Meet Mats Einarsen: the man who wants to make the Internet 5% nicer
Some time ago I did interviews with people that inspired me. And then I didn’t feel like it anymore. But when I recently saw what my mentor back in my Booking.com days, Mats Einarsen, was up to, it was time for yet another interview.
That is very noble.
If you say so!
You worked in industry for a while now, combining behavior science and data science. Making a lot of money for companies like Booking.com, OpenTable and Kayak. Did you go soft on me?
In short: yes. The longer story is that I wanted to take some of the way we think about conversion optimisation and apply it in a different domain. By that I mean being goal oriented, tactical, metrics-driven and research-driven. So I wanted to do that but look at good and bad behaviour instead of revenue. To my surprise, I have to say, when I started looking I realised there is a lot of research on prosocial behaviour and it’s very inaccessible for non-academic practitioners.
Compared to what?
Well, I’d say particularly compared to things like nudging and cognitive biases, which has a lot of popular books associated with it.
Richard Thaler even got the Nobel Memorial Prize recently…
Are you using some of the money you made to make the world a better place?
I guess I am. But it’s not that expensive to read and summarise psychology papers.
“After the Trump election, when shitty behaviour really started coming out of the woodwork”
Let me rephrase that… Why Hacking Kindness? And why now?
So I started this about two years ago a bit after the Trump election, when shitty behaviour really started coming out of the woodwork, and on social media in particular. I’m not really thinking about the politics of it, but there seemed like there was a watershed moment around then were online and offline harassment either became a bigger thing or it became much more talked about.
It probably was always there.
Agreed! But, so I started looking into what we can do, as the builders and shapers of the Internet, and this is the outcome so far. However, since two years ago, I think we’ve all realised a lot of the bad tone online is either artificially generated by very powerful players who will not be influenced by nudges, or it is flat out caused by outrage as a conscious growth tactic by what the Americans call pundits and in some cases politicians.
So Twitter was the use case you had in mind while creating it?
Twitter, Facebook and Reddit. But due to the reason I mentioned, over the last year I’ve fanned out a lot, from trying to improve my 4-year old daughter’s behaviour to imagining how to keep internal mailing lists civil. The funny thing is Facebook and Reddit now is as much a source of insights than a target case, as they do some real research that they make public. As much as it’s been criticised, the infamous Facebook emotional contagion experiment has a very valuable and actionable message on how you can alter the emotional state of your group, for example.
Do you think the reception that study got had a negative effect on what companies like Facebook share? Has it affected your work?
Hugely. And it’s a huge loss. I don’t think people realise how little incentive there is inside large companies to share insights like that, and having seen Facebook’s PR disaster with their paper, everyone now defaults to not sharing anything. So the consequence is we now get a lot of papers on data science methodology and very few insights from the inside on how these large networks actually affect people.
Influencing behavior is incredibly difficult. How do we know Hacking Kindness works?
Is it that difficult? It took you, what, 2 minutes to talk me into doing an entire interview. And you’re not even paying for coffee! I’d say that was quite easy.
Ok, so, for every concept I’ve included, I usually looked for a peer-reviewed systematic review documenting multiple independent experiments on the concept in question. I really don’t want to get the rug pulled from under me because some replication study invalidates all my cards. And that’s not going to happen. However, I’ve also included some concepts that are either very interesting, particularly strong or that just seemed really worthwhile for other reasons. The Facebook social contagion effect is an example of that, and Mimicry is another interesting factor that maybe doesn’t have as much evidence.
“With any behaviour intervention, you always, always, always have to validate that the outcome is what you intended”
And mimicry is also closely related to social contagion, isn’t it!
Good point! Anyways, with any behaviour intervention, you always, always, always have to validate that the outcome is what you intended. The effect might be ever so real, but your implementation in your field can always mess it up. That goes for everything, not only Hacking Kindness.
Do you think companies can make money with making their customers behave better?
You really bring up money a lot. Do you feel the concept of kindness is very challenging to making money? If so, maybe we should talk more about why you feel like that?
Why should commercial companies care about kindness?
Ok, why should they care…
So let me give you two different approaches. The first is very pragmatic. A lot of companies have very practical problems around kindness and prosocial behavior. A retailer will have problems with theft, for example. Or with customers abusing the frontline staff. Or with their own staff mistreating each other. Or with getting people to sign up for sustainable alternatives. I could go on with very practical, costly problems related to kindness. Moving the needle on these problems can be measured with your normal money metrics.
The second is less pragmatic and more strategic. Look at how Lyft can round up your price and donate the remainder, or AirBnb’s OpenHomes program that let you offer your home in natural disasters. Now, they are compelling programs in their own right, but they also add this intangible value to the basic price proposal that is harder to think about in straight money terms. It’s one thing to walk away from AirBnb to Booking.com for a lower commission, but it’s an entirely different thing to say to yourself that you are going to stop being that person that offers up your house when there’s a wildfire!
There’s also the element of what’s called the Helper’s High. Doing kind acts releases different neurotransmitters than what you get from rewards in typical gamification loops, but they still make us feel good and incites us to do more kind acts. And then you get a feedback loop, potentially around your company. But this is really more Growth Hacking Kindness, which is an entirely different thing from Hacking Kindness. Ultimately, kindness builds retention, and that’s the essence of a successful business.
“Doing kind acts releases different neurotransmitters than what you get from rewards”
And the companies should use your cards to figure this out?
They should use the cards to figure out how to present it to their customers.
What about a kinder society? Do you see applications in the growing digitalization in cities? And what about the analogue city?
There’s one thing that I have really appreciated since I made the cards available online. People seem to see solutions to problems in them that I never imagined. I’ve heard from people working on mental health issues, managers, people in education, and indeed city planning. I did not think about that, and my first edition of these cards were aimed at online systems. However, I realize that is too restrictive. When it comes to digitalisation of cities and how we live in cities, I think the tactical kindness approach can definitively have a material impact. But I don’t want to make any statements as to how as I’m not very familiar with the topic.
Lastly, what’s your dream with Hacking Kindness? What will be the project in 2025 that you will be most proud of?
My dream is to make a dent in the Internet [laughs]. To be specific, my goal is to make the Internet 5% nicer. It feels doable, doesn’t it? It almost sounds like too little, but it would make a difference and if we can figure that out, the rest should be easy.