Measuring happiness: AI and mental health with Nic Marks

Listen to this episode:

About this episode:

In this episode of Behind the Data, Matthew Stibbe interviews Nic Marks, Founder of Friday Pulse and author of 'Happiness is a Serious Business'. They discuss the intersection of AI and mental health, the importance of measuring happiness in the workplace, and how organizations can enhance employee experience through data. Nic shares insights on the significance of happiness as a metric, the challenges of identifying causes of happiness, and lessons from Indigenous cultures on measurement. The conversation concludes with a discussion on the misconceptions surrounding workplace happiness initiatives.

Timestamps

00:00 Exploring Therapy AI and Mental Health Support
04:38 Introduction to Friday Pulse and Measuring Happiness
11:58 The Importance of Happiness in the Workplace
18:09 Identifying Causes of Happiness and Frustration
23:30 Lessons from Indigenous Cultures on Measurement
29:44Conclusion and Insights from 'Happiness is a Serious Business'

Key discussion topics

Therapy AI can provide valuable support for mental health: Exploring how AI can contribute to mental health support.
Happiness is a key indicator of productivity in the workplace. Friday Pulse measures employee happiness on a weekly basis to provide actionable insights.
Happiness metrics can guide team leaders in improving workplace culture: Organizations often overlook the importance of emotional well-being. Bad systems can hinder employee performance and happiness.
Simple measurement scales are more effective than complex ones: Indigenous cultures offer unique perspectives on measurement and data.
Happiness is a serious business that impacts organizational success: Effective communication and appreciation are vital for team morale.

Transcript

Matthew Stibbe (00:00)
Hello and welcome to Behind the Data with CloverDX. I'm your host, Matthew Stibbe, and today I'm delighted to welcome Nic Marks, who is author of Happiness as a Serious Business, available in all good bookshops, and founder and chief statistician at Friday Pulse. Great to have you on the show, Nic.

Nic Marks (00:19)
Really nice to be here, Matthew.

Matthew Stibbe (00:21)
Well, so before we dive into Friday Pulse and measuring happiness and all that wonderful stuff, I like to start these interviews with one common question. What emerging technologies or cool things with data and analysis have caught your eye recently? What are you geeking out about?

Nic Marks (00:38)
Okay, I'm geeking out about something which is kind of niche, is therapy AI. And I've been trying it because I'm interested in creating a tool for managers that uses data on happiness. And I think they don't have enough support. And I was thinking I'd like to create a coaching AI for them based on the happiness science of happiness that they could put in. got an issue or something they want to work with. It can be confidential. Haven't really worked it out.

And someone said, well, you should try some of these therapy AIs. And I trained as a therapist when I'm young and I'm like, you know, why can you do therapy with AI? And do you know what? It's not the same, but it's really good. And because it's available 24/7, so you don't have to have your booked hour, which has advantages and disadvantages because... but if you're thinking about something like teenage mental health, how are we ever going to be able to afford that to really do it? And we don't do it.

And so I can imagine a future where you have like a combination of group therapy for teenagers or for anybody, and you have an AI therapist in between. So I was experimenting with that and, you know, I've actually found it genuinely useful as a reflective sort of process. I'm very into that as a process anyway, but I haven't been in therapy for years because life is okay, but just, it's like a diary almost, and you can talk about it. And I, I'm actually quite astounded how good it is.

Matthew Stibbe (01:56)
I can imagine, and I've seen a few of these out there. There's one, well not a therapy thing, but I've seen an AI tool where you sort of upload all your stuff, like your letters and your emails and I don't know what, and it kind of creates, or blog posts or books, Nic, and it creates like an AI avatar that is you as an AI based on the material you provide to it. I think that's a similar thing of just like suddenly some new things are possible that were.

not possible for, maybe it's good, maybe it isn't, maybe they're usable and maybe they aren't, but I could totally imagine. And there's such a long waiting list for mental health support in this country, everywhere, it would be if you could do something that combined group and well-trained AI and occasional or regular human contact, it's got to be better hasn't it?

Nic Marks (02:48)
I think so. that's, you know, my, uh, my logic. We're wanting to explore it for Friday Pulse. You know, it could be great. I mean, yeah, it could be a little avatar of someone with long hair and a beard, you know, um, and I, I haven't absolutely looked into the details of how we would create that, but someone said to me, this will be much cheaper than you think it will be Nic. You know, I think of it as an expensive thing because I'm not a techie in that, in that sort of world. So I think, that's magical.

But just using this therapy made me realize actually how powerful it could be. And of course it's, you know, I'm paying $19.99 a month. And if you think about a therapy session, it's normally 80, a hundred pounds, you know, and you'd be having four. So it's sort of 5 % of the cost, you know, it's interesting.

Matthew Stibbe (03:33)
Could you imagine

With Friday Pulse, listeners, we'll get to what that is in a minute, sort of somehow surfacing the data about your company and Friday Pulse into an AI sort of business or HR coach, for example, like with an MCP connector or something geeky.

Nic Marks (03:49)
I imagine that could be where we're going. It's still just the germ of an idea at the moment, but my real shock as someone who has trained as a therapist was, it's really quite good. It's a little bit soft on me. It's like when you go to therapy, you normally get into more challenge and you are definitely missing some of the human human. And also I know how to extract kind of what I want from it. So I don't know how a total novice is, but I expected it to feel really ropey and it...

Matthew Stibbe (04:13)
Mm.

Nic Marks (04:18)
it does not. Feels good.

Matthew Stibbe (04:20)
Yeah,

I think this is a fast-moving space. Interestingly, everyone on this podcast that I've asked that question to so far has come up with some variation of we got to talk about AI, haven't we? So it's a thing that's exploded into our world recently. So now tell me about Friday Pulse. What is Friday Pulse? What does it do? How does it work?

Nic Marks (04:38)
Yeah, it is.

So I used to work in government policy about how governments measured population wellbeing. I worked in Think Tank World and I ended up having really quite a commercial idea, which was, I'm just as critical about how organizations measure employee experience as I was governments measure population wellbeing. And could you create something for organizations to use to get a better feel, better read ⁓ on how employees are really doing? And so...

That was the nub of it when it started. I mean, I first did some statistical work on this, I guess, 2006, but I set up the company 2012, which has morphed and changed and all sorts of things. And it's become Friday Pulse and Friday Pulse is called Friday Pulse because on a Friday we take the pulse of teams and on a Monday we feedback that data. So I have an expression that feelings are data. And basically we...

We convert how people are feeling into numbers so that organizations can take them seriously.

Matthew Stibbe (05:48)
And I saw on your profile, one of your contacts or ⁓ people said happiness is a metric worth measuring. How much data do you have to get to be able to measure happiness?

Nic Marks (06:01)
Well, I mean, if you ask someone, how happy are you? That's a data point and it's probably very relevant to how they are, so if I asked you, okay, we are recording this on a Monday. So think back to last week. How happy were you in your work last week on a scale of one to five where that's very unhappy, unhappy, okay, happy. How were you last week? Yes.

Matthew Stibbe (06:28)
Me? probably

The lower end of that, maybe a two. Some very stressful things going on around with my building work.

Nic Marks (06:35)
Okay. So

Yes, so that is meaningful to you. And that is a data point. Now there will be some randomness around that because you probably haven't really settled on where you are in the scale. But if you ask that repeatedly every week, you'll kind of settle into my set point is happy or my set point is okay. It was a worse week. It was a better week. And you move up and down the scale. And so then you've got something that is meaningful. So

I think this can work on quite small scales. When you go to populations, you obviously... into sampling methodologies and how you do that. And typically in the UK, I might do a 1500 person survey in the US. I probably want two, three thousand because it's bigger. And then you get something that's within two to three % statistical significance. So that becomes when you're serving a sample. But in an organization, if you're serving everybody, really the randomness comes in in response rates. If you have lower response rates, you've obviously got...

data is less reliable on the higher the response rate is. So actually I think one data point is meaningful.

Matthew Stibbe (07:38)
Okay, so I know I'm aware of the Edelman Trust Barometer that measures at a population scale, sort of trust in government and various other things. Are there reliable, regular, national level surveys of happiness?

Nic Marks (07:46)
Yes.

So in the UK, the Office for National Statistics collects data on wellbeing. It started to do that in 2012. It's actually why I left the Think Tank world because Cameron announced he was setting up this unit and we did a little bit of advice to them. And I thought, well, I've achieved my policy goal. I didn't really know how policy was going to go over the next 10 years, which was quite different than what we were expecting in those sort of happy days. Felt quite optimistic. And...

So actually the UK is the first government that's really committed, well, the second government because Bhutan had a measure before that. But ⁓ the first Western government to really commit to do it. And we do have data on that. And I don't think it's influenced policy enough because policy seems to have got very much into an emotional story around, know, and the world has changed. But it is there and it is interesting. It is used sometimes within policy and they gather data.

I mean, I would be slightly, I'm so in the field that I'm critical of what they do, but basically it's 80, 90 % good.

Matthew Stibbe (09:01)
I used to design computer games for a long time and ⁓ I had to, after I sold that business, I couldn't play computer games for years because I was so critical of like every little, why didn't you do that? That didn't work, that graphics, you know, just being, and now I can enjoy them again, but it took a long time. So maybe, maybe happiness is gonna welcome you back when you're ready. National happiness, so Bhutan measures gross national happiness. ⁓ How?

Nic Marks (09:12)
Yes.

Yes.

So look, I worked with Bhutan, the Bhutanese government. Well, I mean, I can't say I worked with them. I was one of a group of Western people interested in this area that did work with them. I'm not an advisor because they didn't really take my advice. They've ended up with quite a complex measure.

Matthew Stibbe (09:51)
If they don't take your advice, you're a consultant.

Nic Marks (09:53)
Okay,

I'm a consultant, perfect. And I was advocating for a much simpler one and they basically have a long questionnaire which they ask about well-being and happiness on eight dimensions. And some of those are super interesting, but statistically for me, they make the error of mixing up drivers and outcomes. And I'm very, very keen to have a clean outcome measure. And then you look at the drivers of that outcome measure. And I wanted them to have a clean outcome measure

of a very simple question like, ⁓ happy are you? Now the issue with Bhutan is that people are very compliant. They're quite subservient to the government. So as soon as something is a government measure, they're probably gonna say they're really happy. So there will be all sorts of difficulties about doing that. they went down a route where they asked like lots of questions, but it became to me totally ridiculous because they would go out into a village which might be three hours walk from a road.

And they would ask the head of household a questionnaire that took two hours to administer. And I don't believe they were collecting robust data. I mean, I've worked on big surveys like the European social survey where the questionnaire is an hour and a quarter, but my goodness, are people trained on how to do that and the level of methodology. And they are sending out people who aren't trained to do a questionnaire with people and they're looking to tick boxes. I don't believe... there's too much error in the data for me. And it's also too detailed. They get lost in it.

And I would have much preferred they did something like they got all these regions. would say that 58 % of people are happy in this region. 45 % of women are, 65 %... whatever the things are, farmers are like this. What will be is the rural poor in Timpu, the capital, will be less happy than the rural poor out in villages. That's going to be certain because they got less access to resources. So they could have had some very, very simple data which opened up

the important policy questions. And I think they just got sort of paralysis by analysis are going too deep into stuff. So I'm a very pro simple data when we're collecting it on subjective stuff.

Matthew Stibbe (11:58)
And what's the secret of asking the right questions to keep the data simple but measurable?

Nic Marks (12:07)
So I asked you what my favorite question is, which is how happy were you at work this week? Notice I've done two things. I've asked you how happy you are. So it's an emotionally tone question. I've asked you, I put it in a context at work. There will be some bleed into that anyway. If you're unhappy at home, it's going to come. And then I've time bound it, which is each week. And time bounding is really for me important because that creates a very dynamic indicator. I'm not against the surveys which ask people how happy are you at work in general? I think they're useful.

If you're going to do an annual survey, you want to do that because people won't remember the year you can get a general. But I think what's much more exciting is the much more real time data. Week is a really good compromise work because we work in week sprints. You absolutely could measure it daily, but you'd have a problem with response rates. I have tried it. We tend to find Tuesday is the least happy day of the week because it's kind of psychologically the furthest from the weekend. But, whatever it is interesting daily. If you're working very fast shifts,

you could use that data if you're going to have a stand up every morning. I'm very keen that people talk about the data if they gather it, otherwise it's just black boxy. So a week is a nice rhythm to work with, but time bounded works. You can do a month, but that's about the longest time bound you can do, otherwise you forgot what's happening.

Matthew Stibbe (13:20)
I can barely remember what I had for breakfast this morning. Okay, so why does happiness matter in the workplace and what data do you have to support that assertion?

Nic Marks (13:22)
Yes.

So happiness really matters in the workplace. And you know, I'm using a word which is a little provocative, but it's the one that people relate to. So you could use a really dry word like wellbeing, but if I asked, was your wellbeing last week? You'd be confused. Whereas happy, everyone can answer. So that's very good. It's emotional tone. It's basically a reflection on how well the week. And the reason for that is that from a sort of evolutionary and biological...

perspective, happiness is a good bad signal that when our inner and outer worlds are aligned, we feel good. When they're misaligned, we don't feel good. So what you're looking for is that alignment between the individual and the environment. And then you can immediately see there's going to be a lot of value for an organization there because people's internal motivation is what you want in the context of the external goals that are going on. And happiness allows that to happen.

Now it's not only a signal of good bad. So when I'm asking you how happy were you at work last week, I'm picking up that good, bad signal and I'm putting it statistically into a one to five scale. But also when we feel good, we do good work. And we know that, we know that, it's energy, emotion, the word emotion is in move in, they move out. And so you're trying to...

align people's motivation with the business goals and you're trying to help them do that. And so if you help people feel good, they're going to do better work. And what does that come out with statistics? Well, happier teams, happier team members are 20 to 30 % more productive. And that's across all industries. It's the more creative a role gets, the more collaborative it gets, the more customer facing it gets, the more that happiness matters. If you've got people...

on the end of a telephone speaking to somebody that they're only going to speak to once in a sales process, it's going to have a smaller effect. And in fact, we know that from the data, there's a big study of BT call operators and they showed, it's a really good study because it's really well designed, but they showed that happier operators sold 10 to 12 % more in a week than the unhappy ones. That is actually a small effect, but the great thing about these call operators, they've got no teamwork. They've got this very, very clean data.

So actually you can work with it really well statistically, but it's not a real office where, or real work where we tend to be collaborating. And you think about it, it's a pain to collaborate with people who are painful. You know, it's like, you know, and it's great to work with people who are positive. Now that, you know, there's limits to that. You don't want, you don't want so optimistic that you're delusional, but there's, you know, there's a really good sweet spot that most people work in and, and that, that's what you want to achieve.

Matthew Stibbe (16:10)
So there's always this question about the difference between correlation and causation. And if I'm getting this data about, my team or my colleagues happiness, how do you identify what might be causing either up or down with movements? I mean, how much control, how do I turn that data into something I can do something with?

Nic Marks (16:15)
Yes.

Yeah.

So, behind your question is it's very difficult to prove causality. And actually that BT study I mentioned is the first published data that proves causality in an observational world. So it's been people that done it in the lab, they've manipulated people into good moods, other people into bad, they've shown the people in the good moods, come up with three times as many creative ideas or quicker at problem solving or do more output. So there's been this lab-based data for, since the 80s, 90s.

In fact, some even earlier than that, but really good data on that. But getting that... an observation has been really hard. The BT study did it and they did it in a really clever way. So they work in BT call operators, they looked at how the weather changed. And so they saw that on sunnier weeks, the operators felt sunnier and they sold more. And the causality can only run that way because you can sell more and feel happier, but you can't, you're not King Canute.

Well, he couldn't do it, he? But you're not in the process where you can make the sun shine. So that is the first causal proof of it. I've got lots of inter-temporal correlation coefficients which show the same effect, but it's strictly not proof of causation, whereas theirs is a strict proof of causation.

Matthew Stibbe (17:46)
That's interesting, sort of the validity of the assertion, but actually I'm thinking, let's take the assertion that happiness is good for work as a given. How do I know what I can, if I see my team's happiness declining, how do I figure out what it was, I suppose, ⁓ or how do I figure out what to do to make it go back up? Do you see what I mean? How do you...

Nic Marks (18:09)
Absolutely Yeah. Yeah.

I mean, there's two main things. And the first is their daily, weekly workflow. What is going on in there, right? I've got a chapter in my book, which is called Bad Systems Beat Good People. And daily frustrations, weekly frustrations hold people back. Those frustrations don't have to be IT systems, but IT systems, admin systems are...

Matthew Stibbe (18:25)
we will experience that.

Nic Marks (18:35)
I've got an exact figure on a bit, something like 27 % of daily frustrations are about IT and admin systems, okay? About another 25 % are about interpersonal relationships, and then it gets into things like workload, but IT systems are the second behind relationships that are, and those are pretty tightly close together in a couple of percentage points, but I'm not even confident of the difference because that was in a 1500 person survey, so there's probably some.

Nic Marks (19:04)
Bad systems, you know, we know it's frustrating when we're interacting with a system, you know, and that could be a computer system or other one. It's probably the key way that you can first improve your team happiness is to remove friction. Because people like to deliver, we like to accomplish things. Actually, people would like to do a good job. And there are things that hold them back. And so you're trying to work. So in not only on a Friday Pulse, ⁓

Friday Pulse asks people that simple statistical question, psychometric question. We then ask basically in text boxes, what's gone well for you this week? What hasn't gone well for you? And then we feed that back in a learning data rich feedback loop to the team leaders on a Monday morning, sometimes Tuesday morning, depends what they choose, but basically the beginning of the next week to have a stand up, right,

you know, our happiness was okay last week, let's talk about what's gone well. And I always say psychologically, talk about what's going well first. Don't skip straight to the problems. If you go straight to the problems, you have a very different team meeting than if you go to the things that are going well. So we all naturally as human beings like appreciation. And so call out your team for what they're doing well, find out what's going well for them and build on that because that is their energy. That's when they feel good, they're playing to their strengths and all those sorts of things. So

that is what you do first. And then when you've got that good vibe in the room, what's frustrating, what's holding you back, what can we do to remove that friction? So I say really, there's two questions. You should ask your team every week what's going well and what isn't. And then you should get into a habit, a repeated habit, which is appreciating each other. So thanking each other, doing the human thing of saying, thank you, Matthew, for the support that you gave me for this, or, you know, or well done, Matthew, for doing that. And that can sound really trite, but if you were a team leader and...

Let's face it, most team leaders have had no training in people skills. 80 % of team leaders, line managers, supervisors in the UK have absolutely no training in people skills. They've been promoted for technical or length of service. And how are they supposed to... they've been thrown into something which is very complicated. People are complicated. Relationships are...

Matthew Stibbe (21:19)
Can you imagine if people got promoted and get put into jobs doing some data management because they were friendly and optimistic and got on with their team? I mean you wouldn't dream of it but people who are good at data suddenly become data team leaders or something and as you say no no or little training.

Nic Marks (21:37)
And it's probably their only way to promotion. You know, there's that old saying, which is that people join an organization, they leave a manager, leave a manager, you know? And one is that's probably true in my data, teams are three times more important for happiness than organizations. Either membership of your team affects you three times more than a membership organization. So you can go to an organization which is relatively happy. You could be in a rubbish team. You can have it the other way around basically is what it's saying. So it is true, but actually it's an organizational failure.

because they promoted people without thinking about their people skills and they haven't trained and supported them. So Friday Pulse is basically my statistical way of helping team leaders by giving them the data every week that they can have a sensible conversation about the teams and it guides them into, right, this is how we're feeling and that's scary Pandora's box. If you don't know anything about emotions and feelings, I mean, I'm trained as a therapist, so it doesn't scare me, but

having a repeated time when you can go and talk about how things are in the team. Like I suggest a half hour, 15, 20 minute meeting every week. So if you've got frustrated about things, don't let that, you don't necessarily, because you go, I'll speak about that on a Monday standup. So you've got what's called a contained space in the therapy world, but you've got a contained space for it. The team leader comes into that meeting with a bit of data so they know what's going on. So they're not so frightened. And then also just build on those positives and then...

compassionately look at the frustrations of what can you fix? There might sometimes be things you can only lend a friendly ear to because the system is out of your control, but even sharing a problem helps. So it's just creating that data that people ⁓ can do that and helping and supporting team leaders. It's a difficult job being a team leader. Relationships are tricksy. And this is to help them do that better.

Matthew Stibbe (23:30)
There's one other area that I wanted to discuss with you. It sort of came up a little bit in our pre-show conversation and you were talking about what we can learn from Indigenous people about measuring, about numbering systems and I just thought that was fascinating and tell me, for the listeners, tell me that story again and then we can explore it a bit.

Nic Marks (23:43)
Ha ha.

I

read this in a book called Alex's Adventures in Numbers or something like that. It someone reflecting on numbers and he was a numbers geek. I, you know, as you know, I went to Cambridge and did maths. That's my natural habitat. Though I'm not actually a mathematician. I didn't like it, got too pure. I'm a statistician. I like working with numbers. And I read this thing and I never thought about it before. I've always liked five response code. I've always thought it's really simple. Certainly an odd one. I work with seven ones, but I like an odd response code.

And he gave me the reason why short code is better, which is that indigenous tribes have no words for numbers over four five. So they'll have one, two, three, four, many. And birds do this too. They can't really count very high. They can just see they've got many. Now what humans are really good at is doing more or less. So if there was a bowl of fruit with 12 apples in it, another one with eight, you'd immediately see.

Matthew Stibbe (24:36)
Hmm.

Nic Marks (24:49)
that the one with 12 had more. That's instinctive, that's like going to a bush, how many berries has it got on? There's a good evolution of reasons why we'd see what's more. However, if I asked you how many was in them, you'd guess or you'd have to count them individually. And so, whereas if it's four, you can immediately see it's four. So when we're using response codes, people use these naught to 10 response codes like the net promoter score, how much would you recommend this product if someone's naught to 10? We feel naught to 10 is good because we've got 10, base 10 numbers.

All sorts of things like that. Actually, subjectively, we've only got base four or five. Think about it with bushels when we count and you see it in a cartoon, someone stuck in prison, how many days they go one, two, three, four, cross it out. Because we can see four, you know, it's why Roman numerals went five suddenly became a V. In fact, even Roman numerals, I know they put IV for four, but actually if you look on a clock with Roman numerals, it tends to be four ones. They've gone back to the more ancient signal at the bushel.

So I think very, I have a very strong opinion on this and that sets me out even from people who do subjective work. I don't like nought to 10 scales. I think five is simple. And when I do my data, it's also really simple because you can name them very unhappy, unhappy, okay, happy, very happy. And that means that I know these people are okay because they've ticked the box that says okay. And part of my thesis around this is okay isn't really okay. And okay is low energy. So happy is you've got...

Matthew Stibbe (26:08)
Yes.

Nic Marks (26:17)
things are good, in and out, world aligned, you've got energy to go forward. Unhappy is a signal to change. So actually you've got energy for that. Okay, it's the low energy spot. And if I was gonna diagnose most of the problems in organizations, it's they got too many teams stuck in okay.

Matthew Stibbe (26:35)
Yes, and it also I think when you've got a scale of zero to ten there's a sort of decision fatigue around it. I can't really differentiate, I have to engage some mathematical part of my brain to go seven or eight, I don't know seven or eight.

Nic Marks (26:49)
People don't use the whole scale. Actually

when they... I've never done this observational study, but I read about them where they're watching people answer. And if they've got a whole bank of noughts of 10, people get to this one eight and they go, maybe I should change that one back. You know, it's, it's, it's non-instinctive. Whereas one to five is instinctive. just go, yup. You know, so, so, ⁓ and you want to work with people's instincts because emotions are, they're very bodily. They're not precise one, zero variables.

They are feelings, they have different strengths to them and we can differentiate between those strengths but don't try and put too much specificity.

Matthew Stibbe (27:28)
This is this applies to not only emotions, I think, but actually sort of subjective human judgments that you want to capture in data like, for example, sales team assessing the probability of a deal closing this quarter. Right. You've got zero to 10 or, you know, zero to ten or one to 100 percent. I mean, but if you go, OK, it's definitely going to close this quarter. Probably. Maybe. Definitely. You know, you've kind of got to land somewhere.

I think this is a really interesting concept for people who are listening, they're trying to capture all kinds of data that looks, or they want it to look, scientific and objective and data-driven, but actually is involving some human judgment or some human feeling or subjectivity.

Nic Marks (27:57)
Yeah.

Yeah, I mean, normally what happens anyway is people then tend to categorize those variables to understand them. So they go highs above 70%. I mean, just go straight for that. I mean, I'm quite keen on three point scales. You know, is it, know, is it, is it, you know, hot, warm, cold, you know, that's, that's not bad for a lead. And it's probably as much data as you've really got. And when you, I mean, when I present graphs in the book that I, although I use five points in the data collection, I only show three, which is happy, which is the happy and

Matthew Stibbe (28:26)
Mm.

Nic Marks (28:44)
very happy together, the okay and the unhappy. And because that's, that's what we can understand when I grow up, you know, I don't want to make it too complicated. And, you know, and I just present it like that. And then you can say the happy, do this compared to the okay, the okay. And it allows me in my book to bring a few bits of data together on productivity and staff retention and show that a happy employee is worth twice as much as an okay employee over their lifetime with you

because they're more productive and they stay longer. So you get a multiplication effect like the lifetime value of a customer, you know, will be how much they're spending and how long they stay. It's the same the lifetime value for an employee. You've also got the cost of acquisition with an employee. You've got the thing. So it's the same metric as lifetime value of a customer, but for an employee and happiness is 2.2 in my book actually precisely, but over two times more valuable than an okay one.

The unhappy they've gone before they've added any value.

Matthew Stibbe (29:43)
Well, now, so we're almost out of time. It's been a very enjoyable conversation, but I'd like to ask you one final question. Tell me about your book and what... Yes, give it a good plug. But what particularly can data geeks learn about happiness and happiness as a serious business?

Nic Marks (29:51)
Yes.

So the book's called Happiness is a Serious Business and it is the empirical case for why happy teams are more successful and it is also how do you build them. I have a particular data led approach to it. So Friday Pulse is based on a cycle which are called measure, meet, repeat. So that's my offering to the community about how you can improve happiness, it gets into that friction and flow. It does also go into culture

and what's the culture to support it. So, but it's very easily written. I've written it for people leaders at all levels really. ⁓ And happiness is fun and interesting. I mean, there's so much more nuance to it than a good bad signal, which I go into at the beginning. And I also go into the mistakes that organizations make. ⁓ And I've got a chapter called, 'Pizza, ping pong and parties don't work' you know, cause they, you know, and I'm not being the Grinch, but they're going to have a little effect

in that week and then in a day, but it's no good stressing people out and then giving them pizza. You know, you've to get into instinct. So, so it's a, it's a fun exploration of that. And, it's for data people, it's got some data in it. And for some, might not even go far enough. I mean, I I'm trying to write to an audience there, but, all the references there, if you want to look at it in more detail.

Matthew Stibbe (31:24)
Fantastic. As I said, available at all good books, bookshops. Well, yeah, well, yeah, true enough. I just on that last point, I back when I was a journalist, I used to go and visit all kinds of companies and they would take me tours around their facilities and they would show me the ping pong tables and the gyms and the fancy canteens. You know, for the most part, I never saw anybody in there.

Nic Marks (31:26)
Well, actually only available on Amazon. That's where 70, 80 % of books are anyway.

Matthew Stibbe (31:52)
one of these, certainly never, I was taken to a very big famous well-known company, no names, and they showed me this amazing gym, huge, nobody in it. And I'm thinking, well, you know, that's just like, that basically, that serving as a token to put on your job recruitment, we've got a gym, but nobody's got time or energy to use it. So anyway, yes, pizzas and ping pong don't work. Well, on that bombshell, Nic, it's been delightful having you on the show. Thank you so much.

If you're listening to this and you'd like to get more practical data insights and learn more about CloverDX, please visit cloverdx.com forward slash behind the data. Thank you very much for listening.