the floor learns that "saying something changes nothing." Handling that consistently is harder than distributing the briefs.
As for the hot spring town, I think there's one condition: the ryokan have to get along. You're sharing data among competitors, so without trust it can't even start. In Awara Onsen too, concerns were raised in the discussion of the shared survey. Even so, we started from a form everyone could get on board with, by leaving out the satisfaction portion. Don't wait for perfect agreement — leave something out and start. A region capable of that judgment will make it work.
Conversely, the hard case is a region trying to reproduce someone else's success story as-is. You often hear "we want a townscape like Kinosaki Onsen." But when the locations and guest segments are different, you won't arrive at the same picture. Data is a tool for knowing where you yourself stand. Used as a tool for comparing yourself to others, I don't think it lasts.
Your philosophy doesn't have to be finished from the start
So where should a property whose philosophy isn't yet in words begin?
You don't need to write something magnificent first. Our own credo wasn't complete from the beginning either. When you keep reading guest feedback every day, you start to see, in reverse, what it is you actually value. So the order can be flipped. Start distributing first, and build the words while watching the feedback that comes back. That's a perfectly valid way in.
In Closing
What Hotel Yagi is doing is not "analyzing surveys with AI." It is reading guest feedback every day, sharing it across the entire team, and accumulating improvements — then connecting that accumulation to management decisions, capital investment, and tourism strategy for the whole region. AI functions as the foundation that supports that flow.
What came through most strongly in the interview was that AI does not change the organization. Rather, a culture of valuing the guest's voice is being accelerated by AI. Finding problems people couldn't see, having people decide, and having people execute. Repeating that sharpens the service, grows the organization, and changes an entire region.
Hotel Yagi's initiative is both a successful case of AI adoption and a question posed to the service industry as a whole: how will you carry the guest's voice into the future?
Behind the Build — From the Designers' Perspective
From here, we share the perspective of basicmath Inc., the company that designed and delivered this system: the background to its development, the thinking behind its design, and the product that grew out of this work.
Our relationship with Hotel Yagi and the Awara Onsen area began with a Japan Tourism Agency demonstration project. Over the three years since, we have worked alongside them on marketing, DX, and applied AI in the field.
While considering measures for fiscal 2026, we proposed the system described in this article.
Capture guest feedback, have AI analyze it, and deliver it every morning to every employee as information for improvement. The point was to keep that entire loop turning continuously within daily operations.
Adoption itself must never become the goal
There is one thing we prioritize above all when introducing technology: never letting the introduction of technology become the goal in itself.
What problems exist within day-to-day operations? What improvements are possible? And how can those improvements be connected to a better guest experience and better returns?
Technology is nothing more than a means of achieving that purpose. We are always conscious of that order.
The customer's voice is critical information for improving a business
For this proposal, we focused on how guest surveys are used at accommodation properties.
Even where surveys are being run, there are cases where response volumes fall short, where tallying and sharing take too long, or where the collected feedback never reaches daily improvement.
But we believe that — in lodging and far beyond it — the customer's voice is critical information for improving a business. Problem-solving, revenue growth, service improvement: the hints for all of them are in what customers say.
Collect the feedback, analyze it, incorporate it, improve. Sustaining that cycle is what builds the value you will offer in future, and what makes a property, or a region, one that people keep choosing.
Our Perspective
Dozens, hundreds of guests every day.
Far more over the course of a year.
The opportunity cost of failing to capture
enough of those voices is not small.
It was on that thinking that we designed this system.
Five Things We Prioritized in the Design
01 — Keep the questions to the bare minimum
A survey with so many questions that people abandon it partway will never gather enough responses in the first place.
Rather than listing everything we wanted to ask, we began the design by narrowing down to only the questions that genuinely mattered.
02 — Make the form clear and easy to use
Input fields that are too small. Options that are hard to tap. Answers that take too long. Small frictions like these accumulate and become the reason people don't respond.
We designed on the premise that a respondent should be able to finish, with as little stress as possible, in a short amount of time.
03 — Design the actions that actually collect responses
What we placed particular emphasis on this time was designing not just the form itself, but the behavior that leads a guest to respond.
Leaving a QR code in the corner of a room might mean nobody ever notices it. And even if they do notice, without a trigger to respond, that's where it ends.
For this project we produced a dedicated flyer, handed to guests at check-in with a word of explanation. We also added an on-the-spot prize draw available immediately after responding, creating a path that naturally leads guests to answer in the gaps during their stay.
Not simply providing a usable form, but designing right through to how the feedback actually gets collected. That work is the reason survey responses rose from an average of roughly 36 per month to around 150.
04 — Be able to keep changing the questions
A survey is not something you build once and finish.
What you want to know changes with circumstances and seasons. Looking at the responses you've actually gathered often raises new things you want to ask.
So in this system we prioritized being able to add and change questions while continuing to operate — without affecting the analysis of responses already collected.
Being able to keep changing the questions while you use it is, we believe, another essential condition for turning the customer's voice into continuous improvement.
05 — Be able to keep tuning the analysis
We also refined and optimized the analysis itself through repeated discussion with Mr. Yagi.
The important thing is not to treat it as built-once-and-done.
Watch the actual results, adjust the analysis method, and move it closer to the information that property genuinely needs — without those changes costing significant time or money. We prioritize that flexibility precisely so the system remains usable in the field over the long run.
And Then, Into Software — mitsumonoAI Forms (working title)
Building on practice in the field at Hotel Yagi and elsewhere, we are turning this system into software: "mitsumonoAI Forms" (working title), scheduled for release in autumn 2026.
The goal is to let far more operators easily start the same mechanism Hotel Yagi practices — turning guest feedback into daily improvement. We're developing it as a platform for survey collection, analysis, and operational improvement.
From "collecting" all the way to "improving"
Typical form tools center on creating and collecting surveys. Services that add analysis features are increasingly common too.
But analysis results alone don't necessarily change anything on the floor.
What is the problem? What will we improve? And was that improvement actually carried out? To connect the customer's voice to business improvement, what happens after the analysis is what matters.
With mitsumonoAI Forms, our aim is to connect
Collect → Analyze → Turn into tasks → Track → Improve
— that entire cycle within a single system.
AI surfaces the insight; people make the call
There is another important principle we took from the operation at Hotel Yagi: don't hand every judgment to AI.
AI can read guest feedback and present the material people need to notice trends and problems they had overlooked.
But whether to actually act on it, what to prioritize, how much to invest — the final call belongs to people.
As Mr. Yagi of Hotel Yagi put it:
AI gives you the noticing and the spark. The human question is how well you can catch it and act on it.
That principle carries straight into the design philosophy of mitsumonoAI Forms. Not AI running the business or the floor in place of people, but using AI so that people notice, decide, and act faster. That is the system mitsumonoAI Forms aims to be.
The Cycle
Collect → Analyze → Turn into tasks
→ Track → Improve.
Don't make collecting feedback the goal —
connect it through to the operational
improvement that lies beyond it.
What began at Hotel Yagi has now spread across all of Awara Onsen, developing into a system where ryokan, the tourism association, and local government all draw on shared data to raise the value of the region.
We are taking the knowledge and design philosophy gained in that field and delivering it to more operators as mitsumonoAI Forms.
AI gives you the noticing and the spark. The human question is how well you can catch it and act on it.
The brief that arrives each morning has become a communication tool carrying philosophy to the floor.
From improvement at a single property to community development across a hot spring town —
the system that carries the guest's voice into the future is still evolving.