The science behind how our opinions are formed / Data for Bluffers #10

28 March 2022

Do you think your customers make their decisions based on the great creative or product features? Listen to Tom and Ed discuss how people do in fact make decisions and how you can influence them.

Lorem ipsum dolor sit amet, consectetur adipiscing elit. Ut elit tellus, luctus nec ullamcorper mattis, pulvinar dapibus leo.

Tom

Welcome to another edition of the data for Bluffer’s podcast. We’ve got a good one this week for music lovers out there. Doesn’t matter if you’re not a music lover, there’s still some great takeaways, but really we are discussing what we can learn from sort of the, the makeup of the musical L up in terms of quality and fame and how we can use that to better execute marketing campaigns in terms of reach and success. And, you know, if you, if we look at the, look at the bands out there today, there are some great bands, and there are some quite frankly, really poor bands. And in both categories, there are bands that are really famous. You know, we get really famous brands that are great and really famous brands that aren’t great. And really the, the point there is the quality of the output is not necessarily the determining factor of their success. And really, I wanted to kind of discuss that with ed and, and how we can look at the data and, and understand that and see how we can use those lessons in our market thing. So, ed, can you explain why both rubbish and great bands become famous?

Ed

Unfortunately, not exactly right. Otherwise, you know, I mean, the world’s greatest there in our band, but there is quite a lot that, you know, science can teach us about this. And I mean, what really comes to mind is quite a famous exp Aman that was conducted nearly two decades ago now at the university of Columbia, by Duncan Watts who went on to work for Yahoo, he was kind of a, a mathematician network scientist. And then he did some work with sort of sociologist psychiatrist, Matthew Salick, and basically what did, was an experiment in sort of the early days of music being shared and streamed on the internet. I think for this experiment, they actually had to build their own streaming platform, right.

Tom

That was Spotify version one.

Ed

It was a stream platform and they, they chose, you know, carefully selected a, a bunch of songs to go on it. All of which were by new unheard of band. Okay. Every user, as they went on this website could choose which music to listen to and could say whether they like that music or not, all they could see on the website was the name of this band, which they’ve never really heard of, but they could also see how other people have rated that song. Okay.

Tom

Like a ranking or

Ed

Something. Yeah. So like a ranking or a score. Okay. And what they found was that people were more likely to give higher scores to songs that other people had already given higher scores.

Tom

Okay.

Ed

They initially tested this by splitting all of the people in the study into what they call different universes. Yeah. And the people in each universe could only see the scores given by other people in their okay. Yeah. And in each of the universes, it wasn’t the same songs that did well.

Tom

Oh, right. Okay.

Ed

It was different songs in each universe suggesting the influence of other people’s rankings was driving people to kind of like the song more.

Tom

That’s really interesting. Yeah.

Ed

To make this even worse, they were, or not even worse, they were one step further and they actually started lying to people right. About what the ratings of these songs were. So they would show them fake ratings. Right.

Tom

Okay.

Ed

And they found doing that they could improve scores by showing fake ratings.

Tom

Right. It, so almost like take the worst song in the list and say, that was the number one. And then all of a sudden everyone started voting for it.

Ed

Yeah. That’s that sort of thing. They, it is worth saying there, there was a, a small caveat in this that they found, there were some songs that just never did well. Right. Okay. There are some songs that were just terrible and never did

Tom

Well. Well, that’s good. So really terrible music will always always lose.

Ed

They sort of went to like music industry executives. I think it was after the, this, the response was a little bit like, yeah, we know, right. They’re sort of like, we can tell who the bad bands are, but when it comes to picking the hit makers, we’re actually not that good. And the difference potentially go well, comes from the fact that there is some random behavior at the beginning, which decides whether people kind of get traction and start getting fans.

Tom

And I guess there’s, there’s parallels there to everything in life, right. There’s tons of products that, you know, are not differentiated by their feature set. You know, everyone talks about a us P but if you look at project management philosoph, I’ve gotta find, I’ve gotta find a new example of, of, of it that like just my, my go to does the same stuff apply there is, is, is that yeah,

Ed

Definitely. I think the interesting feature in all of these examples that we’ve mentioned is that you tend to have a large number of failed bands or bands that don’t have very many fans and a very small number of massive super bands that this is a particular thing that we find across a lot of products and service is in markets. And so, you know, when it comes to phones, there’s a very small number of super brands for phone. This is a stark difference to say, for example, the distribution of people’s Heights. There’s lots, lots of people around the average and a small number of people are a lot shorter than average. And a small number of people are at tall than average. Now what we find when we have any sort of, kind of social behavior around a product is you very much stray away from that. And you have a small number of very large brands and a large number of very small brands from a mathematical and from a behavioral science perspective. The interesting thing is, is why does that happen? And I guess, I guess a very base level, what is required to make that happen?

Tom

This is linked, I think to, is it the, kind of the Kevin bacon stuff we talked about last time in the, I guess if you look at the number of successful actors in the world, there’s, you know, versus the number of actors in the world, it it’s the same thing in terms of a small number of very successful people versus the, the long tail of a everybody. Yeah.

Ed

Almost certainly it’s the same, it’s the same sort of thing. Okay. You know, a small number of actors get all the big lead roles and this might be controversial, but a lot of them, you would not say are orders of magnitude, better actors than the other people trying to fill those. Yeah. Okay. Now the kind of the, yeah, from the mathematical perspective, it is interesting why this could happen. Right. And we’ve talked about one example. People tend to be not very good at making I say independent decisions. You know, people, people that influenced by what other people say about a product sport in more subtle ways, you know, in terms of how many people they see with a different phone or something like that. Right. You know, the fact that you see loads of people with an iPhone gives you an idea that an iPhone’s probably a good phone.

Ed

Uh, it doesn’t have to be, you know, that you go on and check the ratings of the iPhone to find that information out. And then there is another method by which this can happen. What’s often called or was initially called, I say, competitive advantage. And that’s this sort of availability idea. So the idea that, because something is more common already, more people just see it. So it’s instead of, instead of making necessarily a, a, a, a direct assessment based on lots of people think this is popular. People just see it enough that they okay. They know that thing exists.

Tom

I see people drinking Coca-Cola so I will drink Coca-Cola as opposed to Joe blogs, homemade co brand that no one’s ever heard of.

Ed

Exactly. And that, yeah. Not, not because you counted how many people using Coca drinking, Coca around you and said, oh, that’s more than Pepsi. Yeah. But just because you are kind of aware the brand is yeah. Is available. Yeah. And so this is something that was, uh, sort of first detected, I would say, in, in, in network setting by price, who was a mathematician who was looking at citation networks. So networks of citations and scientific papers. So that’s, when, when you write scientific paper at the end, you write your references. So who have you read that has influenced you to give you the thoughts and also what evidence are you drawing upon? And what they, he found was that you could generate these big disparity and number of citations, which is often uses a measure of quality of scientific paper is how many citations it has.

Ed

And he, he found that you could generate that big disparity in number of citations, similar to the thing we see in the size of bands, um, where you have a small number of papers with lots of citations, lots of papers with very few citations. And you could generate that just by understanding that the more you are a paper’s already been cited, the more light people are to be aware of it, because they’re gonna have read a paper that cites it. So that’s kind of like this availability can also lead to this big distribution. Yeah. This wide distribution between papers without very many citations and a small number of papers with lots of citations. Yeah. And this whole thing does have a name so called Matthews law, which comes from the, the line is, is about the rich getting richer. So it’s this general idea that once you have some fans or some followers, you know, or you are a successful band with lots of fans, it’s easier to get more, another kind of name for a similar effect is the first mover advantage this idea that, that if you are the first into a space, getting those initial customers while there’s very little competition gives you a head start once other people try and enter the market.

Tom

Yeah. That makes sense. Um, and, and, and I guess that’s how Facebook started with their, their campus piece. You know, they, they, they certainly weren’t the first social network, but they went after quite a niche area to become first in that area or, or, you know, more popular in that area, if you

Ed

Like, I think, yeah. That’s definitely the case. I think the important, when, when you start thinking about this in terms of networks, what you find is that sort of a availability. So whether, you know, you know about it because your friend is doing it, but also the behavior aspect, the fact that you’re being led by your friend’s behavior work together and almost supercharge the effect. If I want to know about something, okay. What do my friends do? But also which those, your network to bother joining I’m gonna join. The one that most of my friends are on, because that’s the most useful for me

Tom

Just trying to make, you know, bring this back from the marketing side. It it’s really your, your product gets amplified. If the people that know about you are grouped in some way, you know, if you’ve got a thousand customers and they’re spread randomly over the entire globe, people won’t be aware of you it’d be quite small, but if those thousand customers have some form of connection to each other, other people are gonna see your brand as big. Um, they’re more likely to be aware of it and more likely to follow it, I guess, from the availability, you know, using that availability example,

Ed

It sort of shows the power and is kind of the reason why this thing that we’ve mentioned before on the podcast, the idea of kind of appearing bigger than you are. Right. You know, if you are starting off a start now, you’re not gonna be competing with Microsoft this year, but you could be the biggest brand on a street or on a business park or in a particular city. And that is how, you know, you know, the, a lot of the social networks kind of grew, right. And we’ve, we’ve spoken about that before. And the fact that in that case, it was because their product really, really depends upon having their people interlinked, but actually it’s true for any product that the more your customers are interlinked with each other, the more likely it is that you appear kind of attractive. Yeah. Okay. When people do that assessment of what do my friends think of this product?

Tom

And, and, and you said the word attractive there, you know, what this has got me thinking about is if, you know, we, if we boil this down to the school playground, you know, you get, um, popular kids and unpopular kids, right. And, and, and, you know, by, I think from what we’ve talked about, the popular kids gets more popular because people see him or her with lots of friends and, you know, want to gravitate towards the more popular kid. But also there’s the kind of traditional use of the word attractiveness there. Right. You know, it’s, it is attractive people, you know, who will generally be the more popular kids at school, you know, in a beauty point of view, is there a concept of beauty at attractiveness that exists in citation networks, film, stars, you know, businesses that isn’t necessarily about how something looks, you know, they’re just another thing that makes, makes something more appealing than others outside. Just, just how available it is, you know, how, how popular it feels

Ed

That loops round to what we were saying at the beginning about music, right? It’s not, it’s not that people necessarily like what other people like explicitly, because they’re trying to copy them. What the Watts experiment shows us is they are liking the music that their friends, like, aside from how it actually sounds because of our, our understanding that other people are also like that music. And I think the same is true for brands, right? That people, a lot, a lot of people who have an iPhone, right? My partner’s a very good example of this. She insists on having an iPhone. She couldn’t tell you why she insists on having an iPhone, because you don’t go out and test drive, you know, a bunch of phones and then decide which one you want the best. You just kind of have to have to make an assessment of the world around you. And so as a brand, if you can appear a lot of times in that person’s assessment of the world, then you are like the music that, that in the experiment they’ve, you know, given a high score to.

Tom

Yeah. Makes sense. Yeah. That’s really interesting. So then if we, you know, if, if we use, let’s take a, a B2B SaaS product, what’s a more important point about determining its success, at least in its, you know, till it hits the, the brand availability of, of Microsoft say is how well connected its users are. So when someone new comes along where they’ve likely heard of it from, or seen it within their existing, so circle be that professional circle or social circles, but if all their customers are, are randomly distributed, it’s unlikely to surpass a product that maybe has inferior quality, but highly connected customers.

Ed

Yeah. I, I think that’s, that’s fair from, uh, the point of view of, of new products. They definitely require almost some luck and it’s kind of built into a lot of startup, what you might call philosophy, an entrepreneur philosophy, right? This idea of having a niche and, and trying to exploit that niche in a lot of cases that sort of presented as you know, because you wanna understand your product market and therefore make the product that fits the niche the best. But I actually think a lot of the impact of having that niche comes from building a group of customers who are able to sort of share your product with each other and then your customer base can grow from there.

Tom

Yeah. And I guess that’s the, the social proof, the case study thing as well, that, that fuels that, you know, that if you’re in insurance business, um, you feel better about hearing their, your, this people working, whether insurance business and seeing insurance business logos on the website.

Ed

Yeah, exactly. Like, you know, people put logos on websites all the time, and there are some logos that, you know, for very big businesses for like Microsoft is a classic example of this many, many, many startups that have a Microsoft logo on their website for a wide variety of connections. They may actually have with Microsoft,

Tom

Someone from Microsoft signed up for a newsletter, one

Ed

Is ugly. Is there to kind of bring authority to the products.

Tom

Yeah.

Ed

And that’s one type of authority, or it doesn’t tell people that products for them. Yeah. That’s where having, you know, the logos of all of their peers, for example, gives that idea of kind of group attractiveness to go back to our attractiveness example. Yeah. If you, if you work for a large insurance company and you go on someone’s website and it says basically that all the other large insurance companies are using them, you start to be worrying why you’re not, you know, much more than like long before, you know, what the product does. You’re worrying. Why don’t we use this? What are we missing out on?

Tom

So it’s almost the, the ideal, um, logo display on a website, you know, the obligatory, this is what we work with should be people that your prospects are, are, you know, it’s either in the same industry or ideally they already know, you know, that that would be, that would be super powerful if you could orchestrate that.

Ed

Well, yeah. If you could, if you could like put the logos on the website of companies where the person knows people that work there right. In similar roles,

Tom

Product feature, uh, 20, 23, January 23, we’ll, we’ll get that in there

Ed

And make it GDPR compliant. And then

Tom

Yeah, exactly. Yeah, yeah. Maybe 20, 24 with mouse. We might engineering to get that though.

Ed

But I mean, the example, a good example of this is, I guess, is, is Uber when they started right. Their, their niche, they used to talk about, uh, would be to try and, you know, get enough Ubers outside a particular board meeting that they knew was ending, you know, for a large, like fortune five, 500 company, because they know those people in the room are talking to each other. And know if that meeting happens once a month and the first month they get two of them get Ubers back to the, to the airport from there. And they all, they have a really great experience. Then the next month, three of them get it, four of them get it. And then by the end of the year, companies booking everyone Ubers before the meeting even happens,

Tom

Where does kind of, well, does it, where does of mouth and referrals fit into this? Cuz it, you know, if I’m, if I’m hearing, you know, if I go and I dunno, buy something and someone says to me, you know, check this, check this out. I’m, I’m more likely to go and check out that brand, be it the iPhone, you know, be it the, whatever it may be, you know, is referrals just a kind of, part of the engine of what we’re talking about here.

Ed

Word of mouth from a referrals are another way or the, or the real life way that we see those, you know, that little score that was next to the, the music that rating that number of thumbs up in real life. We hear that through word of mouth, subtle drops into conversation, seeing other people with products, if they’re physical products, that’s how as humans, we kind of see what’s going on around us and assess whether something’s popular. And that’s why in the real world, those soft signals are more important. You know, if you’re on Spotify can see what the most popular song that’s played is, but most people don’t choose the music that way, right. Their friends suggested it, or they’ve heard it around, you know, in a lot of cafes that they’ve been in. Eventually they’ve been like, I keep hearing this song, I’ll look up what it is. You know, those sort of softer effects show us what’s popular much more than actually looking at the rating of the song.

Tom

It, it really sounds, you know, if we, if we replace song choice with marketing campaign, there’s a degree of randomness to how successful these campaigns are. So when, when it comes to, you know, looking back historically over your campaigns over the year, you know, you see a super successful campaign to say, right, let’s replicate that. But actually there’s probably a degree of randomness in why that campaign. And if, you know, if you could rerun history and say, take the same amount of campaigns that you ran that 12 months and run them in another 12 months, the successful campaign might be totally different.

Ed

That’s definitely true. I, I think one way of thinking about that is that a campaign that’s a hundred times more successful is not a hundred times better. It may be no better at all. It may be ever so slightly better or worse. Yeah. It could even be worse. Right. And anyone who uses social media for marketing will know that if after you post, you get a bunch of comments within 24 hours, then that, that post might be seen a hundred, two a thousand times more than if that post didn’t get those comments in the us 24 hours. And that’s that kind of first mover advantage. Right. You know, it misses out on its first mover advantage and it never gets another chance,

Tom

I guess that reminds me, you know, you hear the kids, you know, they’ll post on TikTok or Instagram. And if they don’t get a certain amount of likes within a certain amount of minutes, you know, they’ll, they’ll delete it. Um, I guess it’s, it’s that type of that type of principle. But the other thing that’s making me think that, you know, a lot of work, a lot of work goes into, you know, as I say, the planning of the content and the messaging, maybe not as much work goes into the execution, you know, how do, how do we get that? How do we, how do we, how do we use that first move advantage? How do we get that, that early boost? But so there’s probably not enough re um, recognition of sometimes the randomness of it that actually it might not get the start that it wanted, that isn’t always a reflection of the content. Um, it might just be, you know, the, the dice didn’t roll your way that time. And actually, you know, take, have another go.

Ed

That’s actually kind of a really important sort of positive message you can take from this is the idea that, you know, sometimes things do well when they don’t deserve it, but equally a lot of the time campaigns don’t do as well as they probably could have done social influences. In particular, I’ve seen this quite a lot on LinkedIn is people talk about, you know, recycling posts as an easy way to generate content. The reason they do that is that there’s kind of a tacit understanding that just because a post didn’t do well the first time doesn’t mean it’s not gonna do well the second time. And understanding of, of this ran, these random processes enables us to kind of not make decisions too quickly or make judgements rather too quickly. It’s really common for videos and things like that to go viral years after they were initially posted because they kind of get picked up and they get that buzz and they get lucky. And you can kind of think the same with your campaigns.

Tom

You, you get the viral video or the video it’s not yet viral from two or three friends and all of a sudden it, it, it hits that tipping point.

Ed

A a really a good example would be, you know, Jim, Jimmy Carr got in some hot water, shall we say at the beginning of this year, over his Netflix festival, but that been on Netflix for a few months at that point. Yeah. And it wasn’t that it had, there was gradually more and more and more disquiet about it. It was that suddenly clip got shared suddenly that clip went viral and then it became kind of a scandal overnight.

Tom

So looping it back, I guess. I, I think my, my key takeaway here, which is unless you’ve got, you know, your number one in your, in your category or your, a global brand, you know, with the availability or the brand recognition of a Coca co or, or a Microsoft understanding the kind of the peer groups or the social networks where you have traction is actually a big advantage. You know, if you want to get as big as a CocaCola or Microsoft, you’ve kind of got to start by understanding these, the smaller social networks now, and of a eventually they will, they will spread and connect. And, you know, one day, if, if, if, if it aligns, then you’ll be that, that big, but actually just leaving it to random chance and, and, you know, spreading your message everywhere, puts you at a disadvantage if you’re really trying to be a, a kind of super large successful company.

Ed

Yeah. I, I think that’s, I think that’s a fair summary. I think that what people can sort of take away and think about are those, those sort of two methods by which this competitive advantage comes about. So that’s availability looking at, do people see your product around, and that comes from being kind of accessible in the first place. And then on top of that, the, the attractiveness and that idea of, of how, you know, surveying their friends to see how popular it is. If you can, if you can align those two things, then you can give your product the best chance. And then finally, the other thing is we talk, you talked about, you know, leaving it up to random chance. If you wanna roll a six on a, if you think of it as rolling a six on a dice, the more times you roll a dice, the better chance you have of getting a six at the same time, don’t, don’t read too much from the success of a campaign or of a sales tactic or of, you know, whatever you do, you can get lucky as well as getting unlucky.

Tom

I think that’s a really interesting point to, to, to kind of leave it on, you know, we as humans, we, we seek patterns, but yeah, actually it could just be random. Don’t, don’t look for that pattern where there’s not always one,

Ed

You know, you could, you could get led, led in the wrong direction quite quickly. I think,

Tom

Um, I think that was fascinating. I hope everyone enjoyed listening, um, share it with, with, um, with people and we can almost conduct a little experiment for ourselves. You know, the more people you share this with let’s test, if it, if it grows even quicker for us, because that’s really what we’re talking about, the availability of, of our podcast. So yeah, with that, we’ll see you in two weeks, uh, say goodbye, ed, goodbye.

 

Friends in conversation | Herdify

Sign up to the Herdify newsletter