After 11 Years In UX, This Is The Mistake I See Everyone Making.

🔬 The Observation That Prompted This Rant

• We measure satisfaction, intention to use, overall liking — and then we go back to our teams and say "users don't trust it" or "satisfaction is low" and expect that to be actionable

🧠 How Experience Actually Works — A Quick Neuroscience Detour

• Experience isn't one thing — it moves through layers: sensation → perception → judgment
• Sensation is the raw signal reaching your sensors; perception is your brain integrating that into something meaningful; judgment is the conscious evaluation you emit at the end
• Most UX research only captures the judgment — the tip of the iceberg — and skips everything underneath it
• Knowing someone rated satisfaction a 3 out of 7 tells you nothing about what to change

🍷 The Sensory Evaluation Parallel

• My master's specialisation was in sensory evaluation — how do you extract what someone actually sensed from what they perceived overall?
• The wine, perfume, and automotive industries do this routinely: trained panels isolate attributes (texture, pitch, smell profile) and rate them independently from overall liking
• We can and should do the same with software

📐 Hassenzahl's Model — The Framework I Keep Coming Back To

• Three levels: intended qualities (what the conceiver aims to produce) → perceived qualities (what the user actually experiences) → final judgment (satisfaction, purchase intent, etc.)
• The gap between level one and level two is where most products fail — you can intend a premium feel without ever checking whether users actually perceive it as premium
• Decompose until you can't decompose further: "premium" means nothing to an engineer — "high-pitched sound perceived as alarming rather than reassuring" does

💡 What I'm Actually Asking UX Researchers to Do

• When evaluating a product, go beyond overall satisfaction — ask about the attributes that compose the experience: reliability, accuracy, responsiveness, tone, whatever is relevant to your context
• Use rating scales so you can track change over time and compare across studies — even imperfect numbers beat no numbers
• If you don't have time or budget to do this with users, do it internally — train your team to evaluate the attributes so that when you go back to the developers, you're speaking their language

⚠️ The Cost of Not Doing This

• You end up doing redundant research rounds because you never captured the full picture the first time
• Your feedback loop stays shallow — one round of iteration, and then the team doesn't know what to do next
• You are shooting in the air, and the product improves slowly or not at all

Transcript

SPEAKER_01 0:00

In today's episode, my opinion of what we should all do as user experience professionals when it comes to releasing new products, features, and services. I'm switching the UX on AI digest for this episode, which is an opinion episode. And please excuse me in advance for the eventually ranting-oriented episode. Because I want to share what I do think about some of the process that is being done, or most precisely not being done, not being conducted as user experience professionals. And it's with a little bit more than 10 years, maybe 11 years of experience that I am observing this thing. So what I'm about to talk about today is what I think should be done at all times when you are evaluating your feature slash product slash technology with your users and that you're coming back with some results. So, first the observation. So just a reminder about the definition of user experience more or less. It has to do with all the emotions, thoughts, perceptions, sensations that someone has before, during, and after interacting with something. And here I want I want to remain as broad as possible. Something. So as you can see in this definition, we are talking about many things: sensation, perception, and so on and so forth. And so this is collectively labeled as experience. And so that's why we are using the term user experience. So that's the experience of a user. And so once we define those terms, we can think about okay, how can we define further experience? Experience can be conceptualized at various levels, and at least one way of understanding it better is probably the process through which we experience something. So I'm a little bit biased because I did a neuroscience degree. So I know a little bit, I'm not an expert, but I know a little bit about the process of perceiv perceiving something. Which is when you interact with something first, and it's not the first here, I'm just using that for the sake of simplification. First, you you feel that something. So for instance, if I want to interact with my computer, first I do see it. I see it, but I I I don't know that I see it, I just perceive it. Meaning I see my computer on my table, and I can grab it. But first I need to see it. And here the word see is like directly I'm I'm talking about the perception and something that it reaches my consciousness. But before that, the signals like the electromagnetic waves, the let's say the colors are reaching my sensors, which is my retina in this case, and then this is being integrated in my brain to shape an overall image of what is my computer like and where it is in space and so on. And at the same time, I have a million other things happening in my body and my brain, and I also have memory playing, I also have my motivation, maybe I'm motivated to do some work, maybe I'm not. All of that is playing at the same time to form a coherent image that I will be able to digest. And I do not want to let's say complexify this too much, but we can imagine that there is also the attention playing out, so we could want we could be willing to focus on the computer more than other objects that are surrounding us because this is the center of our attention at this very moment, because we are driven towards doing some work with it and not with other objects. So the ability to focus that is of utmost importance. But nevertheless, we can imagine that we can understand that the process through which we use something goes from sensation to perception to action. And so sensation by sensation I mean all the signals that reaches your sensors, so it can be your hand if you touch something, your mobile phone, it could be your eyes if you see something, a car passing by down the street. It could be your ears, you hear something, so that's the signal reaching the sensor. But then when the signal reaches the sensor, of course, it's not yet integrated, and that happens very, very, very fast. So there is an integration playing in your brain, and then all of that is starting to make sense. Why am I saying all that? And so I'm sorry for this kind of side note, but why am I saying all that? It's because this is important when we consider user experience. We are measuring and we are trying to evaluate how people interact with technology. So let me try to complexify this a little bit further. If I use a new software as a service to accomplish my task, that means I would have had to first sense it, then perceive it, then act on it. And then there is all the complexity downstream of me evaluating my interaction with this piece of software according to a million other criteria that are playing, and only a handful of them are accessing my memory and my consciousness. So, for instance, if I use an app, I am able to say probably, probably, that this is XYZ amount of useful or this is XYZ amount of intuitive, but based on what criteria? Because there is a million, there are a million, sorry, that can play out and inform my final judgment. Let me give you a real example. When you eat a dish and that you like it, are we able to say why? Or are we able to say why not? Sometimes not every time, and sometimes a dish is kind of complex, and so we don't have the training to be able to identify and to separate all the criteria that lead to your liking. But ultimately, you liking this dish has some explanation behind. Ultimately. It's because it resonates with your taste, and ultimately it plays with the sensors in a given way. But for it to play with the sensors in a given way, it means it was conceived in a way that resonates with you. So that's all the goal of let's say making products that suit users' needs and that fit their expectations. But there are several levels to that. There is a sensations one, there is a perception one, there is a cognitive one, we can think about Maslow's hierarchy of needs. So there are multiple levels through which you can let's say please a user. And for instance, if I use a software for accomplishing my job, well maybe I will put more emphasis on it being reliable, secure, useful, and usable than it being beautiful. Certainly, but at the same time, I do have things that I base my evaluation upon. That's really important to have in mind. Let's say we consider a doctor. A doctor is trying to make a medical diagnosis of XYZ, how can I say, uh, disease. They need to use some tools to assist them. They they need to, or if not, like we are entering the field of doing medical diagnosis without any tool, and this is very complicated, and so on. That's why tools were used, were developed. So at some point, you need to use tools. It could be radio imagery, it could be a stethoscope, it could be anything. So you use tools, and then based on the information you gather, you are able to make a diagnosis. So let's imagine right now that we develop new software as a service or whatever a software to assist and to help the clinician make his diagnosis. The clinician will think, okay, how is that software? How does that software look? Can I drag this window here and that window there? Because I need to compare X imagery with X with uh Y imagery. So I need to be able to compare and I need to be able to visualize, for instance. Or here I need to be able to put some notes. So here the criteria of usability will play. But that manifests this usability in concrete ways. Usability means something for this person at this moment in time. It's not just usability, yes, but there are things behind this term usability. And so then imagine, we imagine that we want to we want to answer the needs of this clinician through other means, or probably we want to answer more needs. So let's say instead of having to drag Windows around to visualize all the tumors and so on, well, we can have it through a chatbot with AI embedded. And this chatbot will provide the answers to the clinician. And so we can think, oh yeah, it it does give the answer, it's way quicker because we have less tasks to perform as a user, or it takes it takes less time and so on and so forth. So there are criteria. We are able to compare probably the chatbot use from this new state to the previous actions, and we are able to say, oh yeah, it's more usable, it's more pleasant, whatever, it's more reliable, or it's less reliable, but it's more usable. Yes, we are able to compare that, but we do so because we have, or at least the user, when we evaluate, has reasons to believe so. Like there are really practical applications, practical implications behind this new perception. So, this is the same thing as with your recipe that you like, it is the same thing as with the mobile that you use. It is the same thing with the car that you drive every day. It is the same thing. We have a perception, it reaches our consciousness, we evaluate it, and we emit a judgment. That's how it works. And sometimes it's done consciously, sometimes it's done unconsciously. And I do believe that we can be trained at, I'm not saying we will have a one-to-one relationship between our evaluation and our sensation, but we can be trained at identifying the reasons why something clicks or it doesn't. We can be trained, for instance, I'm really much biased because I'm a user experience professional and especially researcher, and I worked in that field for like 11 years. There are a lot of things that don't click with me, and that do click with me, and it's not an overall judgment of the technology that I that I do or the the products. It's I am able to explain the whys to some people, and I could go on rant mode for like an hour to someone speaking about why the iPhone is not behaving XYZ properly, or why the metro architecture is uh probably suboptimal. Because all of the all of the details, let's say, let's say I have a tendency to analyze and I have a tendency to dive deeper when it comes to user experience. But that's probably because I was trained to do so. And I think I do believe that everyone could be. And even if you're not an experienced professional, user experience professional, probably you do have this ability. Some of us are really picky when it comes to our experiences. But what I'm referring to today is that it's possible to emit an evaluation to to to have an evaluation of something and know the reasons why. So I can say, yeah, I like this dish, and I'm able to say why I like this dish or why I don't like this dish. And this is of utmost importance. Okay, why am I telling you that? I'm again a bit biased because I did my master's degree in neuroscience, but more precisely, precisely, sorry, I did it with a specialization in sensation, sensory evaluation. So that means that the specialization was how can we extract? How can we extract the sensation from the perception? Meaning, how can we isolate when someone says, Well, this dish is pleasant, how can we isolate that it's because it has a creamy texture, or that it's because it is um it has a pleasant smell, or whatever, and even if we say pleasant smell, what does that mean? Is it minty? Is it uh whatever uh cinnamon? Whatever, I don't know. And this is from a while back, but you know, it's the ability to isolate, and you will tell me, well, we are not robots to be able to isolate everything at all times, and I'm not saying that, but let me give you an example of why that's possible. On earth, there are some people who are paid to do that for some things, for some areas. With wine industry, for instance, we have uh I don't know how that's called in in English, to be honest, I know only uh French name, but we are people who are professionals at describing the wine, not saying if it's pleasant or not, describing the wine with sensory qualities, and we also have people able, I worked in the automotive industry for some time, to describe the sensory qualities of a motor in a car, or with the sound of a door when it locks, and so on and so forth. Now, let me ask you a question. How do you think that Coca-Cola recipe came to existence this way? How do you think that you like XYZ yogurt the way you do? It's because the recipes, especially with with with um with food, but also perfumes, we can think of perfumes. Like we have distinct teams. We have the teams in charge of evaluating how it is perceived, end of process with someone. Do you like it or not? That's it, more or less. I'm over-exaggerating and oversimplifying. But we ask people, do you like it or not? We see if they buy it or not, that's it, or if they consume it or not. That's the end vanity metric, let's say. But behind that, behind the surface, behind the iceberg, uh beneath the tip of the iceberg, we do have all the conception, all the all the design. And it looks like what I'm saying is really, really basic and is really like obvious, but it doesn't seem like it. I have been working in the user experience field for over 11 years, and there are some basic, basic things that I'm not seeing, really. So, okay. If you make a yogurt or if you make the new Coca-Cola competitor, you will have to say if you will have to tell if this recipe is being liked or not, first and foremost. But you would have had a process to reach that end product. And so if this thing is not being is not being uh let's say consumed or liked or whatever by the user, by the customer, you will need to know why. So that you can come back to it and make the necessary adjustments. You need to know the why. So if it's because your target doesn't like yogurts and you are making a yogurt, well, okay, that's a pretty obvious reason. But if they do like yogurts because they do consume yogurts, but you want to improve on the recipe, you need to be able to tweak the ingredients. So you need to be able to go back to your I don't know how that's called to your chef in this case or to your um ingredient formulator, and this applies also to perfumes and to wines and to everything. That's my point today, to everything. You need to be able to come back to them and say, look, it was not liked, but if you say that, how do you think they can act on this feedback? They cannot act on this feedback because they don't have enough information. That's it. We need to provide as much detail as possible and to be the bridge between between the perception, the sensation, and the physical properties or attributes of your product. So that's it. And now let me tell you. I did this training, which is on neuroscience, but uh with the specialization in sensory evaluation. Yes. But then I realized this is not new. There is a model in user experience which is called the Ascensal's model. I don't know if I'm pronouncing the last name correctly. So Ascensal shared a framework to conceptualize user experience. And it goes in three steps. First, we have like the intended qualities of your product. So when you do something, when you make a product, you do have an intended quality that you expect to reach. So let's say using a MacBook to type on the keyboard, it should produce a XYZ feeling. So for instance, a premium quality feel, and so on and so forth. And then the teams need to define what does that mean to Have a premium quality feel. If I type on a key on the keyboard, what is the sound it should generate? What is cheap sound like? What is premium sound like? We need to be able to compare. And then, once we know, we need to be able to isolate the physical properties so that we can generate that feeling. And this is not science fiction, what I'm saying. This is real. Like, how do you think that we have great products? How do you feel that we can all think about a great product because we know that we have references for that? Evan, like if I tell you right now to think about a great perfume or think about a great computer or think about a great car, you will be able to think about that. And if I ask you to think about the best car you have been in, you will be able to tell. And even that car, you will be able to say, look, I didn't like XYZ, even though the car in itself as a whole was great. Well, that's a proof that asking someone to judge a product overall is not the only thing you should do as a user experience researcher. You should isolate all the components that lead to a final evaluation. And so you will be able to say, Oh, look, I have been in this car, and I remember, well, the steering wheel sensation was amazing. Yes, that's what we want as designers and as conceivers of technology or objects or experiences. That's what we want. We want to have the details because we want to be able to go back to the engineers, designers, developers, and so on, and to say, look, this person or this group of people, they said that XYZ aspect of the experience could be improved this XYZ way. I hope you get the point. Okay, so this is Ascensal's model. He states basically that when we conceive an experience, we do have something in mind. So we want to generate a feeling. So let's say premium or soft or whatever creamy texture, whatever. So we do have several things that we have in mind. And so that's the conceiver's space, the conceiver imagination, let's say. But then it doesn't mean that it will be perceived this way. You can think about all the premium qualities you want to embed in your product. It doesn't mean that the end user will perceive it as premium. And that's the second stage in the Acensel's model. You need to evaluate that product with the user and ask them how do they perceive it alongside the dimensions that interests you and that you think should be assessed. So how premium does it feel, how creamy does it feel, and so on and so forth. And this, even that, is like an interpretation of the attributes, because premium, what does premium mean? So when you have a doubt, try to decompose what you are assessing. Because the more you decompose up to the point when you cannot decompose anymore, then you have your relationships between the attributes of your product and the sensation it provokes and the perception it provokes. So if you only evaluate premium, yeah, that's great. If your team internally has all the framework to relate premium to what it means when conceiving the product, fine. But if you don't, you need to be able to evaluate that with your users. Evaluating all the subcomponents of a premium feel sound, touch, aesthetic, style, and so on and so forth. You need to be able to do that work. And so that is that is the second level, which is is it is the intended perception actually um reached? Okay, and then the final level is okay, now that we have evaluated the perception along these dimensions, what is the final final judgment? Is it satisfactory? Is it easy to use? Is it I don't know, will people buy it? This is like vanity metrics. This is usually what the business is interested in. But there is a danger to that. It's that if you only do this and not the rest, you have the tip of the iceberg, you don't have the the full story, and you cannot come back to your designer saying, look, the satisfaction is a bit low because the premium feel is not there yet, or the satisfaction with this chatbot for medical diagnosis is low because the chatbot is unreliable, because it is giving XYZ answer like it's it's it's uh variable and it gives a different answer every time. For instance, in three cases out of ten, it's randomly it randomly outputs wrong wrong things. Whatever. I'm sorry, I'm I'm taking wrong um weird examples. But hopefully you can get the point, which is you need to be able to have a bridge or to be a bridge between what the users feel ultimately and what you will tell your designers and your engineers because your engineers are not thinking about they are not talking to users, this is your job. Your engineers need to be able to know, okay, what is the sound that I should that I should output here in my car. For electric cars, we have ways now to there are ways to change the sound that it makes because the motor doesn't make as much sound, so for safety purposes, when a car passes down the road, the car needs to have a sound on top so that people hear that a car is coming. And so there is a whole discipline of generating sounds for cars, right? And it needs to be somewhat safe, but at the same time pleasant, and so on and so forth, right? So imagine your car engineer, your sound engineer, they will not think about pleasant, not pleasant. This is not something, it's not their vocabulary. Their vocabulary is the frequency of the sound, the layers I should add, and so on. And so for that, you need to be able to go as close as possible to the language they speak. So you need to be able to tell them, look, this sound it evoked alarm for them and not something pleasant. Whatever, because they felt it was high pitch, for instance. So, okay, I hope I make my point across. The idea is to isolate as much as possible the sensation from the perception. The sensation is a description of the attributes of the product or the feature or the thing you're using, and the perception is the final judgment that you make about this product. And so there is a whole discipline which is called sensory analysis, and this is something that we oftentimes know in some discipline in some other areas, like the wine, like um, let's say perfumes, all of these things, because we naturally understand that they are the people, let's say, evaluating how it is perceived, and they are the people formulating things like in the lab. We oftentimes think, conceive them in a lab, like they tweak a little bit the ingredients, they change a bit the proportions, and then it changes the overall aspect. Well, but this is a caricature, of course, because this is real, and this is an example that we oftentimes think of, but it can apply to everything, to everything. And if you don't have clarity right now of where what direction you should go in when you change your product, it's because you have not identified all the attributes that that contribute to the final perception. That's it, and this applies to everything, and so that's my rant today. I'm sorry if it feels like a rant. I will try to hold as much as possible. But yeah, it does apply to everything, whether you're doing a software as a service, or a physical product, or a dish, or a perfume, or anything else. It does apply to everything. You need to be able to distinguish. And so, as I said, this is a discipline called sensory analysis. You don't need to be a sensory analyst to do that. It's just a question of when you ask questions to your users, like not only please do not only ask if they are satisfied or not, please do ask about more things like the attributes. And it can be done qualitatively, but I would really, really encourage you to do it with scales. Because scales, they will give you numbers, and the numbers of over time you are able to use them to track evolution and to make relationships obvious between your attributes. Now, the last thing, of course, we need to think about something that is really important, which is not everyone rates the same way. Even if I ask you, to what extent do you feel this sound is alarmful on a scale of 1 to 10? So that's an attribute that goes before in terms of the sensation and perception chain before the final judgment. So alarmful sensation, someone can rate it as a four and think that it's highly alarmful. Well, maybe not a four, but a six. And someone might rate it as an eight, and for this person, eight is high alarm because we don't have all the context. For some people, like we don't have the reference points these people sometimes they have heard in their life sounds that are way more harmful than that, and some others a little bit more harmful, but their reference points are the way they are, right? So there is a whole theory when you administer skills to people. Some people advocate for let's say standardizing all the results because not everyone rates the same way. And some people say no, no, we shouldn't standardize these skills because the the differences with which people rate is important in itself. Anyways, I don't want to enter into this debate. I think that it depends on the on the on the end need. But ultimately, if you do administer skills to people, it will help you track the changes and it will help you, even if it's not statistically significant, because it will oftentimes not be, you will be able to go back to your to your the person in charge of conceiving the technology slash product slash service, whatever is being used, and to adjust and to tweak the components. Because they will have reference points and it will make sense to them. But yes, and then of course, all of that begs more questions, so you you can be asking, okay, but what are the attributes I should evaluate? And this is the point. Um, so in sensory analysis, as I was saying, not everyone rates the same way if you're not trained. So if you ask someone to rate a wine on some attributes, and he's an expert, he will be able to rate. Like, let's say the color properties, the texture, the taste, and the taste, they will have more attributes. Like they will have they will have like how can I say like a wheel of attributes. And it's the same for coffee. Maybe if you're not familiar, you can look at coffee attributes and you can have a lot of attributes and descriptors. And so that's the first thing. Like, if you're not trained, you will find it difficult to describe something. That's the first thing. I am not able to describe a wine. I'm not a wine professional, I'm not able to describe it. But if you if you're trained, you will be able to. So that's one thing. The second thing is once you're trained on how to describe, you can be trained on how to judge, on how to evaluate the presence or absence or the degree to which an attribute is being met. And so you have two steps: first the attributes and then the rating. And then finally, all the people need to agree on these ratings. And so, concretely speaking, that looks like you have a group of people testing things together, and then they fine-tune the ratings accordingly so that altogether they arrive at something coherent. So, altogether, the group of people outputs a rating on the attributes that is that is more accurate than the sum of its parts. So it's like an instrument of measure which is composed by several humans. And these people, we ask them to well, we ask them, we train them to let's say ignore the the rest of their perception, the rest of their feelings, the rest of their memory, and so on and so forth, because in the moment we generate a context that makes it easier. For instance, if you taste a wine, it will be done in a black cup. So that you, if you only evaluate the taste, you will not evaluate the vision, the the visual aspect. So that's an example, of course. And that applies to a lot of other things. I worked in the car industry and and we had like rooms in which you had no sound coming from the exterior that could pollute your your perception of the sound. So there are multiple aspects, and so you can imagine that the people evaluating the attributes of the product are being trained to be like an instrument. And an instrument should be reliable, should be efficient, should be effective, right? So trained to be an instrument, like if it were a robot, but it's a human. But it's a human. And I think this is important that it is still a human because we don't have yet robots which have let's say the same sensors depth that we have. Like we haven't reproduced like a human ear or a human tongue probably to be able to reverse engineer how something is perceived or sorry, um sensed. Anyways, so where do I want to get with all this? Okay, if you work in software, let me give you an example. If you work in software and you are working on a new chatbot and you wanna have a guess at if it is, let's say appreciated among your users or not, you could go the route of well, measuring satisfaction and that's it. Or measuring intention of use, which I don't like too because people are really bad predictors as at saying if they will use something or not. And you could do that if your guess is should I use it or not, should I release that on the market or not? Because at at the very at the very first time, this will be your only preoccupation. But imagine you're well established and you want to improve your product. You don't need to ask questions regarding will is it is it is it um liked or not. It's not only that, you need to improve the product. You know the product in itself is liked, you know it fits needs, you know it answers some real problems, but then you need to improve it. Like the attributes of this product, and for that you need to be able to come back to the formulator to tell them look, XYZ criteria, XYZ attribute, physical sensation should be changed. And my point today is that we tend to do that in some obvious environments like chemistry involving chemistry, um, like perfumes or wine and so on, but we we don't do that in all areas, and it's it's um it's a pity in my opinion. It's a pity because we double the efforts doing research studies, we double the time, we do redundant studies because we lack part of the picture. It's like gray areas. It's like if I release a chatbot and my users are saying they don't trust it as much, and then okay, yes, I go back to the developers of the chatbot and I tell them, well, they don't trust it, they will tell me why. Well, I can tell them with the quality, qualitative information, but ultimately it will it will last one round of iteration, but then that's it. They will be able to have references point reference points. So if I tell them it's not reliable, yeah, okay, it's not reliable. Okay, so I need to increase reliability, but how does that work? Increasing reliability. What does that mean? Not reliable. So I guess I hope, sorry, I hope I made my point across. Always try, it depends on the scenario you're in. Okay, I want to make this clear. If you're pondering whether you should release something or not, because it answers a need or not, maybe you don't need to do that. But if you're a well-established product and you need to improve it, you need to be able to have a perfect bridge, like a flat and smooth bridge of relationship between how it is perceived and what you should change on the product. And this is real. I worked on attachment, emotional attachment to a product. And I was able to identify up to the attributes that we could place in a product to create emotional attachment. I'm not saying that I reached emotional attachment, but I worked on this field and in this domain. So this is something real. Even if you don't aim for emotional attachment, maybe you aim for usefulness or intention of use or satisfaction and so on, maybe you aim for that. And so you need to evaluate the attributes of your product. So I'm begging the user experience researchers to do that. And finally, my only point is that you can do that in several ways. If you don't have the means to always recruit, because that there is a an important point, which is sometimes we have limited time, not sometimes, oftentimes, we have limited time to do our research with end users. We can only ask them so many questions, such as um, well, to what extent are you satisfied with this product, and that's it? Or or maybe um uh rate the extent to which you agree with the following statement, and and the question is uh is it easy to use? Is it uh it does it answer your needs? And you you you don't have time to answer to ask all of the questions, like to what extent do you find it's reliable, to what extent do you find it, I don't know, uh accurate, and so on. We don't have time to ask all of these questions. So if you do have time, please do it. I really encourage you to do it. Go down the rabbit hole and ask all of these attributes and make sure you can reproduce this setting across studies so that you have points to compare over time. You can track the evolution and you can go back to your chemists, let's say your chemist, quote unquote, so that they can improve the product. If you cannot, with end users, at least do it internally. Train your team internally on the sensation and on the perception of the product. Because if not, it will be evanescent. You will tell, oh no, it's not reliable, and that's it. And the team is like, it's not reliable, but they don't know what it means. So then you are shooting up in the air. It's crazy to me to see that only in some industries we do that. So a good methodology is heuristic analysis and attribute evaluation. I I don't see this method often. Well, the method itself is not attribute evaluation, it's it's rating scales. You basically rate the extent to which something is present or something fills a criteria and that's it. And then well, you do that and you and you have your your attributes being evaluated. And you can do that internally if you don't have if you don't have access to users, but at least please do it. Find a way to do it. So I hope I made my my pon across. Evaluate the attributes, come back to your chemist, iterate on the formula, and you have a winning recipe. Thank you for listening today's episode and see you in tomorrow's episode. Cheers, bye bye.