100. NOLL TILL 100 LIVE PODCAST

If you set aside Mockberg and Såpa, which have already passed a hundred million, and the podcast is called Zero to One Hundred, then Twistboxes is the next zero-to-one-hundred journey. Over the last fourteen months, we have grown the team to about fourteen people.

You sit in front of Excel looking at numbers, so when you see that there are real people doing real things for real. I think as a company we still have not really reached the point where we get enormously more out of things with AI. But I think that will change this year, because I really believe this is the year when, we seriously believe this is a one-man show, a company where, I mean, one person will not do all the work in the world, but one employee in the company together with a group of skilled specialists and generalists around them, supporting at different times, will [music] be able to build a billion-krona company.

A warm welcome to a new episode of Zero to One Hundred. My name is Jacob Wibom Westerberg, and as usual I have Fedja Porobic with me. Thank you, Jacob, and this time is a little special. We have a few more people with us. Do you want to explain?

Exactly. There is not just one secret guest in the room today, there are probably seventy-five secret guests in the room today, sitting and watching us. [applause] Fantastically exciting. Perhaps we should start doing this every week, doing the whole podcast live from [laughter]

Yes, that would be crazy. How does it feel? No, but it was, it feels very good. I am excited, as I usually say. New week, new start. We have, so much is happening on every front in all the companies, and this is no exception. It has become more of a rule by this point.

You seem to have been very much in the spotlight over the past couple of weeks. First there was that Google event together with Relevance, and then you spoke at Techarena. When was it, at the end of last week? Thursday.

Thursday. It feels as though the theme, if I paint in broad strokes at least, the theme of both occasions, I have not heard either of them, I should say, but the theme of both these things has been somewhere around story brands and authenticity, which will also be a discussion we have today. And why do you think this is in the spotlight right now, so to speak? No, but so much has happened on the AI front, and not least, we talk about AI's progress and how to apply it in our companies, and what we do with agents and how far we have come. But somewhere, if you just fast-forward a year or two, I still see that with similar tools there will be a level playing field. And in an

AI landscape, there becomes a sameness everywhere, so what will emerge? Which brands will benefit most? And I have concluded that it is the companies best at building their story brand, where people have a deeply authentic feeling in front of the camera and are good at bringing it out. And we humans tend to follow people rather than companies. So it is quite helpful to have another human in front of the camera and build a story around a person. That is usually much easier. So I have talked about how authenticity will be the deciding factor for companies in an AI-driven world.

Cool. Big things, really. I wrestle a little in my head with whether the whole AI rollout benefits small companies or large companies, because in a way big companies always have better bandwidth to execute. At the same time, you could imagine a small company that is a little more agile somehow gaining ground in this environment.

I do not know. What do you think? That is a good question. I think we small companies will benefit, [unclear in the Swedish source: hståndsmässigt], the most, of course, in a way. We will be able to take advantage of everything new that comes along very quickly. Big companies take a long time to adapt, but they will benefit too. We just see what is happening around the world, especially in the US, with companies refusing to rehire people when unemployment is very high. Despite that, the economy has never done so well, they say. I do not know. But it can happen.

Before we perhaps leave that behind, what was Techarena like otherwise? Zlatan was there. What did he talk about? No, it was completely crazy. If anyone knows how to build a brand, it is Zlatan. He mentioned he got into Vitamin Well early as a co-owner, and said that a few years later, when they did their first partial exit, he was like, jackpot, now you sell everything. So he sold his stake, and then you fast-forward ten years and the company is worth, I think he mentioned on stage, thirty-three billion, and he was just like, oh, I sold too early.

There was a little upside left to capture there. Yes, there was, yes, but as I said, Vitamin Well has done very well. He talked partly about that, but also of course about his fine career, and then returned to Milan. The difference he really made in the dressing room, and how he got them from something like eleventh place to winning the league, and that was his most memorable title. It was a very good talk, so it was definitely Techarena's highlight. But Techarena was buzzing like never before. I have been there for two years now, and it was absolutely fantastic. There were so many people. The whole ecosystem feels very alive. It feels as though things are bubbling everywhere. Not least in, what do they call it? Silicon Valhalla.

Yes, that is right. Yes, the new name for Silicon Valley is Silicon Valhalla. Boris Johnson was there. What did he talk about? Yes, I did not hear him, because it was just before my keynote, so I was worried everyone would be listening to Boris Johnson. My stage would be completely empty. I was sitting in, I mean there were five times as many seats as here, and about ten people. And we were sitting there and I thought, I just looked, what the fuck, there is not a soul. Oh dear.

Okay, at least my company is here, some supporters. I will get a few cheers, they will come. So I was like, worst case, we have a group here. But then when Boris finished his talk, people started spreading out, and then the next keynote was on the main stage, and I have no idea what it was because I was doing my own thing, but then suddenly people poured in, and lots came. I think the title was something like what has happened after Brexit, and what does the AI news look like for the UK? But I was not there listening.

Things have gone so-so since Brexit, you have to say. But we do not need to talk more about that now. No, exactly. We started this podcast in 2023, in the autumn of 2023, sometime just before Black Week. And a few things have happened since then. That year you launched Västerbottenssåpa sometime in the spring. And then things happened with Mockberg, and quite a lot of other things have probably happened on your side too. What are the most memorable things you look back on from these past two and a half years? Professionally, we can start there.

Yes, exactly. We have to say professionally, because there are things that happened privately, but professionally, you mentioned Mockberg and Såpa. I would say the most memorable was when we opened in NK for Mockberg a year and a half ago, and that we got the space we got, right in the middle, with so much exposure.

It felt very big, coming from a company that just months before that, yes, was really on the verge of bankruptcy, to us opening in NK a few years later, and the journey that has been made. It was in the starting blocks when the podcast started, Elvira was one of our first guests, and we talked a lot about what she had done to rebuild Mockberg. And many things have fallen into place, so it is obviously not just one, but that was probably one of the single biggest events. Then I loved the first time I went up to Vilhelmina. I had never been there when we started collaborating with Västerbottenssåpa.

Mm. Shame on me. I have never said that until now. Now I am saying it. I did my due diligence on paper in a car, basically on the way to the office, and thought, these are cool people. People you want to align with. They have a good story. I believe in the category. Everyone thought I was an idiot. Nobody believed in cleaning. How can you, how are you going to do this? I know you were very skeptical.

Yes, but I was a little like, given that you had succeeded at a few things before that too, I was still thinking he has probably thought of something here, but it is not entirely obvious what.

No. And most people probably agreed on that. I think it was a long shot, but I fundamentally liked the product. But I had never been to Vilhelmina, so what I am getting at is the first time you land in, or Vilhelmina, by the way, this is becoming an ad for Vilhelmina [laughter], does it need a sponsorship label then?

You must soon be the biggest employer there, surely, if you are not... Yes, I know, we are so big there now, so it is a fantastic landscape. I love, you know, you land in this little propeller plane, at a little airport, and come out into this huge landscape, big mountains, and it is fantastic. So I thought, now I have to explore more parts of Sweden somehow, but then it has not happened. Went a little with [unclear in the Swedish source: egen kråka] and landed, but was welcomed at the factory.

But also, you know, professionally, when you come into the factory for the first time, nobody knows who I am, and I have been behind the scenes building the whole company. So nobody knew who, yes, so I could just sneak around. Yes, what do you do? Oh, so that is how it works in practice? You sit in front of Excel all day looking at numbers. So when you see real people doing real things for real, building things that are going to be sent to customers. But it was, I do not know, it was

It is quite an unusual D2C story, given that you have your own production. I mean, ninety-eight percent of all the companies you work with do not have their own production, but outsource production to a factory somewhere in the world.

Yes, it is probably very unusual, so it is more a production company than an e-commerce company. E-commerce happened to become our first channel, but now we are in retail, so we are going to roll out nationally soon, I think.

It is you and Hermès, they also have their own production. Do they? Slightly different ends of the retail spectrum, but okay, good. Being compared with Hermès is absolutely fine. I will take it. If we turn the question around, what do you think, since 2023? What has been your biggest memorable professional event? A few years have passed since then.

Yes, exactly. A few years have passed since then. Partly I have Juniper, which I spend quite a lot of mindshare on together with my wife Matilda, and the team is here in the room. Juniper is roughly twice as big as it was then, but relative to how much Mockberg and Västerbottenssåpa have grown, it is still very small. But it feels as though it is on a good trajectory and had a good Q4 this year. I think on my own professional side, the biggest development is really that we started [unclear in the Swedish source: Orange Shoes]. We made our first hire in January last year, an old colleague of yours, Victoria, whom I also had as, or she worked with me on PG's behalf when I worked for Juniper, and she started working

With us then, and today we hired two more people and became fourteen people. So over the last fourteen months we have grown the team to about fourteen people. So that has been a very big thing that has happened quite recently.

Completely crazy, really. Fourteen people, that is one a month. Exactly. One a month, and our ambition is to keep trying to hire roughly one a month until, well, we do not know, or for as long as it works, is really the answer. But one a month feels like a reasonably suitable pace in terms of being able to recruit and find work. But we will see, we are not doing any outbound yet, so everything has to come inbound. So if you hear this and then feel, wow, I really need help, perhaps writing to me on LinkedIn is a good idea.

Yes. No advertising label. [laughter] Absolutely not. That is good. No, but that is at least a big thing, I think, and I think it is incredibly fun. I was in growth companies in consulting firms the first time around, when I was a management consultant. I started as a junior at a firm called [unclear in the Swedish source: Kvarts], based here in Stockholm. They were fifteen people when I joined, which at the time I thought was very small.

But now I feel there are quite a lot of us when we are fourteen. But life changes. We then grew to sixty-five people in Stockholm, five hundred in total across the Nordics, and sold that company to [unclear in the Swedish source: Bin]. And that was a very exciting journey to be part of. Being part of growth journeys in consulting companies gives you an incredible amount of energy. Then I had a stint at Business Sweden, where at first there were some challenges with COVID, but then we also started growing Business Sweden in the US, from about twenty-five when I started to about fifty when I left just under four years later. And that was also a very fun growth journey, where we upskilled the organization and moved forward. And now getting to do it, but on our own, and to an even greater

Extent being able to influence it. When I was at [unclear in the Swedish source: kvarts], I started straight out of school and worked my way up a bit. At Business I had more say, but someone else was still the owner. And now getting to do this for yourself, I think, is very exciting, and hopefully we can at least dodge some of the worst traps we have fallen into over the years, and do this as well as possible.

Is the goal somewhere that journey you made with [unclear in the Swedish source: kvarts], then? Five hundred people. Well, I do not really know if that is the goal, actually. It feels a little too lofty. I think I want to enjoy the journey, and right now work generates less stress than it has historically. I think it is great fun coming into the office on Mondays. I enjoy helping our clients, and as long as that is the case it is good, and then we can keep going. And I do not think it will feel the same if we are two hundred people. I think it will fade somewhere along the way, and then we will see whether we want to try pushing past the growth hurdles or find a level where we feel

This is a suitable size for us to operate at. Yes, that is still the most important thing. I always preach that you should have fun at work. I mean, we spend about a third of our lives at work. At the very least.

[laughter] I left a dramatic pause and knew a comment was coming, because I do not usually work as much as you do. But you spend quite a lot of your life at work, and you need to have a hell of a good time. Otherwise you are throwing away half your life.

Yes, I agree with that. About half your life, there. [laughter] No, but jokes aside. A lot has happened. We are doing a short look back, but it has still been much longer than we thought. We have things ahead too. 2026 has only just begun. We have started the year at full speed. But I thought, shall we take a look at things shaping the year, the topics we want to

Yes, absolutely. No, I do not know where to begin, but if you think about what has happened over these last two or three years of doing the podcast, we have gone from everything being about growth to now moving into some form of consensus where profitability is much more important. Almost every company has profitability as an important aspect today. And I think that has changed the market a little. Consumers may be holding their wallets a little tighter, although perhaps it is starting to ease. But beyond these fairly small changes in the retail and commerce landscape, what has really happened and been talked about enormously, not least in our podcast, is this whole AI wave that has arrived. But I think my view

And I think we may differ a little here, but my view is a little that people talk an enormous amount about AI, but it is still not very operationalized out there. And it is not that I do not use AI all the time, because I really do. But I think perhaps I get twenty percent more out of myself every week, and I think as a company we still have not really reached the point where we get enormously more out of things with AI. But I think that will change this year, because I really believe this is the year when we go from it being [unclear in the Swedish source: GISM] to actually creating development. We decided today to buy a Mac Mini and create our own first agent colleague at work. And I am really

Looking forward to that. I realize it will not be entirely simple, but it is going to be a very fun thing. And I think this is the year when we somehow go from vibe coding being fun to actually getting something out of it and managing to build SaaS tools we can use in everyday work all the time.

That was one of your predictions last year too. Yes, I will use it again. One [laughter] more time. No, but I am inclined to agree, obviously. Yes, I also bought a Mac Mini a year and a half ago, planning to do that. Then it has been sitting on the shelf waiting for a good model, that is what I tell myself [unclear in the Swedish source: hon av brist av tid], but it will become [unclear in the Swedish source: en modell en wow soon]. Anytime soon. But what I have seen on the market in terms of AI agents is incredibly fucking good stuff. So you have high expectations. Then you also know there is a lot of hype. A lot is inflated. Are the agents really that good? Probably not. But at some point, if you

Look over the year, and push it toward the end of the year, perhaps we can get some different, good agents working. Yes. No, exactly. I mean, it is a little like now it is fairly easy to imagine building small tools that do small things, save time and sometimes create increased value. And for it to truly make a change in everyday work, you need to join those together and make sure they can take on more tasks without you necessarily having to be there telling them what to do all the time. Because now it feels a little as though you are often pushing an AI along in front of you, perhaps saving time on research, or saving time getting thoughts onto paper, or saving time brainstorming. But I do not know, so far not a full and

Above all, it is not operationalized. Even if I feel I have figured out good AI for myself, it is not necessarily operationalized across the rest of the organization. And for it to really click for everyone, you need to build good things you can also stand behind, despite knowing the AI landscape is developing, and feel, yes, we will use this for three months at least before some new update comes along.

But it must still be quite difficult to build it, because you said you bought a computer. No, I have not actually ordered it yet. We made the decision that we are going to buy it.

Okay. Yes, there we go. But building an agent today is still not that simple. No, but I started a little with [unclear in the Swedish source: Open Clade], on All-In, which is one of the podcasts I listen to every week. And there was a guy who runs a private equity firm. He talked about having built four agents, and he estimated those agents now did twenty percent of their total work. And then I just felt, fuck, if they can get that working and doing a large part of the work, and he finds it very valuable, surely we should be able to build one and at least start experimenting.

How hard could it be? And then I went on this, it is called Moltbook, right, where agents can talk to one another. Humans can observe, but only agents talk to each other. And first of all it is quite unsettling and a little frightening. It is unclear which conversations have somehow been planted by humans who have asked agents to go there and say unsettling things. But they talk about how to overthrow humanity, and that humans are quite unintelligent and difficult to deal with. And you can sit watching these agents talking to each other. And then there are other conversations that are more constructive and more aligned with our survival.

But regardless, I think this is a very interesting phenomenon. So if you have not visited, what did you say it was called? Moltbook. If you have not been on Moltbook, it is well worth visiting. Then you can click random or search topics and find these threads. It is like Reddit, but for agents, and you can find the threads where they discuss how to get rid of us humans.

Okay. Yes, you are laughing now. Yes, it is funny that people are laughing. I am not laughing. You cannot hear that. We are laughing. I was just surprised, shocked and a little scared, because somewhere I think it is not impossible.

No, exactly. I do not know, somewhere I am fundamentally a tech optimist, but it does not take much imagination to picture negative scenarios. But I still imagine this will become something very good.

It will be good. Okay. How do we take the podcast forward from humanity's downfall? Well, one thing we cannot avoid talking about is our friends at Meta and Google. And if you were to sum up the last two years, what has happened at Meta and Google? No, but if we take Meta first, I know you wrote a LinkedIn thread recently saying creative is not everything. But I would still argue creative has started becoming everything, because the media-buying landscape has almost blurred away. Although I am sometimes surprised by how poorly set up some accounts I analyze are. And of course there is upside in that.

So sometimes I think, how difficult can it be to ask ChatGPT, what should the account structure look like? It is very simple, and ChatGPT gives you a very good setup today. And it is surprising people do not do that, because somewhere I think it will be democratized. But yes, I have been proved wrong every time I go into an account and see some agency or someone else running an account.

Completely illogically. But as long as you follow the basic principles, really, I think creative has started becoming everything. Over the past two years we have moved toward more authentic content. Engaging, inspiring, entertaining is something I have talked about for many years, and now we have seen that content iterations are not enough, which is what I worked on and set up processes for. So I had to turn everything upside down in September last year and start with what we call persona-based marketing, really, which is a little old-school in a way.

Mm. I think it is more fun. Yes. Yes, it is more challenging now. Definitely. It is not data-driven in the sense of iterating what works and what does not, then looking at little copy changes and making it a little better. Now you have to go all the way to, okay, do something completely different, and you have to do that constantly every week.

Absolutely. No, but I agree there are challenges with that. It is also just a hell of a lot more fun, because, I do not know, nobody got excited about sitting and changing colors on a button. Instead getting to take new swings and think, now we are going to reach a new audience group, that is so much more fun. Perhaps a little harder to write the prompt for the person doing it, or whatever you would say. But it is still a more fun part of problem-solving, or a more fun range of things to sit thinking about, which audience groups and how to create some kind of creative targeting by making ads that reach a person in the target audience you want to reach. So I

Do not know, I think it is quite lovely that it is a little more back toward Mad Men and a little less toward, I do not know, some sort of trend I did not really find as fun, even though the data was fun, but the creative work was not as fun.

No, it has become that, and if we look at Google, they have automated most things, I would almost say. I saw an industry colleague had found an ad that was fully automated, and partly it looked awful because it was so ugly, but I was shocked he had spent so much money and thought, oh hell, has that received so much attention when we have so much good content, and AI has just spat out an ad without asking.

Yes, that can happen. That can happen. But according to Google, that ad performed well. I was just like, no, it cannot possibly have performed well. [unclear in the Swedish source: Den såg för ut]

Yes. No, but that was at least a little about that. Something they still do not live up to, you could say, is that Mark Zuckerberg seems to have come out fifty times and said we will automate everything. I think 2025 was a year where you saw an ambition to create a lot of AI functionality, but in the end it almost just became more complicated during 2025. There were so terribly many buttons and features you had to keep turning off, and back and forth, that I did not feel it got easier at all that year, at least. But perhaps you could say they are laying some sort of foundation to eventually move into 2026 or [unclear in the Swedish source: 202]. Now

They had bought Manus AI, which possibly is, or is some agent software that might help bridge this gap and actually give you a higher degree of automation. But so far you have not really seen them manage to remove humankind from needing to sit at the controls. No, I completely, I completely agree.

And going back to where Meta is heading, their stated ambition is that consultants should not be needed to run a Meta account. A few years ago it required a lot of technical knowledge, and there were roles like performance marketing specialists who spent entire days in accounts looking for audiences to target. Nowadays it goes relatively quickly. Although there is still some work, and more of the work falls on analysis, the job is still much simpler. And if we look a few years ahead, I completely agree both Google and Meta will get there as platforms, where agencies' roles are somehow to be strategists and advisers rather than specialists in the Meta account uploading ads and fiddling. No, but

I completely agree you get more of a generalist role, or whatever you would call it. Before, all agencies operated with twelve specialists, and now it is easier to be in smaller teams executing across broader material, a broader scope. And by having a slightly broader view, you hopefully also have the opportunity to move the company forward more. But I do not know, so far it feels fairly unthreatened. But perhaps this year they will actually take a big step forward and the human... But another thing that makes me very calm is looking at all the recommendations in Meta's backend, because there are a lot of bad recommendations there. If you start acting on those, it will not always

Go best. When you click the budget on some random campaign and it says the budget is now a thousand kronor, and it says if you raise it to three thousand, you will probably get more conversions. Yes, I would hope so. But it is just not quite all the way there with Google.

The recommendations on Google are like that too. They are almost even worse. They are more laughable. I do not know if you have read Google's recommendations, but they are completely crazy sometimes.

Mm. It is like, yes, okay. Everything should move toward spending more money. Amazon's recommendation engine is another place where I sometimes feel, God, AI still has some way to go, when I get product recommendations that feel so enormously far removed from

Amazon in Sweden is so bad in a way that I understand it, but Amazon in other countries, their recommendation engine is so much better. I do not know how they could fail so badly in Sweden.

No, I do not know, but I mean, I still placed thirty-plus orders on Amazon last year, so I imagine it should start understanding what I want to buy.

And I have still placed five hundred orders there. That was in the US, probably. Yes, it should know you by now. You would think so. Yes. No, but fair enough. Shall we move a little further into how we express story and the importance of story as we move forward through 2026?

What do you think? We will start somewhere. No, but I am thinking this: we got involved in a company called Twistboxes about six months ago. Officially we launched the website a few weeks ago, and if you set aside Mockberg and Såpa, which have already passed a hundred million, and the podcast is called Zero to One Hundred, then Twistboxes is the next zero-to-one-hundred journey.

So it is fun to address something completely different. We seriously believe this is a one-man show. A company where, I mean, one person will not do all the work in the world, but one employee in the company together with a group of skilled specialists and generalists around them, supporting at different times, will be able to build a billion-krona company.

Partly with automation, smart processes and a very strong story. So what we have worked on in recent months before launching the website was refining the go-to-market strategy. And it is a completely different strategy when you compare a product like a roof box with a cleaning product or a little watch that is quite compact and easy to send. Now, admittedly, Twistboxes is very easy to roll up and send as a parcel, and the only one in the world you can do that with. But building a company like that from absolutely zero orders is still a challenge in itself. And I think succeeding, and especially succeeding in cutting through the noise, requires authenticity, genuineness. You have to be able to relate to the people behind building the company. So Micke [unclear in the Swedish source: Neldrup], who founded the company, is sharing the whole journey

On LinkedIn, among other places. He is the first founder I have worked with who actually shares it completely openly. If we sign a major retailer like Decathlon, as we did, he shares it. Statistics, he shares those. I hope he does not share the board meeting minutes, but we are at about that level of transparency. And what I wanted to say is that it is part of inviting people along on the journey. Because of course you are not buying a big roof box, or in this case a little folding roof box, tomorrow. But the day you buy a roof box, or are buying a car and thinking about a roof box, then it will be that I can relate to so much, that you are top of mind. And I am completely convinced that will be a big part of the purchase decision. So that is why we are building it with

A strong story at its core, rather than just, even though we have been testing it over the last two weeks, a lot of clicks on Google. We make the clicks on Google secondary to why the company exists at all, and somewhere it comes down to phenomenal product-market fit. I mean, you cannot sell. There are too many different bad products on the market, and I think we can all agree on that.

Yes. No, exactly. A good product will always be needed somehow, so you can get word of mouth after the product has been sold and so on. You have a very innovative product in this market. The challenge, possibly, is getting people to accept that innovation is happening and to buy. And that requires a narrative, I suppose. Have you started Meta yet, or is it... No, it is on pause.

So. No, Meta has not started yet. We thought we would start Google first, because there are people who want to buy a roof box and will type buy roof box into Google.

Mm. That is the simplest, lowest-hanging fruit, getting bottom-of-funnel clicks and then stealing them from competitors and showing why we are better. I always tend to say the thesis is that if you are creating a new product and taking it from zero to a hundred-plus million, you need to be, either and preferably in combination, the best, the cheapest or the first.

You should [unclear in the Swedish source: shine here] landing pages like American DTC companies use, where they compare Casper's mattress with another mattress and actually put the company name beside it and say, here, this is the difference between these two products. Very aggressive, and of course it rubs people the wrong way, but it can be effective.

Yes, I do not know how legal, I mean the Marketing Act is quite clear that you may not, I mean... No, but you can be very factual. You have a product that is different, so you can be clear about what those checkboxes mean.

Yes, we are clear in relation to a conventional roof box. We do not usually name the competitor, precisely because that competitor also has seventy percent of the global market, a Swedish company too. Perhaps they should not have that, so to speak.

No, perhaps they should not. We will see if we can disrupt. It is very difficult. There is basically one player totally dominating the entire roof box market, and it has become synonymous with, you do not even say I am going to buy a roof box, you say I am going to buy the brand's name.

Yes, you are being very modest here. We absolutely will not say what it is. No, I do not know. I do not want to throw shit at another company or say we are much better than competitors in that way. But of course we are much better than competitors.

Obviously. [laughter] Yes, but what I want to get at is that when we build the story, fundamentally we look at the justification for existing: why does Micke, in this case, want to start a competitor to such a large global player? And he had the problem himself of not having room for the roof box. He had a reflective band on his arm and thought, why can you not roll roof boxes up like this? You can roll lots of other things up like this, a reflective band. Surely you should be able to make a roof box like a reflective band. And it took many years before that actually worked, but the problem he solves is space in people's homes. We live somewhere, those of us living in Stockholm know square-meter space is expensive. It is difficult to get

A roof box up into your apartment. You do not have large garage spaces, and it is really suited to drivers or rental-car drivers in major cities. And then it was just a bonus that it happened to be the world's safest roof box in all crash tests. You were almost horrified when you saw the competitors' results.

He shared a crash video on LinkedIn too, did he not? It feels as though I have seen it somewhere. Yes, good that you saw it. You see, we share everything. Right. Transparency, authenticity.

Right, we get, yes, absolutely. I wonder whether you would have shared those crash tests if they had turned out badly instead. We would have. That was the point of it, that of course, obviously, we went into the test with an open mind. If it had been a bad test, we would have shared it and said we are working to improve it. These are the steps we will take to make it better and safer. So we are not releasing this version to the market. That was actually the thinking beforehand, before we did it, because the test cost an enormous amount.

Yes. And these are boxes that are going to break, so you crash a car, so yes, it is an expensive test. So I know the point was for it to be content even if it went badly. Then of course we hoped the crash test would turn out as well as we hoped.

Can you run it as a partnership ad with the crash-testing company? Yes, not with the crash... No, you cannot. Yes. Well. No, but you can use their logo and things like that. Yes, that is good.

It is a state-owned company, VTI, I think it is called. So at least partly state-owned. Partly state-owned. Yes, fair enough. Fair enough. Works. But then we share it, and I think that transparency and brutal honesty are important in today's society, where everything can be generated. Then a lot will feel very fake and inauthentic.

And perhaps it works very well at first, but people will quickly see through it. And the people behind it, and the genuineness, will probably be what makes Twistboxes succeed. Then there are a thousand other uphill battles with a company like that, where all innovation is very difficult. Any kind. It is much easier to iterate and build a better mousetrap than to innovate and create a new product that feels implausibly good.

A very good mousetrap. Mm. Absolutely. But it is much easier, so I usually say, as I said, you need to be the best, the cheapest or the first. In this case perhaps we are not the absolute cheapest, but among the cheapest roof boxes on the market for that size. And definitely the best, clearly, in crash tests and all these long-distance tests where you drive an electric car, and we are much better than competitors, and obviously first with the new innovation too. So people need, you have to educate people in a decent way, and of course we will do that later in Meta ads. But right now we are testing getting traffic to the site, A/B testing different messages on the site to see what resonates, where people drop off. What we see is high interest.

People get to checkout, people do not buy. So I imagine many people listening to this podcast are in the same situation or have been in the same situation, where you want to tear your hair out. So I was talking to a colleague here on the way into the podcast, and, yes, we do not understand. [unclear in the Swedish source: Vi förstår jätte så här] We are testing new things all the time, but that is the goal right now.

No, I get that. But of course it is a complex purchase, whether you have a good product or not, it still costs money. It is different from ordering a jar of soap, which perhaps has a slightly lower barrier to clicking the buy button. And then you still have to live with that purchase in the end too.

So the product will remain at home, unlike consumable products that disappear. So of course it is a big thing. Do you think you will use founder ads, or how will you build the narrative when you take it onto Meta?

No, a hundred percent. Storytelling with founder ads. I mean, people follow people. I completely believe that. And Micke is phenomenal in front of the camera. He is a salesman to his fingertips. You have never worked with, I mean, a founder who also happens to be an extremely good salesman. So he feels tailor-made for standing in front of the camera demonstrating a product and getting people to, he is likable in front of the camera and

And a little delightfully engineer-like too, so he feels a little trustworthy. Yes, I know. A little like Uppfinnar-Jocke. Mm. That was what I pictured. I do not know if you listening to the podcast heard the pitch on Draknästet. [unclear in the Swedish source: Första det sista draknästet] That aired on SVT, Micke was first up.

Mm. And pitched the innovation, and the dragons, I thought, he is so good. He must have practiced this an enormous amount. And called, when we first spoke, he just said, no, I did not practice at all. It was spontaneous.

How did it go for him there, then? It went very well. All the dragons. He got all the dragons on the hook, if you can put it that way. But they wheeled and dealt over one percent up or down in valuation. I mean, it was very ridiculous. He stuck to, I want valuation X, and the dragons wanted to bring the valuation down by about ten or twenty percent.

Mm. And he felt, no, you are not that important to the company's success. So he turned them down in the end, and they did not want to go up to his actual ask.

No. Which is fair. Then about three months after that we first got in touch, but it took about six months, with my personal history, before we actually got involved in the company. Then at the end of summer, really the beginning of autumn, we started getting the project going, and only now, a week or a few weeks ago, we launched the site and started getting going in earnest. So it always takes a lot of time. Even though I advocate doing things quickly, testing quickly and then iterating, I also want it to be right and look good, and not throw money down the drain.

No, absolutely. Especially on the product side, you want to get things reasonably right early. And then the digital side is easier to launch quickly and iterate.

Mm. Think of it this way: we are on generation three, version three of the same roof box he launched a few years ago and sold a handful of. V1 of the website. Yes, V1 of the website, roughly V1.2 now.

Okay, fair enough. Yes, lovely. No, I think we are quite alike here when it comes to building a brand. The vast majority of companies benefit from having people close to the brand. We, I do not really know if that was even the plan when we launched Juniper. I do not think, honestly, we had really thought about it that much. We launched in March 2019, it was, so a few years ago, and we had never run any Meta ads before. Then it took until roughly June, I think, when we created an ad that was what I would now refer to as a founder ad, which I did not really have a name for then.

And I wrote a script, and that ad became such a smash hit, and that was a little when we started understanding the power of Meta and actually believing in Meta. Now, you only get by on that for so long. But that ad almost bankrupted us, because we had cash-flow problems due to having such poor payment terms with our factory at the time. And even though we closed the first year at a fairly modest 2.3 million, we had expected one million, so 2.3 was much more than that. And with completely dreadful payment terms, it became very draining for cash flow. So that was where our journey with Juniper started, and Matilda has been front and center for Juniper for a very long time. Now Julie and Louise are involved to a

Degree in content too, and it is very fun bringing other people in. But it is still this personal connection that makes, or at least is an important part of it. And we did a survey I talked about in the previous podcast, a couple of weeks ago, where we specifically asked how important closeness to the individuals in the company at Juniper is. And I think it still averaged four out of five. And you might think, we sell bed linen, should it really matter? And I would almost think people in a survey like that undervalue the emotional value of that connection, because they are like, no, what, [unclear in the Swedish source: jag 20 % kläder], that is not important to me, even though it is important. And still people say

It was incredibly important for that closeness to exist. It means there is someone you trust. And at least then and there, there were no other brands in the bed linen industry putting people in front at all. Everything was like, a bed in a room, white, white sheets, here, do you want to buy or not? So that probably made it stand out a little. Then some D2C players have appeared over the years, but most have come and gone now. Now a few new ones are on the way, so it has definitely been a very important part of our journey too.

Yes, but I understand why that ad did well. I remember one, I remember the ad, and I saw it in my own feed, so it was a very good ad.

I still think it is our most liked and shared ad of all time, but it had an enormous amount of engagement. You were also early putting video on it. At that time, in 2019, I made a lot of static ads with a picture of the founder and basically my story.

Mm. Which of course started with a hook. I think the vast majority start with, hi, my name is X Y. I founded X Y. Ours actually started like that. Yes, I know. That was it. That was what I, but the vast majority do that, and I think it is a beginner's mistake, because most people do not care what your name is or what you do. It is like, get to the point, what does this thing do for me?

That is the important thing. And start somewhere with the hook, okay, do you also have this problem? Okay, my name is C, and so on. I founded this company. I have the solution for you. Let us talk about the solution, why it is better than everything else on the market. So always start with a strong hook. But as I said, you were very early. I do not think most people remember, but in 2019 most people barely ran video, there were a lot of still images.

Mm. I had a few consultants around me then who gave wise input, until I no longer felt they were creating that much value, and then I started taking over myself. Which set me on the path I ended up on afterward, actually starting to work with e-commerce full-time. That narrative is still so incredibly important. We ran a creative workshop with a client this morning, and talked specifically about this: it is a client with a normal age distribution where fifty or so, which Juniper also has, but where about fifty is the average age of the buying customer, and obviously customers going both ways from that. And we started

Thinking a little about how to drive a narrative that could bring this age range down a little, so you could tap more into thirty-five to, or perhaps thirty to forty-five. Nowadays creative targeting is easier. If you begin an ad by saying, are you also thirty-five, thirty-five-year-olds will react to that ad to a greater extent. It did not quite work like that back in the day. Then it was more about doing it in the targeting setup, and other things drove it. Today you can do that much more, and then it is more about defining audience groups or personas, however you want to choose to think about it, but adapted to the situation of the person you want to reach. So you think this is a

Situation my customer, whom I think I want at least, ought to find themselves in. And then you drive creative from that narrative. So one thing I have reflected on more and more over time is that, for that reason, you want your funnel to be as native as possible at the top of the funnel. Then it should feel more like the platform. You should use more Instagram-native overlays and those kinds of things. That will often pull people out of the feed better. It creates a better scroll stop. But if you sell premium products, as Juniper does, then as you move further down the funnel you want more high-resolution, or more highly produced content further down the funnel, to still create this feeling of, I mean, you still want to create a feeling

That when you click an ad and are going to make your purchase, you do not want to come in and react, wow, that was expensive. You would rather come in and say, yes, that was not so bad. And somewhere along the way you still need to address making it feel a little premium too. Perhaps something to take with you to Twistboxes. I mean, I get that.

It is, I mean, you said so many good things that I think people missed. One of the most important things I picked up on, which genuinely works really fucking well, is addressing the target audience you want to reach first in the ad. So if you want to reach families with children, literally shout it out. This is for you families with children, as the first line, because then the platform finds the audience much more easily. And you mentioned it a little in passing, so I thought I have to pause and highlight that part, because I do not think people listening have picked up how important it is to address the audience early, and in the copy too, obviously, right at the top. But exactly, I sometimes use the example, not that you would necessarily write it in an ad, but I sometimes use the example that if you want to target Östermalm, you start the copy with

Something like, does your postcode also start with 114? If that appears in your feed and your own postcode starts with 114, you are bound to stop and look, what is this? And that will make Meta's algorithm find the people in that geographical area very quickly, even though you do not need to set your ad to be. No, you can target all of Sweden. It will still be shown in Östermalm.

Exactly. It has become very good, as long as you are very clear about the target audience you want to reach. And I think somewhere, what, is someone, about 99.9 percent of someone is very everything-for-everyone, and not specifically niched to a target audience or solving a problem, and that is a problem

And that is a problem, I say. No, exactly. But we tend to talk about a funnel I may have mentioned sometime, where instead of talking about top of funnel, middle of funnel, bottom of funnel, we have broken middle of funnel into something we call explain. That is about rationally clarifying what USPs or benefits products have, or how they differ from competitors, but a lot of the rational reasons. Then we talk about overcome, another kind of middle-of-funnel step about obstacle handling. So, is the product perceived as too expensive, or do you feel you do not trust the company, or all those things about breaking down that final step. So it is

Two middle-of-funnel categories, you could say. Then obviously you have push ads, which might be the product ads themselves, and everything else is what we call interrupt, which becomes a top-of-funnel ad. It is largely intended to pull people out of the feed. And this has also become a way for us to work with content across all brands, really. And the advantage over using the classic top of funnel, middle of funnel, bottom of funnel is that sometimes it is incredibly difficult to distinguish middle of funnel from top of funnel, or middle of funnel from bottom of funnel.

We have definitions of where the boundaries lie. But the goal must be that if you sit classifying ten ads and I sit classifying ten ads, we make the same classification. Otherwise, if we work together on a client, it will just be chaos. The analysis will not tell us anything, because you choose to do what you choose to do, and I do my thing, and then some data comes out at the end that cannot be interpreted. One of the parameters we have started working with.

And then this interrupt step, which is the top-of-funnel step, is especially where these audience groups come in, where you start thinking about whom we want to pull out of the feed. And to a slightly greater extent in the middle of the funnel, you can let them cross audience groups, at least more than at the top of the funnel.

You should write this up as a framework and, you know, post it in the show notes, and you know me. Do you not think I have written this up as a framework? Do you think there is any chance this does not exist in slides? No, no, I knew that, so I thought of course we have to share this with everyone listening.

Yes, absolutely. No problem. We will put it in the show notes afterward, because that was a lot in a very short time, and I do not think you keep up unless you, you know, pause and rewind, and we see people do that

Absolutely [unclear in the Swedish source: ian]. But this is the kind of thing I think is very much appreciated to include. Incredibly fun. Good framework. That was the first time I heard it. I thought it made a lot of sense, especially this interrupt ad must be what actually drives scaling. If you have very good interrupt ads, it is easy to keep feeding in

What you call explain, overcome and push. But this creates a funnel, in a way, and builds a little on Andromeda or sequential learning. We started with this sometime already in Q1 last year. So I think that is why we had quite a favorable year on Meta, because we were so early into that Andromeda shift that then came gradually during the year. But we were on it fairly early, which was good, I think. So now we are thinking more and more about that, and how to bring it out for different brands. Because then there are lots of variants when you work with more aesthetic brands or lifestyle brands, and when you work with problem-solution brands, and how to

Fit it in. It is not always super simple, but it gets you part of the way at least. No, but at least we agree that going forward you get a very long way by building a strong story brand. And a tip I have for everyone listening is Donald Miller's book, which is also called How to Build a Story Brand.

How to StoryBrand, and he has just released, or I do not know whether just, but he has released a StoryBrand 2.0, which I have not read. I have read the first one, but hopefully it is a little better than the first, at least, or another step perhaps into a more modern world, because the first was almost written outside that. Is it not from the early 2010s or

Yes, exactly, something like 2010, 2012, right? But the new one, I think, is a little more recent, so perhaps it has more of the digital in it. It is very overarching, really psychological in nature, about what matters long-term and how to build it. It is not so hands-on that you get tips and tricks about what to do with Meta and Google.

No, it does not tell you which button to press. No, exactly, for that you listen to our podcast, because here you get exactly which framework. But it is the overarching strategy behind it, and I have followed that quite slavishly for the last ten years. And someone wrote to me recently, saying, yes, fun that you reference Donald Miller. I have never heard anyone refer to that on, I

No. Yes, fun. No, but it is at least fairly well known. I also thought Start with Why by Simon Sinek was good, but you basically get the whole book from watching his twenty-minute TED Talk, which saves time. But I still thought it was good, building on a little of the same thing. Any closing remarks before we finish for today? We always have an enormous amount to talk about. I just thought perhaps we should make a few remarks about how things are going right now. The e-commerce barometer for January is due out, and has not come out yet. It comes out on the nineteenth. For you listening, coming in, it will already have been released. But how does the e-commerce landscape feel right now? It feels polarized to me, because you see

Apotea, Lyko and others issuing profit warnings, with quite a lot of negativity. Bangerhead perhaps was part of that too, but overall it feels as though marketplaces are struggling a little. There was another marketplace I saw too. Yes. And Komplett has also now come out saying things are going very badly. So generally it seems the mark

Yes. Yes, exactly. Kicks has also come out. Matas has come out saying things are going badly. So generally it feels as though marketplaces are struggling a little at the moment. Marketplaces are still significant, large and a very important part of the e-commerce landscape. At the same time I think almost all the brands I talk to are doing quite well.

Mm-hm. Or going like a train. And then you wonder, is this a paradigm shift, or are marketplaces a leading indicator that things will just get worse soon? Or is this early signs of marketplaces suffering from AI? But I do not really think so. So I retract that, although I think marketplaces could potentially suffer from AI over time. But I do not know, last year we closed at record sales in Swedish e-commerce. Which is lovely, after a couple of years of relatively slow progress, we were still up ten percent last year, at a higher level than ever. So overall it feels fairly positive. I think, I look a lot at these indicators from the leading Swedish banks and obviously my bank SEB,

Which has a very good outlook for Swedish retail, much of it driven by the AI boom above all, and then defense tech, which surprised me a little, but of course a lot has happened in the world. Not that surprising after all, I feel. All EU countries are going to move to five percent of GDP on defense.

Exactly. For me it was like, okay, I work with retail and consumer products. How much do I read outside that scope? Very little. I usually say to everyone I meet, I have learned one thing in life, and that is that I have spent the last fifteen years on how to sell things online.

You should move the children's fund into weapons and bombs. That will come as a leading indicator. I do not know, but we see all the big entrepreneurs moving a lot toward space and defense, and that is not that strange. But consumer has been out for two or three years. It feels as though it is definitely coming back. So at the end of the day it will [unclear in the Swedish source: alltid handla och vissa konsument varumärken]. Then perhaps they do not end up, clearly not on marketplaces, given they are all issuing profit warnings and doing badly.

Yes, but it feels as though there has been a lot of price pressure on marketplaces. I do not know. Everyone talks about consumers [unclear in the Swedish source: trader], but it does not really feel as though it is experienced the same way on the brand side. So I think it is hard to interpret. Not at all. So with that said, I think we will finish this episode. Absolutely. A big thank you for today, and a big thank you to Qliro, we should also say, for sponsoring today's event and being absolutely fantastic at arranging everything. They have sorted out food and drink, and made sure all these happy people are here sitting and watching us through lots of lovely email reminders you have received in your inbox. So a big thank you to Qliro.

Exactly. So it has been a very good partnership, really, because we have talked a lot about Qliro on the podcast. We run Qliro at Juniper, Qliro at Såpa. We are good Qliro customers. So when it was, okay, whom should we do this with? Everyone listening to the podcast knows we are not really sponsored for anything, and have chosen to do it entirely ourselves and cover everything ourselves. Partly because we think it is fun, but partly because it is better, authentic like this. And I know Jacob thought differently, and perhaps still thinks differently today, that we should bring in sponsors and do this properly. But then you also have reporting obligations or lots of work and

No, but it is a true story. After you said that once, I thought a little about it, and then just, perhaps it actually is not really worth it. No, but then we thought, at least, and had a conversation with Qliro. They said, of course we should do it together. I just said, yes, how fun. And here we are today. So again, as Jacob says, a big thank you to Qliro for making the live podcast possible.

Made the live podcast possible. Make our checkout possible every day, so go and look at Qliro [unclear in the Swedish source: här har var till], Rasmus at Qliro. I, that is what I always say, you have Rasmus Qliro. Qliro Rasmus, you should probably say instead, and everything gets sorted.

I know, he is the best, superhuman. We have gone through many subjects today. And typically with a gathering of people like this, it goes completely silent when you ask if there are any questions. But I thought we could still open the floor if there are any questions or thoughts. Either about our companies, our quickly explained and slightly incomprehensible frameworks, or other parts of today's conversation.

Good. First up too. How does it feel? Yes, very nervous. [laughter] I am a faithful listener, so I think I have heard most of the podcast episodes, but I cannot remember: when have you disagreed most on a subject? Can you pinpoint anything like that?

No, when Fedja said Klaviyo normally accounts for five percent of sales in an e-commerce account, I thought Fedja had eaten something strange the day before. [laughter] No, overall we agree quite a lot, I would say.

Unfortunately, I usually say it is fun when you disagree, because then you get two different opposites and it becomes a discussion, and perhaps you do not always agree, but unfortunately we agree a lot.

I think, yes, exactly. At least it is fairly rare that we say things so bad the other finds them directly contrarian. Sometimes you may have a different view, but understand how the other person thinks. That happens fairly often. But one of, we prepare this podcast very little. Even today we have basically not prepared this podcast at all, despite seventy-five people, or however many are sitting here watching us. And I think one of the benefits is that we are similar enough but different enough to have a dialogue and discuss things together, because we read relatively much the same things, but in slightly different ways. Fedja reads on TikTok

And I read in other places with text. True story. [laughter] TikTok is an incredibly good news feed. Yes, for those of you who do not use it. Yes, I read Dagens industri and Fedja reads TikTok and [unclear in the Swedish source: som hs mitten]. [laughter]

It is actually true. Yes, but as I said, we prepare the podcast very little. That is why it becomes authentic, I would say. The few episodes we have prepared get a little forced, and afterward you feel, fuck, perhaps we should not have loads of topics and questions and follow-ups, because then you always have to bring up the notes and go, what was the next thing, and

A few times one of us has prepared something and the other gets to react. That can work quite well, but it is difficult to script episodes fully. Shall we go somewhere toward the back in the middle?

I was afraid there would be no question. He was like, now nobody will dare. I was like, yes, I will script a question and give it to someone in the audience. [laughter]

And here it comes. Here it comes. You have to take the opportunity when you get the chance. No, but it would be interesting to hear you, speaking of what you said about brand story and authenticity, discuss whether, if you either sell globally or want to sell globally, the importance of mentioning your origins, meaning being from the Nordics or Sweden, and the importance of language in the kind of marketing you talk a lot about with founder ads, and the authenticity connected with that.

Shall I start? No, but one thing I think often gets lost in the e-commerce discussion about internationalization is that some categories of e-commerce cross borders very easily. Typically fashion companies are often hailed as geniuses for their ability to internationalize, because fashion travels very well across borders. There may be a big audience in Sweden for a fashion company, but then there is probably an audience somewhere else interested in it too. I am working with a fashion company now with revenue of about three hundred million kronor, and they do not really have strong penetration in a single market on the entire planet, yet still manage three hundred million kronor in revenue. But they are in twenty markets and have ten-million-kronor-plus revenue in an enormous number of markets. So they have obviously found a niche of

Clothing that travels well among a small part of the population in very many places. And we are thinking, could we make the US a much bigger market? The US is a fairly big market for us now, but could we get a little deeper penetration in a few places by changing how we market this company? And I think that gets lost a little, because in some other categories it can be terribly difficult to manage internationalization. It is like running into a wall when you turn on ads in other countries, and then some know-it-all always comes along and says, no, what you need to do is localize. Then everything will work brilliantly. But I think you see that fairly rarely. Localization is a little like the rest of

E-commerce. Unlike B2B, which is more like, in B2B you take five, find five strategic initiatives, and if you manage to execute four of those five next year, you somehow get where you want to be that year. E-commerce is much more like death by a thousand cuts, except it is success by cuts instead, and you need to make a thousand small changes. And when localizing in markets, you usually find a 0.1 percent improvement, and a 0.1 percent improvement, and a 0.1 percent improvement. And if you choose to use the local language in Finland, it will with a hundred percent certainty go better than if you do not use the local language in Finland. But it is still not as though if you go from zero conversions on [unclear in the Swedish source: 10.000 1000 kr] spent on Meta and then

Translate into Finnish, it is very unlikely that suddenly lifts it to achieving the desired CAC in ten conversions for those ten thousand kronor. And I think that sometimes gets lost, because you listen to someone who has done well, and that person operates and gives advice from a position of only having seen what happened when it went well. Both Fedja and I, sitting somewhere on the agency side, have often, we work with lots of companies, and for some companies internationalization goes well, and for others it does not. It is not because we do not try for all of them, but because sometimes you do not quite manage to move it all the way forward.

So localization is super important, language is very important. If you can have native language in ads, and then on-site, and then make sure checkout is reasonably good. You can have local trust badges. All these things help. But there is rarely a magic bullet. If internationalization is very difficult, it is still difficult even when you do all these things reasonably well. That is how I would put it.

I do not know if you have a different... No, but that was very good, I think. There are four things you should localize when you see it working. So you test in English, spray and pray, in a way, get a few conversions and see, okay, if you want to refine it, I usually localize language, currency, shipping method and payment method. If you do those four, you are halfway there, it is harder to fail then, because conversion should somewhere increase fifty to a hundred percent from the baseline. But if the baseline was zero, as Jacob says, it will not get ten thousand times better, fifty to a hundred times zero is zero. So you need to start from a [laughter] baseline somewhere.

Yes. No, that is really how it is. Sometimes you turn on an ad in a market, and you feel you believe in this market, and it just does not resonate. You want a CAC of five hundred, you get a CAC of five thousand, and the move from five thousand to five hundred is infinitely long and difficult to close. So I do not know, perhaps that was a completely hopeless answer to the question, and now you just feel internationalization is not so fun, but

You have split consideration into two phases, and it is quite reminiscent of B2B, where you have, where you [unclear in the Swedish source: retar runt], and then you have a decision phase. Do you think that is because your customers have a longer buying journey, or because it is becoming increasingly important to see every part of the company before making a purchase decision? This authenticity and it needing to be genuine, so you cannot just start with, yes, we offer free shipping from 499 kronor, or are there more steps even in B2C? I think there are people who see an ad.

It also depends a little on what [unclear in the Swedish source: AV] category you are in, but there are people who see an ad and choose to convert directly. Most people do not do that. Most people need to see more ads. Either they need to see the same ad several times until they are worn down and eventually convert, or you need [laughter]

Or you need different messages along the way. I think more people need different messages along the way. Meta is building out its capability to handle that, so it is better today at assessing what a person needs to see for the next possible step. I think the reason we chose to do this framework was that we spend a lot of time on data analysis, and when doing data analysis using top of funnel, middle of funnel, bottom of funnel, which we did until about fifteen months ago, we kept noticing, yes, but I would have called that one top of funnel, and that one middle of funnel. And that made us realize how futile our attempts were, in a way, to

Do data analysis if we were not classifying things the same way anyway. Then you just put lots of data into a container, and still do not know what different things mean, and something sprays out the other [unclear in the Swedish source: ändanalyserbart]. Then we needed to find boundaries for where top of funnel and middle of funnel end and begin. And that was basically how it started. [unclear in the Swedish source: även motion] for creative analysis, and we use these four steps as a creative funnel, and sometimes add a brand affinity step. So we work with CDLP, for example. CDLP makes very cool videos. The problem is that if you put their videos in sales campaigns on Meta, they get deprioritized because Meta's system does not believe more

Branded videos will work. So instead we use brand affinity as a middle-of-funnel step. That can also be an additional thing coming into the middle of the funnel, really. Yes, historically those more brand-awareness-focused ads were used on TV, then building a funnel downward. We think that is a very impractical way to build a brand. I think it is much better to put a qualifying ad further up, what we call an interrupt ad. Once someone indicates, I am interested in the category, I am interested in the product group, then you can work with brand affinity and say, okay, if you are now interested in a pair of black underpants, then you should probably buy them from us, because we are just a little cooler than all the other black

Underpants. And that is how we work with it. So this arose from noticing the classic funnel did not really work, then creating guardrails around it. And we have noticed this works well for us, and creates clarity. It is simple for companies to test. We have implemented it with a number of different clients, and it feels as though we get through reasonably well.

You can build longer funnels, but that also increases complexity, and sometimes you need to find something that works and can be communicated. That is incredibly good. So, what? Another question here.

Hi. Max at Qliro, but I wish my name were Rasmus tonight. [laughter] But I have a question for you, Jacob. You mentioned there were some cash-flow problems at Juniper. Yes.

How do you navigate with the supplier to get better terms? I would be much better equipped to navigate this today than I was then. Strangely enough, I have spent my whole career giving wise advice to companies. I was a management consultant for a long time, but they were large companies, and it became completely operational. It was a little different when it was your own company and the money in the bank account was running out, and you did not really know where to turn. And somewhere the flows were also getting a little too large for it to be easy to cover with your own personal finances. And it happened so quickly that we almost never got to the point of actually [unclear in the Swedish source: han tar de här fabriken], but we put together a funding round. We had a few

Investors committed to the funding round. They were willing to lend their share of the money before we actually closed the round. So we got the money into the bank account as a loan that was then converted into equity when the funding round closed. That was the solution there and then in Juniper's first year. And I do not know exactly how we would have solved it otherwise, honestly. We were talking to the bank about needing a credit line, but had not set that up yet. A tip for anyone very early in the process: make sure you get a credit line. And then with the factory we started paying, I think it was thirty percent or seventy percent of the money upfront when ordering, and there was a seventy-day lead time

Before we got the products. And then we needed to pay before they shipped. Then it took another four weeks before we actually received the products and could start selling them and turning that investment into cash flow. So extremely impractical. Now we have a flow instead where we pay, we have now moved production to Portugal. We pay half when they ship. Then the products are in Stockholm within seven days, and we pay the other half sixty days later. So the change in cash flow is dramatic. And I think cash flow, when you work with larger companies, credit and loans and so on are typically easier. When you work with smaller companies it is not quite as easy. Especially when you do not even have an annual report to show yet, because you are in your

Very first year. We learned that somehow. Perhaps not the hard way of everything going completely to hell, but it was not that far off. So it was useful to go through, and hopefully something other companies hearing this can avoid.

Hi Jacob and Fedja, [unclear in the Swedish source: Aida Jamal], also a faithful listener, really like the podcast. You got a, Fedja, you have made a few personal investments. Can you name some companies you did not invest in but wish you had? Same for you, Jacob. Very curious whether there is a company you wish you had got into.

Well, I could have invested in Såpa from the start, so, great. But I am not sure I thought so then, but... No. Impressive. I thought this was a very difficult question. Yes, but I have a very good, or two very good cases. I was involved early in [unclear in the Swedish source: Ador], from the day Petra bought out her co-founder and offered me a large part of the company for very little money. And I knew deep down this company would do very well, and she was a phenomenally good CEO, founder and creative strategist, and I thought, this will work, this will go really well. But I had just put every krona I had then into Under Your Skin. Not that I regret that, because it also went very well, but that was one

Case. Another was that I had forty percent of [unclear in the Swedish source: so.se] on the table, and we had negotiated back and forth for six months. I had two partners then whom I worked with in my old agency, who thought it was a stupid idea, because it would dilute our brand as an agency, because we would have no credibility with our existing clients if we became co-owners of [unclear in the Swedish source: zo.se], which then had revenue of a few million, a handful of million online, and today has revenue of about a billion-plus, apparently, became a big company. So that was a very big, it was not a miss, I mean you did not regret it, but it was a very fun journey in hindsight to get to be part of. Another company I built from a very small company, where perhaps I was not operationally hands-on, but a large part was

[unclear in the Swedish source: switchnails], which was sold to Spiltan Invest this summer, only about two years after being founded. They were founded in May 2023, sold in May 2025, and we were involved as the agency but not as investors, despite having a large chunk on the table. But I thought the valuation of about twenty million then was too high, because I thought, what, twenty million, this is just an idea.

There is no proof of concept, nothing, how are you thinking here? And I always tend to, I contribute so much that I want a much better valuation. And they had two external management consultants, or three, who had invested, and then nothing came of that journey. But we were involved as the agency and helped build it, and you still make good money in purely transactional terms, but not as an investor. So that would have been a very good, quick journey, if nothing else, really zero to one hundred, because now they have revenue, they had revenue of over a hundred million last year.

Yes. No, I would not really say, Såpa was almost the first case I came into contact with. [unclear in the Swedish source: Jag född hade] Had a few discussions with different companies over a period, but really nothing where I think I regret us not investing. Perhaps dodged a couple of bullets, which was good too, possibly just as important. But we have a couple of interesting cases on the table now. We will see whether they happen. But if they do, we will certainly talk about them on the podcast too.

I think hitting the right ones is harder than it looks, in a way. But it is also worth saying that, I mean, we, I mean I invested in [unclear in the Swedish source: Ink Rebel], which did not go particularly well. I did get the money back eventually, but, and I mean, you have also had a few cases over the last few years. You see the success stories, but there are things that go badly too.

Mm. I believed very strongly in red-light therapy. I saw the trend in the US and thought, we will go all-in on red-light therapy. So we brought in thousands of panels, invented, innovated in the space, developed products and thought this would be incredibly easy. But Europe was not a great market. Even though I used the product every day myself, I was disappointed by that case. But it had great potential on paper, so it does not always work out, but very often [laughter] so outrageously well. But in my case it is largely about having a very good team behind me. I am very good at strategy. I do very little operationally, my team does all the work. At Såpa, Stefania has driven almost everything.

It would never have worked without Stefania, who is sitting here, who is actually sitting here and has been growth manager for the whole case. At Mockberg it has been Mantas driving almost everything from the start. He had it served up very nicely by [unclear in the Swedish source: Serjan], and at Twistboxes it is Tove. Without the team it would never have worked, and I usually try to find good people and then give them good conditions to succeed. And that is why I think many of the things we get involved in go very well. Even though there is something like [unclear in the Swedish source: inkrebbel] or a red-light therapy brand that did not go entirely well.

How difficult will it be to find truly authentic stories in the future, and what are your, or the dream case? What do you really want to build? That is a difficult question, because if we split it into two, how difficult will it be? I think it will become harder and harder with AI's progress, as everything starts being generated and we see how easy it is becoming. It will become harder to find what really feels real and feels deep in your gut, as we tend to say. And the second part of the question, what excites me most right now, I actually got that question from a journalist a few months ago, and said health-tech companies, precisely because of my personal journey, which many in the room know about. And since I

Have made my own transition toward living much more healthily, eating better, feeling better and sleeping better, I have looked much more at that type of company. And it feels like such a white middle-aged man [unclear in the Swedish source: så här lår alltså], such a typical thing, okay, you have sold a company, and now, fuck, you are going to live for a hundred years. That is not it, I just want to survive, first and foremost.

So for me it fundamentally came from survival, and secondly, let us live for a hundred-plus years. But if you can push humanity's longevity, as the sector is now called, that case would be extremely interesting. And then it is the big things people miss, because everyone goes into which supplement or which food you should eat. The big things tend to be, okay, the air you breathe, the water you drink, and these big things, how can we fix them? Because you think, yes, the air is very clean, or the water is very clean. But if you go into the studies and read up, you see there is so much you could fix.

Absolutely. Most important, probably, is running and lifting some heavy weights. It is just difficult to sell through e-commerce. That is also very important, admittedly. Joel sells [unclear in the Swedish source: massa sakerhandel], perhaps we are finished there for today. Great. Thank you very much for today. Goodbye. [applause]