Sample Size

On Neuromarketing, Cristina de Balanzó the Main Nut of Walnut (I/III)

Edgar Sánchez, Senior Neuromarketing Consultant, Ph.D. Neurosciences

Carolina Castioni, Junior Consultant




Edgar Sánchez (ES): Hello, I am with Cristina de Balanzó the main Nut of Walnut Consumer Neuroscience Consultancy. Christina thank you very much for taking time for this conversation.


Cristina de Balanzó (CdB): Thank you, Thank you.


ES: Cristina, let’s go to the basics first. What is Neuromarketing?


Cristina: I think you can find many definitions about what Neuromarketing is, but I think the most important thing is to understand how the power of the unconscious in the decision making processing. I think this is our fundamentals. You know, we exist because unconscious importance. Therefore, we need to understand this, if we want to predict, understand or change behavior. This is for me, I mean obviously we can overcomplicate the definition and say this well is something that comes from the cognitive neuroscience to understand, marketing activities, this is something that describes. I think this is the essence of what Neuromarketing is, is to understand the power of unconscious and decision making process.


ES: Great. What are the complexities of Neuromarketing?


CdB: I think we have many complexities. Since we have started. Talking this got a lot of detractors and people that cannot see how Neuromarketing can be there, in the center role of these new tools. The big challenge is actually to we as an industry is to make all these insights into something actionable. I think this is a key challenge that we have, you know, like first to give a role in this. I think we have been listening all these speakers [In the Neuromarketing World Forum] talking about importance of integration. I think integration is key because maybe when Neuromarketing was started, we were talking about replacement [of the traditional Market Research]. I think this was quite wrong, so now we had to tweak to reposition ourselves in the industry. Enough to be relevant for the added value that we are offering, because at the end of the day it’s gonna be more expensive because we are adding something. So the key challenge is to really provide to the client the evidence about the added value of our work.

ES: And that’s a complex thing

CdB: Yes, It’s very complex.



ES: So, your conception of Neuromarketing is a complement to the traditional of market research.

CdB: Yeah, I am always saying this. On this I don’t have to make my mind. First time I started in this field. I have been always saying the same: we are working for integration and not replacement. I know so because I think neuroscience is having a dimension that was hidden before, but to understand human beings, and it’s the complexity, we still need to observe what people do, to describe this behavior. at the same time to talk to these people and to get these… I think this is the way we humanize the data, as well, you know like the emotional way has all been something that just people can’t articulate, but then neuroscience establishes these facts, you know things that you want to talk about based on something. So that’s why the whole paradigm makes sense.


ES: Good, what’s the way to measure the success of Neuromarketing projects, is there any kind ROI, the return of investment?

CdB: Well, depending on the metrics of the client. This is going against you, you can establish, you know, how to measure, how you’re going to measure the effects of the investments, of the return of investment of every single project. You can go both project by project, you know, you can help them to optimize a piece of communication, you know, to see if these are particular piece of coms has worked better, you know, is the way business decisions, the things the Neuro has to be able to achieve, being this client or to have a role and a voice about brad decisions. This is a thing what needed to do. But I think in every, in a project basis, you have to establish this ROI you are looking for, and anyone any other kind they are looking for as well.


ES: Ok, good. Talking about methodology, one piece of methodology of Neuromarketing, how big should be a sample for a Neuromarketing study? on the one hand, and on the other what would be the composition of the sample for the study?

CdB: Well, I think we come back again to the business objectives and the sample you need to look at. I mean, in terms of generally you need depending on how homogeneous is the sample. With EEG, you need a minimum of 40 people in order to have a splits in terms of gender or use it- People are talking about 30 to 40 and you are able to split this data. I think depending on the question and how robust, you know, you want to go for, you need smaller samples, I would say. But there is depending on which techniques you are. I am talking on the EEG and GSR. If you are talking about Eye-Tracking, you can do a qualitative work just ten. But then you can have a massive quantitative study of Eye Tracking, that means 200 people visiting the store. Or in Implicit Testing, normally we have samples, this goes online, and for reducing the noise, you need a kind of 130 as minimum in order to proceed. But again I think we can answer sample size restrictions according to the technique that we are using and then how homogeneous is this sample- Because for instance you go from consumer to health, the sample is much more homogeneous and therefore we can go for if we normally need these 40 people, we can go with 15, because these doctors, you know. health practitioners like hematologists are quite homogeneous and therefore we can go with much more sample size.



ES: Anything is perfect, Neuromarketing is not an exception, so what are limitations of Neuromarketing?

CdB: Well, I mean, I think there are many limitations, and neuroscience is something that we are getting close to the answers, like how advertising works, you know, that is a massive question with no answers I think neuroscience is adding an additional insight to this kind of questions that no one has been able to answer. So I think neuroscience effectively is helping to understand this, you know, has put on the table many things that before. just we could maybe test by our intuition, you know, it’s like this like this is gonna work because these are some elements while now we have an evidence like our faces, as our colleagues are saying, are important as to stories that put our culture into work, that’s why helps to get the engagement. So I think now that we have started to have a little lot of brand learnings and we are getting a kind of body of knowledge, you know, to have this intuition to something that has been scientifically proven. So neuroscience can make tangible, you know, what, many many years ago meanings as well, many people as well have been saying, you know, how this element is actually giving the whole engagement. This rational level, it didn’t make any sense, but now it can make it tangible, and I think neuroscience is what is good at, you know to show scientifically a proof that this is happening.


I want to continue reading the second part







Interview with Steve Genco leading author “Neuromarketing for Dummies” (I/III)



.Neuromarketing. its complexities, methodological and other neuro-points



Hello, I’m Edgar Sánchez (E.S.). I am with Steve Genco (S.G.), leading author of the book “Neuromarketing for Dummies”.

E.S.: Steve, thank you for taking time to join this conversation about Neuromarketing.

Steve, let’s begin with the basic things, What is Neuromarketing, what is the definition?, your definition of Neuromarketing?

S.G.: Well, Neuromarketing is basically the study of consumer behavior in terms of the process of the human brain and applying a huge amount of knowledge that has been developed over the last couple of decades in the brain sciences which I include Psychology, Neuroscience and also Behavioral Economics. The application, those insights which those include topics of Marketing in Consumer Behavior, Consumer Decision Making, and basically looking at the consumer as kind of full picture of the consumer brain that we have today as opposed to the kind of limited models that we used to have before this new information became available.

E.S.: Great.



E.S.: What are the complexities of Neuromarketing?

S.G.: There’s a necessity for Neuromarketers to have a pretty strong scientific background. It’s not a do it yourself kind of certain methodology. There is a whole specter of methodologies. Some which are very detailed require Ph.D. level, scientific training in order to do and some others that are easier to perform and to understand how do they operate the data and still get a pretty strong background. So I think the biggest complexities of Neuromarketing, you have to be serious about is the science. Neuromarketing is a field that at least today people come into from other fields where they’ve already established credentials and scientific knowledge and to apply that knowledge to Marketing and Consumer Behavior.



E.S.: Generally speaking, what is the way to measure the Neuromarketing projects, is there anything like a ROI ( Return of Investment). What would be those measurements, those successful measurements.

S.G.: Ultimately, Return of Investment is kind of the ultimate measure marketers need to justify the cost to companies that they worked for. Measurement is a piece of marketing, ultimately its marketing itself is an effective or non-effective cost to testing , marketing pre-testing, post testing and monitoring. Any large product company or small product company for that matter needs to determine how it’s going to allocate it’s spending, it’s expenses in order to maximize the overall ROI. So that’s the kind of long sophisticated answer.

In fact, companies tend to measure their research results I think in a less quantitative or precise manner. And it’s often a kind of a satisfying sort of model where things are going on ok. You know we’re selling, increasing share or maintaining share, and as long the boat is kind of floating along, at an appropriate level. If something bad happens and then is a flurry of activity. What will we need to do, what will we need to adjust and then you might see a lot more research. And then you might see changes in terms of how the company has whole response to the marketing situation.


E.S.: Talking about the methodologies, I would like to ask you about the sample size. On the one hand, how big should be the sample size? and on the other hand, how important is to have a representative sample for the studies?

S.G.: I think sample size is actually somehow overrated problem. I think that a lot of times marketers and market researchers are more comfortable talking about sample size because it’s an issue they have to deal with in every kind of research. So the issue the same sample size in Neuromarketing  is something they can correspond to and talk about. But in fact sample size is a matter of statistical power. Sample size is not, there is no number that it is appropriate, sample, you need to know the strength and the effect that you are measuring and then you need to be able to calculate, you know,  power calculation or to determine what is the sample size in order to extract that signal. You know, from the noise of the data.

So for some kinds of effects, brain effects they are very prominent signals to noise ratio and because of that, in order to identify them, you don’t need a big sample.

Now, the second part of the question. The important issues are which is what is the population that your sample is represented. It may be the case that a signal can be identified with twenty or fifteen people but if you’re interested in what that signal represents in terms of a kind of consumer’s behavior or the kind consumer’s response, and you’re interested in the response of particular segment of the population then you have to be sure that the people that you’re testing are in the right sample, are in the right group.

Second, but a lot that we’re interesting it is comparing segments. So, the sample size issue kind of goes away, if you want to compare so say, that in order  to determine the presence of a signal in twenty people let’s say for EEG or some  other kind of methodology, well so you need 20 people to determine the straight signal, in a particular group, say pair of men and women you have to go to 40 people because you have 20 men and 20 women, because  each of those sub-segments used to have an adequate number to be able to identify the signal.

You also need to be able to do young and old, now you’re up to 80 because if you take each other’s groups you take young men, young women and older men, older women and you got to eighty people. So, it’s really easy to get the sample size up, if the comparison that you want to make require comparing different groups, product loyalists versus product diagnostic. New users versus long term users, every time you want to compare any of those groups you have to double the samples. That’s a limitation because unlike survey methodology where you start off  with a big sample, you can slice the dice according to maybe identifying an interesting distinction in the sample for people who live in a certain sized city, in a certain region, shop or a certain store, they seem to have an interesting characteristic, you can cut and tease that out of the data.

But in experimental research you have to predefine  those groups, you can’t find them afterwards so that’s how often something that focused to survey research has a little trouble understanding the data pre-specified and you can’t just slice the dice of data. You have to create an experimental protocol that gives you, each self one study, for comparison you have to have more people to have this statistical power to do the comparison.



E.S.: The first part of the question was because I have read somewhere that the golden number for a sample size in Neuromarketing is 30. And that’s why I’m asking you this.

S.G.: For some kinds of study to identify some kinds of patterns, 30 may be perfectly adequate, but 30, of who? 30 women, you want 30 young men, you want 30 buyers of a product that buy it at least 10 times a year. So you often find that the most powerful research has some kind of built-in comparison. An absolute measure does not mean very much, if you gonna have a general population and the general population, the amount of attention allocated to this ad, is whatever the number is, that’s much less helpful than the amount of attention allocated to this ad was 47 and this ad was 39 and given this sample sizes, you know that is statistically significant difference. Now you know something as opposed to a general level.


E.S.: The second part of the question was because I’ve read again somewhere that it’s not important the composition of the sample. If you do Neuromarketing study in Sao Paolo Brazil, you can extrapolate to people in the UK or Finland. Some people are claiming so because they say the idea they claim, that the brain does not have so much variability among people and they are claiming those things.


S.G.: I think it’s a kind of misnomer, I’m aware that these claims have been made and I think it confuses a couple of level of analysis. It’s the case that if you get excited about something, and I get excited about something and we get excited in the same way, so we may have arousal, skin conductance increase, frontal lobe symmetry, that’s the way our bodies and our brain process being excited about something. But you may be excited about something completely different than what I’m excited about. It’s not a machinery of excitement that it’s relevant. It’s what is exciting to you, and that very significantly there is a lot of very interesting research coming out now, cultural differences, even perception the way in which some Asian  cultures and Western cultures perceive the scene. They are using eye-tracking, where they’re focus on the scene, literally the way in which, what seems to be a very hard wired perceptual function can be different in different cultural context, in which people learn how in our culture we absorb information. So there’s a lot of really interesting differences. So the idea in defense  and resilience, Spaniards all respond in the same way, at one level is true, all of our brains work pretty much the same, but what we respond to can be very different.

E.S.: Okay,


I want to continue reading the second part…

Neuromarketing, Interview with Elissa Moses (I/III), IPSOS Executive VP of Neuro and Behavior Science Innovation Center


Hello, I am Edgar Sánchez (E.S.), I am with Elissa Moses (E.M.) in the context of the Neuromarketing World Forum in Barcelona



Elissa Moses is Executive Vice-President of Neuro and  Behavior Science Innovation Center at IPSOS



Edgar Sánchez (E.S.): Elissa  thank you very much for taking the time to be with us talking about Neuromarketing.

Elissa Moses (E.M.): My Pleasure

E.S.: Elissa, let’s start with the fundamentals. What is Neuromarketing?

E.M: Neuromarketing is really a new subdivision of Market Research and of Marketing and really was born out of the fact that Neuroscience has developed so much more understanding about how people make decisions, how emotions are processed and that learning has filtered into the whole marketing research industry. So it’s embracing of these new kinds of understanding. Particularly they are described as a System 1 by Daniel Kahneman and primarily refers to non-conscious response.



E.S.: What are the complexities of the Neuromarketing?

Elissa Moses (E.M.): They are many. Partly or fundamentally are struggling with a deep science and so a lot of people don’t have the experience with this science and sometimes they misunderstand things and they get only part of the story right, they embrace tools they may not understand and it really does help to have a some kind of the guide that has neuroscience understanding, experience and can be a real practitioner in terms of knowing what to do it. It’s not just about knowing the science, it’s about knowing how to apply it to the research and marketing world.



ES: What is the way to measure Neuromarketing projects, the success of Neuromarketing projects?

E.M.: The success of a Neuromarketing project has to do with the value that is perceived, to be derived from it by the clients and the people that are working on it. The most immediate value comes from diagnostics. Understanding depending on what you are measuring and what the objectives of study are, understanding if you lose emotional engagement in a certain point of a commercial. If the visual reaction upon seeing a new package or experiencing a new product is negative, that’s really important learning. And so because of the specificity of knowledge that comes from Neuro-mesurements you have the ability to have great detailed learning about how people are reacting unconsciously beyond if they can maybe  put it into words. That’s the beginning.

E.M.: Over time we are seeing more and more empirical evidence about correlations, and certain patterns of neuro-response in market success. I think as the sub-industry evolves we are going to see much better modeling in predicting value once these neuro-tools are combined with conscious tools and we have a more holistic and comprehensive understanding of consumer’s patterns.

ES: I understand that you think that Neuromarketing is a complement of the traditional marketing, research marketing tools and practices

EM: I think that they both go hand in hand and given that we know so much of the importance of emotions and unconscious of how me make decisions and what guides our behavior, we would be very foolish not to include that in our investigation of consumer response. At the same time, cognition and intention matters and so you really need to have precisely the equation.



ES: Talking about the Neuromarketing methodologies, I would like to ask you  about the sample in the Neuromarketing studies. This means two things: what is the sample size? in one hand. And What is the importance of the composition of that sample size? I’m asking you because on the part of the sample size, I’ve heard or read that some people are claiming that 30 is the golden number of the sample size . On the one hand. On the other, some people are claiming that the [potential] extrapolation of the results is quite high that means that is you do a marketing study in Brazil, Sao Paulo,  you can extrapolate that to people in the UK or in Finland. In terms of the size. Does Neuromarketing work require a big sample size, a representative one?

E.M: Edgar this a very complex question because it has a lot of different elements and aspects of it and I have a very firm point of view of this. We’ll take it in pieces.

First of all, the rules of statistics do not change just because you are using neurometrics. So, if you want to be projectable, if you want to do significant testing, if you want to be representative of a sample, the same rules apply, quantitative research as quantitative and qualitative research as qualitative.

There are more factors involved as well. If you ask me questions about a stimuli, I’m looking at the stimuli responding and trying to be rational when looking at the stimuli. If you’re passively measuring my response to something so to say bio-metric, E.E.G.[Electroencephalogram], even Eye-Tracking, there could be something inside that can be a stimuli, a thought that I have. There is no way for you to know. But I think, I forgot to do something my boss asked me to do…Oh my God!, I’m not showing to you on my face because I can control my facial muscles, but inside something’s upsetting me. You’ll never going to know that. You look at the bio-metrics and go Oh my God! something happened, there was really an event here. That is why you need enoughable quality of a sample to be able to take out those outliers. So the majority of a sample response at a certain place in a response to a major stimuli, you have a good probability of believing that is a reaction to that stimuli. If you’re doing it with too few people, you don’t know.

Now, one of the things that I have concerns of what really small sample size is, is that it’s not very repeatable in many cases, I’ve done studies in my previous job working with E.E.G. we would have a much bigger sample taking groups of thirty, how they responded to an add from the same homogenous sample and it didn’t necessarily imply the same results. I think that there is an exaggeration of the example of smaller sample sizes, because it’s wish fulfillment, it’s much cheaper to have smaller sample sizes. I think you could no more project how people will going to react in Argentina in extrapolate that to Japan and you can if you did a 30 person survey and tried to project it to Japan.

I really think it’s a fallacy. And if in fact 30 is enough, what we are seeing is that diminishing returns with some of these metrics at sample. Eye-tracking is one that doesn’t require many, because it’s much more of a just of sort of universal physiological response, tracks your eye It means that you’re measuring very blunt effects. But if you try to measure something that has to do with how someone really feels about something there is contextual. It has to do with cultural references, personal meanings, bigger the sample you are going to need Otherwise, I think you’re measuring a very blunt effect like pain. I don’t have to have a big sample to know if I pick people with a pen, it’s gonna hurt One or two people.

I used to work for a boss, he was brilliant, he said I only need to stick my arm out of the window to know if it’s raining. I don’t need a really big sample. If something is the universal effect, then the smaller, but if something has to do about how do I feel about luxury, how do I feel about the category, how do I feel about the brand, let’s get serious we have to have a bigger sample.


Go to the second part (out of three)