The next journey starts here: niche ecommerce, Twist Boxes and the PR tactic that costs nothing (Episode 102)

A warm welcome to a new episode of 0 to 100. With us today we have a brand-new, possibly permanent guest on the couch. We have none other than Fredrik Ankarsköld, and I will get into Fredrik's story in just a moment so we can have a chance to get to know him. But before that I want to give a huge thank-you to Jacob Wibom Westerberg for 101 fantastic episodes.

You have been with us on the journey from 0 to 100 as we have built Västerbottenssåpa, which we have talked about literally in every episode. You have heard Jacob talk about the journey from, well, around 10 million, when Juniper was very small and had almost only just arrived on Swedish soil, to Juniper now turning over just over 20 million and running at an even better pace.

So that has also been an absolutely fantastic, incredible journey. Jacob has generously shared all the challenges they have had and all the obstacles along the way. And then, most recently, when he started his own agency. At the start of the podcast in 2023, Jacob was still working at Porobic Group, and then sometime in 2024 we went our separate ways. We carried on doing the podcast for almost another two years after that.

So it is quite incredible. We have survived a lot together. It was a tough autumn in 2024 when I faced the biggest challenge of my life with the widespread, metastatic cancer. And there was a break, so it has been a roller coaster like no other. When Jacob said he was thinking 100 episodes might be the place to stop, my first thought was: should we shut the podcast down or carry on somehow? And here we are today. After the last episode I received so many direct messages from you that, somewhere inside, I knew we had something good going. It would be a shame to kill our lovely, fantastic podcast. So we will carry on and see where we can take it.

With us we have someone who is, in a way, a long-awaited guest, who has usually been behind the scenes. Fredrik has helped organize and run Björn [unclear in source: Påmans]'s podcast, Framtidens e-handel, where you edit and listen and are really the one chiming in with those fact-checks, which I experienced myself just yesterday when we recorded a new episode. It is so nice having Fredrik in your ear, able to double-check things you blurt out. Because when you are sitting in the studio, you usually do not have a computer in front of you, nor do you want to be typing things in. You want to get information out, but of course you want it to be correct, especially when you are referring to articles you sometimes read years ago. So in that sense you have had your ear to the ground. Then we have got to know one another on a personal

level over more than a year now. We have had a few lunches together. You also have an incredible health journey, and we found each other somewhere around that health theme. So I am incredibly glad we will be doing 0 to 100 together now, and of course I want you to introduce yourself so all the listeners can get to know Fredrik a bit better. Tell us a bit more. What does Fredrik's life look like, and what does he do?

Yes, what do I do? I have a lot of balls in the air, but I run an e-commerce business together with a woman called Mia, called Core by D, selling training aids for horses. Mm.

So that is one track. The other is that I run an AI platform that helps e-commerce businesses build SaaS tools. And the third is really the production business. So I am involved; I have helped organize this studio here.

Mm. I am also involved in [unclear in source: Coy Agencies]' podcast, which is also about e-commerce. Wow. And design. And then there is Framtidens e-handel, which I am involved in. [unclear in source: Michalipper] a woman from Spain, but I film everyone and produce it, and help fact-check there too. But I am also there in real time and can send updated questions to Björn's tablet while he is sitting there talking.

Yes, now you cannot provide real-time updates, but it will be different. Now you get to sit in front of the mic and share more of your own knowledge. But you have an incredible amount of experience behind you. Every time we talk, you have new experiences to add. Everything from running a bar somewhere in Greece

on Rhodes. Yes, on Rhodes, to: yes, I have done this and that and the other thing, and this happened, and a few summers ago we took this trip, and when I was little we were out cycling around Sicily. And I am like: wait a minute. An extremely multifaceted, interesting person to have with us, so I am excited and really want to know more. But let us start somewhere on the commerce track. We know you just mentioned that you help run Core by D. What does Core by D do, apart from horses? You said horse equipment.

Tell us, how is it going? It is training aids for horses. It is like a saddle pad, but we also attach resistance bands that help the body, or the body, the horse, achieve the right posture, which is good both for preventing problems in horses and for rehabilitation, because the training is gentler. For example, hill work, varied exercise and that sort of thing are good, but if you live in Skåne there might not be many hills. Then if you want to train on a hill you might have to, well, transport the horse somewhere you can train it, and that is not always the easiest thing. So it is an option for varied training. I really only got involved in the first place because I started, I used to build Shopify stores, and that was

before they bought Tictail, and it did not even exist in Swedish. Mm-hm. Mm. And then I helped build it up, because at that time I was running Lyckopodden. Around, when was it, coming up to 100 episodes, a woman called Agnes wrote to me saying: could I be a guest on your podcast? And I felt: no, I have talked more than enough about philosophy by then. So I taught her to take over instead, and today she makes a living from Lyckopodden.

Wow. But in that scenario I met a woman called Ulrika, and we started, as business partners, and then we started [unclear in source: hypnotication], which was self-hypnosis to help strengthen your self-esteem, improve your focus and things like that. We met because in connection with that, since I was doing Lyckopodden about personal development, I had released a guided relaxation that was a bestseller on Storytel for about a year and a half in personal development.

The only problem was that all the data went to the publisher, and I could not drive customers to Storytel because the margin is too small, and it drives them to Storytel. I felt trapped, so instead I thought: let us start a Shopify store instead.

Mm-hm. Because she is an NLP expert and does NLP training, which is really, well, we will not go too deep into that. But what year was this, just so people can follow?

Yes, what year could this have been? Good God, this was before Shopify existed in Swedish. Before Shopify, 2018–2019. Yes, it could have been something like that. And at the time I also wanted, I thought: yes, but Shopify did not have a way of streaming the music back then either.

Mm. Because I was like: yes, I want to sell self-hypnosis, but I do not want you to be able to, I want to [unclear in source: hitt], instead of you just getting a Google Drive file and being able to download them. There are two downsides to that. One downside is that people do not understand how to do it on their phones. If you give them a link and they are supposed to listen, lots of people did not understand how to do it.

Whether you had an Android or an iPhone. It was not that simple to do. So we thought: let us build a plugin that is a music player built into Shopify. Then I met another business partner, [unclear in source: Anna Karin], and we spent about 200,000 developing that.

What? Yes. Yes. Yes. But the problem was that we were given extremely low priority by that programming company. Because it turned out that, a fucking long time had passed, and we were like: hello, how is it going? They had assigned an employee to us, and it turned out he had written zero lines of code over the whole, because they can see the number of lines of code written, so zero lines of code over the entire period

and then eventually they built a little more of it, so things almost worked. But it was glitchy, broken, buggy and looked very strange. But it was always: yes, but if we are going to finish it, we want more money. And it was like that the whole time.

So eventually we shut that project down. But that is how, so then we sued the agency. No, it was just a matter of shutting it down, basically, [unclear in source: och sänka sig]. Yes.

Fuck, what a cold shower. But the lessons there have been that I have helped develop Shopify plugins. I have also run a self-hypnosis business. And before that, I tried dropshipping one summer with Mia, who I also work with, and we did get it going. I think in the first month that summer when we got it going, we sold a million's worth, at least.

Absolutely crazy. But then all the rest of the headaches came, because we had the wrong agents. They chose products to buy whose quality varied. So, mm, half the customers were really fucking happy and half wondered: 'Yes, but I ordered from you because...' We were getting repeat purchases too. But when the repeat customers got variations in the quality of the products they bought, we shelved that too. But we used a large part of that money for me to hire consultants, coaches, all sorts of people, to learn marketing, Facebook ads.

Dropshipping, I usually try to put it like this when I am recruiting. If you have worked in dropshipping, that is basically the essence, the peak of e-commerce, running at 360. Mm. Starting a regular retail store from 0 to 100 is actually pretty slow motion. You have to build a brand, set up Shopify. It takes time, and it is allowed to take time. It can take six months if you are bloody good. It can take 12 months if you take it a bit easier and try to learn. But dropshipping feels like, I do not know, ten times faster. You find a product, you have to get it to market quickly. You might have hours or days. If you are a week late, someone else, or about 100 other people, have got there first.

Mm. Then you launch that product and start marketing it, and you see: yes, now 100 other people also have the same product. How do you differentiate yourself in the marketing to get the sales? And then that product peaks and dies just as quickly, and then nobody wants it anymore. So you have to find a new product. So it is still a, it is a new, difficult thing to do.

I do not think people understand how hard it is. It is incredibly hard, and I also tend to see it as, it is not drop, lots of people sell courses saying dropshipping is the easiest way to make money. I usually say it is the easy way to gain experience so you can then build something lasting. For example, at Core by D, we developed the design ourselves.

So we ourselves, together with factories, have gone through everything from materials, and for example we are now developing the product further, working on design protection and things like that. Then you get something unique that nobody else can so easily take away from you. Plus, when you have that experience behind you, I find it easier to dare to invest in something you know will take an insanely long time and be very complicated, but which will make you harder to copy later.

Yes. Entry barrier. Mm. That is true, and the harder something is, like when we look at Västerbottenssåpa, which I mentioned in the introduction here: yes, we have our own proprietary recipe, so the recipe is unique IP. It is very hard for someone else to come in and cook up soft soap like ours. Of course you can make soft soap or another cleaning product and try to make similar ads. You can do that in any industry.

It is very hard to get precisely that consistency and soap quality that we have somehow managed to build on from Sara's grandmother's [unclear in source: för rikaste recept] sometime back in the day. So it is a bloody old recipe, and in a way it is, there are lots of soft drinks called cola today, but there is only one brand that sells Coca-Cola.

Mm. Which has undeniably done very well. And most people who drink cola, I do not myself, but most people who do can taste the difference. Yes. Yes. And Coca-Cola also has something unique in its distribution chain.

Mm. Which is also a way of differentiating yourself. Yes. Yes. Which also makes it difficult to get in, because when they own such a large share, you know, they often own the fridges in some restaurants or shops, and then, yes, getting in is still not as easy.

No, they have done, yes, very well. Let us, let us skip cola. I want to hear more about Core by D. Mm, how did it end up being equipment for horses? I was actually, I was an e-commerce consultant many years ago, and Mia had actually asked people around her, because she was attending an entrepreneurship workshop. She had asked the people running the workshop: I need help with my e-commerce business, who should I talk to?

Mm, and I was the person they introduced to Mia. Then, after we had worked together quite a lot, we realized: this is fucking fun, working together, we should carry on. But at the time it was called [unclear in source: core by dreamhästar]. It stands for dreamhä, dream horses in Icelandic. But the problem is that you have several problems. One is pronouncing the name; the other is that it targeted Icelandic horses. So it was not just horses, it was Icelandic horses, an even more niche thing. And at the time it was a bit more scattered, because they had a horse taxi. They sold saddles and lots, lots of different accessories. And we also sold another product, what is it called, called [unclear in source: Bunger Band]. The problem with [unclear in source: Bunger Band], which even when I sat

with Mia was that my question to her was: how big is Band in Sweden? I mean, what even is [unclear in source: Bung band]? [unclear in source: Banny band] is also, it is really a product similar to ours.

Okay. So it is a saddle pad. And for those who do not know what a saddle pad is, how should we explain it? Well, it is the saddle, or it is what you put underneath the saddle.

Yes, so many people who have seen things like PS of Sweden [unclear in source: Questian] have seen that there are lots of colorful saddle pads. Yes. So lots of people use it as, it is something you have under a saddle. But you can also make it into an attractive design, for example. The difference with [unclear in source: Bung Band] was just that her philosophy was that it should be so cheap anyone could afford it. But the challenge was that at her price point it was so cheap that she had to use that kind of quality. The equipment sometimes broke.

Mm. Then people might say: 'Yes, but the bands snapped,' or something like that. Then she could blame it on: 'Yes, but that is because they used it incorrectly.' When we did our research and started looking at manufacturers and things, we were like: no, it is more that the quality is not sufficient.

Then, as Mia and I discussed, instead of building someone else's brand, which she also probably, I do not know where she manufactures them. I am guessing she manufactures them in China, ships them to the UK, and then we have to buy them from her in the UK, so it is, that is

okay, this is a British brand. Yes, exactly. Which still sells a fair amount in Sweden. Yes, not as much anymore. The funny thing was, when we stopped buying these products, she has never contacted us since.

No, I understand. But what happened then was that we looked at it like: okay, can we not solve everything everyone could complain about? Can we not find the manufacturers ourselves, develop our own design where we simply solve all the problems? Then we developed Core by D and got rid of the saddles, because we had an insanely unprofessional saddle manufacturer. He would send things to us, when we got the deliveries, in wine boxes.

Then one time Mia was like: ah, he has stopped sending them in wine boxes. And I said: oh, what has he sent them in? It says pale lager. I was like: fuck, that is a beer box, then. Then we switched to another saddle manufacturer, and it was the same thing, there were manufacturing defects in the saddles, and you know, buying a saddle cost quite a lot. Yes. And then you have to buy a number of them.

Yes. Yes. Yes. The problem with that manufacturer was that when something was wrong it was: that is not my problem, that is your fault. And we were like: but you are the one who made the mistake when manufacturing the product. Yes. Eventually it ended with, and likewise, you need saddle fittings to get the saddles to fit properly. So we shelved that. Now we have focused, we have stripped away all the other products and just focused on our own, where we already have our own design, and now we are developing it further. That is where we are.

That is part of what we do, at least. A very big, a very big journey to make, from being a consultant to having to learn the whole of equestrian sport, how it works, and then particular details within equestrian sport.

And then you have to manufacture accessories. That is incredibly much harder than you think. Do you have a passion for horses yourself? Because otherwise it usually is pretty difficult, undeniably. Well, when I did my military service, I rode in the parade at the palace.

Oh, right. Yes, so I do have some experience with horses, and I know I have an uncle who had horses when I was younger, although we did not spend that much time together. But I had a support family because my dad hit my mum when I was, I cannot remember whether it was six months or a year and a half [unclear in source: så bodde hon och jag åkte i marken], and he told my mum: you have nowhere to go.

No. And my mum thought: I will prove you wrong about that. So she made sure to contact the police, and we lived at a secret address until I was 18, I think, something like that.

Really? Yes. Back then we were called c/o Eriksson. But during that time I had, because my mum was very ill. Care of Eriksson. Yes, exactly. Because we, even at school I was not allowed to give my real surname, because we were, we lived under protected identities.

So you were called Eriksson then. Yes. For a while. For a while along the way, or quite a long while. Everyone you grew up with knows you as Eriksson. But I wonder, I think the people in the class knew I was [unclear in source: vicklander]. But whenever you wrote out the address, where you were listed somewhere, it was c/o Eriksson.

Oh, right. [unclear in source: Är det så fin pappa hade jag]. Okay. Yes, I will not dig into that. I will not dig into that background and the past. But that is something new to me too.

No. Mm. And yes. Anyway, the support family I stayed with had horses. Mm. Mm. And I stayed with them for support because my mum was very ill when I was growing up and had cancer on various occasions. She has had so many illnesses, nearly died of so many different things, that it is incredible she even survived. So yes. That is one of the reasons for my interest in health. Mm.

Which you have too. Yes. Well, I got to witness it in myself, and you got it through your parents or through your mum. Yes. And of course that is an incredible thing to have to go through as a child. Mm.

And it is, yes, unbelievably nice that it has gone well to a large extent anyway. Mm. And that somewhere along the way you learned and took it on board. Because not everyone does. I know lots of people around me who have barely changed their lives at all. They see what I went through. They see what a transformation it has been, and then just totally ignore it. This will not happen to me.

Mm. Then you just put the blinkers on. This is tunnel vision deluxe. No, but I have met people, even one who had [unclear in source: på mage tillfället på grund problemen], and who was just: yes, but I have sacrificed everything. And it is like: yes, but the whole kitchen counter is full of Coca-Cola bottles and bags of crisps, and you know, it does not really feel as though this person has sacrificed everything.

No. Yes, that is, yes, I do not know what to say. Mm. But okay. How is Core by D going? What was your turnover in 2025? What was our turnover in 2025? We must have closed the year by now. And how long have you been involved in Core by D? It has been a few years now.

Exactly. I can look up how much we turned over. We had slightly lower sales because I tried Facebook advertising, or tried to. We have sold about 2,000 of these in total, and they

How big is the market, then? How many are sold? How many? What is the demand? People usually talk about the tangible market, TAM. What I look at is the number of horses there are

in Sweden. I actually do not remember that off the top of my head right now. But it must be about a million or something like that. But I remember we have a spreadsheet with all these numbers. I can bring you along sometime so you can

see. But what we really did was run a lot of Facebook advertising, and we got it to break even. Yes, we have sold about 2,000 products, and they are around 2,000 each. Whatever that comes to [unclear in source: 2000 * 2000 miljoner].

Yes, roughly, that is what we have sold for. I think it is still hard when you go into a niche within a niche within a niche. Think, one niche in itself is equestrian sport

because not everyone needs riding clothes or saddle pads or accessories. Then within that niche, the saddle pad does not work for, okay. It works for everyone, but not everyone has the same need. Yes. Or you have to educate them about the need, that there ought to be a need.

Mm. Instead of the general thing, which tends to be lifestyle-ish. Yes, but the brands you mentioned, Equestrian or PS, sell a saddle pad that is nicely designed. That is very simple, because you see it, it is a lifestyle image.

Mm. Whereas with educational material you have to change a person in some way, and that can take a great many different touchpoints. So it is not just a Facebook ad, click, buy. You will do your research before ordering something for 2,000. Yes. Yes. Yes.

I see we sold 580,000 kronor last year, which is quite little considering we have sold more of other products in one month. But we are heading in the right direction.

But what we have really done is strip away all the other things around it. That has been the big thing, developing our own design, because, well, we had a horse taxi, so we had a vehicle you could transport horses in. All these things, saddles too, selling off all these things. We had everything from riding tops to saddles to every, I do not know, all sorts of different products there, whose stock we have struggled to sell off and now change. But with the new design coming, we have designed it so that, with the current design, there are tabs you attach the bands to, but those are permanent right now, which means the product is only used for perhaps either training

the horse or rehabilitation. But if you go for a ride, just take a ride, you do not want these tabs flapping around. Mm. So that is what we have redesigned in the new design, which we are currently applying for design protection for, so you can [unclear in source: hana kan] simply remove them and then use it as a saddle pad. That means, for example, our competitors just have something like black in quite a boring design, whereas what we are doing now, in the first phase, is really that we have several different colors and make it so you can also use the product when you are just out for a ride.

Then it is no longer tied to just being rehab, basically. No, I understand. So then it becomes more than just the accessory itself. Yes, exactly. And now that we have come this far, when the design protection is finished, we are looking at what products we can add so we do not just, because we have now removed products and focused on one. The next step will be to add and develop new ones.

Mm. How big is the market, then? Is it all riders in Sweden, or all riders who have a certain specific need for their horses to, I do not know, jump, if you want to enter, what is it called, when you

you mean competitions or showjumping competitions? Yes, but I would say it suits everyone, because, and that is what we need to educate the market about, and what we have changed the strategy for. Because when I ran Facebook advertising together with Mia, I brought in various consultants. I even brought in Nils, who works for [unclear in source: Senormal], and he works for lots of different companies.

So I got help from Nils too. For the setup, or to run it too? Yes, both, and to have him as a coach, to be able to say: I want you to sit with me, and, what am I doing wrong, what do I need to improve? Then it reached a point where even Nils was like: I do not know what you, I cannot find anything. I cannot see anything you can improve. And then I was like: ah, fuck.

Then we need to develop how we work. So what we have changed since then is that now we are focusing on preparing to carry out a research study, which makes it easier to bring more veterinarians on board. And then we have started working more, this is more the organic approach, working more with horse trainers, equine experts, people who work with it more physically. And there we have started to see that we can grow that way, but focus on the slow route first.

Then once we have got that going, we can add more Facebook advertising again, is the idea. You said it was roughly break-even. Yes. Why not keep optimizing by 5 or 10%, and then you get 5 or 10% on the bottom line?

I think it was just that, of everyone I get help from. I have not, you have not looked at my ad account. No, I have not. But you have never asked, which is still a bit strange considering you have been very generous with health tips and shared all sorts of things that I have taken on board.

Mm. So I could, you could easily just have put a computer here and said: hey, could you just look through this for me? It literally takes five minutes.

Yes. But it is just like how I geek out about health. Yes, of course, when you are good at something, reasonably good at something, it takes very little time to scan through it yourself and understand where money is leaking out.

Yes, and that is how it is for me in paid social or e-commerce in general. If I pull up, I can go into a meeting, if the meeting starts at 11, I can sit down in front of Google Analytics, Meta and everything at 10:55.

Mm. Never encountered the client before or met these people. In five minutes I scan through all their data, get it into my head, and can then talk for 60 minutes so it feels as though I have prepared this for two weeks.

Mm. And everyone is like: wow, oh. So there are all sorts of insights in a meeting like that. I have come to understand: okay, this is a fairly unique thing, and it just has to do with experience and gut feeling. And gut feeling is built on the experience you have developed from seeing the same things over and over for 10 or 15 years. And when you have done it for 15 or more and seen perhaps thousands of accounts, you learn to see patterns very quickly and see: okay, this is too low, this, so you put everything in context. Now Core by D would be much harder, because it is not as though I have worked with 100 different accessory brands in equestrian sport, but I still have very good insights because I might have

I have worked with five different equestrian brands. Mm. Some of the biggest sellers in Sweden. And when you have seen their ad accounts, their analytics accounts, their stores, how the traffic moves, you find a common thread somewhere. So it is not that far removed. Then we are part-owners of a brand called ZacZess. We invested in it [unclear in source: 2020 ett slutet 2021. Börj].

How are they doing now? It has gone very, for a while it went well, and then we were hit by having burned a bit too much money on testing, experimenting.

So it kept growing, but we burned a bit too much money, so we had to brake a little. Then everyone on the founding team had children. Mm. So in 2021 I had a child, in 2022 a colleague had a child, in 2023 another colleague had a child, and one of the sisters who founded the team had a child.

In 2024 everyone had children, except me, so all three had children, and I was instead struck by my illness. So it has been tough considering everyone has had a lot of focus on their private lives, but it has still gone pretty well. Now we are profitable, and this year we will, we have our sights set on growth again. First there was a shift to profitability, and now it is just foot to the floor.

Nice. That could be fun. It will be a good journey. But they are turning over somewhere north of 10 million at least. That is really good. Yes, we are only scratching the surface of the beginning so far, really.

Yes, it is, it is an important start. What you learn a great deal from is all those mistakes you make at the start. I think we need to share those on this podcast.

Mm. And especially things like purchasing, sourcing, IP. You are now talking about applying for design protection. How did you go about that? Do you need a lawyer for it? Okay, Meta is not working. You tried looking through your consultant network. Okay, who can you sit down with to learn Meta?

How come you have not, like, asked an AI chat? AI is very good today. I would have got help from Claude, shared my screen, had real-time discussions. Yes, there are lots of things we can improve for you.

Yes, I do not know. But I also think what it has been for me is, when you have done so many things at once for such a long time, and laid a good foundation and redone things. That is why I, we hired consultants too, like: yes, we need new input, because with my current expertise I do not have the understanding required. It might just be a little dial you need to tweak for it to work. But yes, what we have focused on instead is: okay, what we know we can do is the long-term work, building relationships with veterinarians, horse trainers and so on, and focusing on improving the product further so we can then scale up the Facebook advertising again.

Mm. My spontaneous feeling is that you have got into, you have got yourself into quite a tough industry. Mm. Because equestrian sport is very traditional, having worked with it myself.

Yes. Then it is a niche within equestrian sport, meaning one product, and later perhaps more products, that are infrequently purchased products. It is difficult; you do not buy this product every month. It is not a refill product, so you have no naturally returning customer.

Mm. Which is also a negative aspect in commerce. What is advantageous is that the saddle pad accessories work for the vast majority of horses. So you have very few returns, I am guessing. Two in total.

Yes, two. That is nothing. And then it is because someone accidentally cut, in one case someone had accidentally cut into the packaging with one of those parcel

But that does not count. And the other was a buckle that broke. Yes, that is not really a return in itself, it is more a complaint about a faulty product. Yes.

Or-ish. At least one of them. So you have zero returns, which is the nice thing about the business. Then that really means: okay, you have to find new customers all the time, so you need to work a great deal on the aftermarket, after, post-purchase really, and post-purchase recommendations, post-purchase referrals, that kind of loyalty where people love the product so much they want to recommend it to others. And that is very difficult. I know we invested in Twistboxes last summer.

We kicked off that project. We launched it at the start of this year, and we have sold quite a lot of Twistboxes boxes now, roof boxes. Mm. But that does not have a naturally returning customer either. It also has an expensive price point. You need to do some research before buying it. So it is not something you see when someone [unclear in source: spåen] and just click to order: oh, I need a roof box.

I will take this one. Yes, 10,000. Sold, basically. Sold, however good it may be. And then there is an educational aspect in that this one rolls up. Oh, wait a moment. Is it safe? How

will it handle the car journey, or what if rain gets into it? So there are lots of questions about more expensive products before you click to order them. What is your plan for this year, then?

For this year we have our sights set on turning over 10 million, and we are still well on the way. We are not yet running at a 10-million pace, because then we would have needed sales of, yes, at least half a million.

Mm. To be able to say: okay, there is a ramp-up towards the end of the year. Because now we are somewhere at the start of peak season; just before summer lots of people buy roof boxes. So we are not there yet, but we are working on it. Now the whole team except me went to China and visited the factory.

Mm. So Micke [unclear in source: Neldup], who is the founder and a fantastic entrepreneur, took the whole team along, and the whole team is really, he is the one employed by the company, but the team there is the investor collective. No,

it is Stefan, he has a, Stefan [unclear in source: Körling] has the factory down there, so it is our own purpose-built factory that basically only manufactures Twistboxes. It is fucking cool. Then we have Tove, who is my colleague and also growth manager on the project, and then Jenny and her husband went down, who are really good and have previously worked on a great many different projects. Jenny and I met at [unclear in source: Molly], where she was CEO, and then they helped build Gant in Sweden. Yes. Wow. What is her role at Twistboxes?

At Twistboxes she is chair of the board, and really a bit of everything. She is the hub, so she keeps the rest of us in check and makes sure things get done. Takes minutes so we have everything written down. Hands out action points, follows up. And they are so interested that they have now done quite a lot of PR work and got this into all sorts of newspapers. It is a rewarding product because it is an innovation. Have you done any PR work?

No, that is also the next phase to focus on, reaching people who want to write about it or talk about it. We have also tried some influencer marketing on a smaller scale, since we are quite a small company. But we have had mixed results there. What we find we have had the best results from so far is if we find, for example, a riding instructor who trains people, either trains horses or teaches people how to look after their horses.

Mm-hm. Because when they start recommending our products, you get much more credibility than if we talk about them, for example. Yes. Yes, of course. Then it is very small-scale. You have to find the trainers. You have to somehow get in touch, and then they have to want to take on the product rather than something else.

Yes. Yes. So it is quite difficult, and then when they do sell, they sell a handful of products. So getting to scale is very difficult. Yes. And my thought there is not, it might not be tremendously scalable in itself, but if you combine it with PR and other things, it builds a whole that strengthens the brand.

And you undeniably build credibility for the brand. Mm. What PR does is precisely credibility. I really believe in PR when it comes to AI-driven search, because I think lots of people will use ChatGPT for shopping in the future. They already do to a large extent, especially for this kind of product, like a saddle pad or a very specific horse accessory.

Mm. When you buy an infrequently purchased product, people will ask their assistant or their friend in the phone: hey, these, you know more, you have these needs, how would this work for me? We talked a little about fluoride and fluoridated water, and that apparently Ramlösa contained more than the limit value.

Yes. When we were at that restaurant. Yes. When we had lunch, I just took a picture of it, and it knows my medical history, my ChatGPT knows about it because I have my health project called Fedjas hälsa. So I put the image in without a single question

mm, and let it analyze the image and give input, and it knows exactly what I, it is as though it knows what question I am going to ask. Is this, can I drink this?

Mm. That is my summary, and not because I am terrified in this context, but it is above the limit value for the product itself, so you should not ingest far too much fluoride.

Mm. Because it can unnecessarily disrupt the body a great deal, and a can, it says, is okay now and then. Not daily use, not larger amounts, and make sure a large part of the other fluid you take in is not specifically Ramlösa. Hmm. I thought: okay. I had no idea, and I am still very well read. So that was a very, a quick-fix example of an everyday use of it.

Mm. I think lots of people are already there, using their assistant for that kind of purchase or information-seeking channel. Then instead of perhaps taking a picture of the saddle pad, perhaps they see the ad or know they have a need, and ask their ChatGPT: hey, I need to find a saddle pad that would work for X, Y, Z.

Then it starts listing one, two, three, and how it lists them is very PR-based. Partly it searches forums, the organic stuff like Reddit, and sees what is being written about a lot right now. Then it looks at authority, it looks at SEO, which sites rank highly. Then it looks a great deal at, and takes credible sources into account, and recommends based on that, because then it can always cite sources. You have probably seen that on ChatGPT, when it cites sources for things.

Often they are very credible sources. Sometimes they are not. It might be some random forum, like Tesla Club Sweden, if you have a question about your Tesla or something else, where somebody has written a post. But sometimes it is: okay, now we have gone straight into Tesla's user manual.

Mm. That is cool. Yes, I use it, I know especially when I buy supplements, I am the same. I regularly go and give blood samples, stool samples, urine samples, breath tests and so on. It knows about my gut flora, my vitamin deficiencies and things. It can see at a completely different level, a completely different level, because it has such extreme access to information. It was the same when I was buying supplements recently. I was like: yes, but according to the test it says something here about B5, and I tried asking it, and then it could explain the whole chain of, yes, I needed, because I was thinking: can I not just buy B5 or a multicomplex? And it was like: no, but based on your blood values and the chain you have, you specifically need [unclear in source: B5 P5P], because it is the variant your

body can absorb more easily without converting it in this way. And if you combine it with copper bisglycinate and [unclear in source: metylfat], then you will get a better out, a better outcome from it. Mm.

How fascinating. And when I, then I could even ask, I went into different shops and photographed the products, the product descriptions, put them in, and then I could get help with exactly which supplier I should use, so that it

Yes. And that is how we will live in the future. So shopping has already changed. Specifically for products that are knowledge-intensive and require some research. That is where you will ask these kinds of questions. Lifestyle brands will not get that impact here and now. A Mockberg will not be affected by you asking questions about which watch you should buy. It is like, yes, there are lots of different lovely watches out there, but Mockberg is quite a very lovely brand.

That is not the kind of question you will ask ChatGPT. But I definitely think riders will have, so I would recommend PR. It costs nothing apart from time, obviously.

You can prewrite articles and send them specifically to journalists. What we do at Twistboxes is perhaps find a digital newspaper, then see: okay, who writes the articles? Then we download 10 or 15 articles by that journalist, which they have written about other brands that we think are reasonably similar to what we are after. Download them, upload them all to ChatGPT at once, and ask it to learn how this writer does it. The journalist's writing style, what is the style they write in?

What prompt do you usually use, then? It is a very simple prompt. Just: can you analyze these texts, because they were written by a journalist on this, this site? Also go into the site as a whole to understand what words, language, usage and level it has. Because there are different kinds of, yes, levels you can pitch it at. Is it more advanced and nerdy, or more accessible and general, or what level does this site have?

Then you have my product, and this product, and this is our site and my story. You already know that. Can you now write an article that would fit perfectly on this site, at that sort of length? And then it does it, tap tap, with quotes and everything already finished. Mm.

Once you have that text ready, you can send it specifically to that journalist, and they love it because they really have very little time, so they can have the whole job served up to them.

Mm. Then they copy and paste. In some cases they put your entire article in as is, if it is not, it should not feel sponsored, because it does not either, since you have taken 10 or 15 older articles.

Mm. Once you have taken them, you can crunch them and make a new one. The journalist uses the entire article in the best-case scenario. Otherwise they might take parts of your article and write something themselves, or rework your entire article in their own ChatGPT and publish it so that it still fits the site.

And however they do it, they have saved time. There is a difference in doing a job like that, [unclear in source: ett väldigt], and then writing quite a specific email: hi, this would fit perfectly on your site.

How do you usually formulate the email when you contact them? As personally as possible, basically. Hi, I read this article and that article. I saw you published this yesterday. Really fucking good. Thumbs up. I have followed you for a while. I run this and this. We solve this problem. If you would like to write about it, I have already prepared an article here for you. Done.

It must not be too long, because as soon as an email is longer than the average few lines, you leave it for later. You read, you might just scroll through it: ah, I will leave that for later, because it is too long. You cannot face replying, cannot face tackling it, because you have, everyone has a lot to do in life, especially if you work in digital press. You have an enormous amount, lots of emails. If the email is too long, you save it for later. So short and concise, with the finished article already attached

either in text form or as some kind of attachment; attach a PDF, that is easiest. Fucking smart. It is simple and you get it out, and you have done half the job for the journalists, so it becomes easy for them to publish. The point of them publishing it is that you can use the newspaper article itself and add traffic to it. But you can also have the newspaper's logo on the site, which creates credibility. Credibility increases willingness to convert and makes people feel a stronger sense of connection with the product. I mean, I might recognize IDG if it is a technical product, or recognize DI if it is something in finance. Well, nice that they have written about this. That is credible. I trust it. I have read, I read about it like that at least sometime, and then you click to order it.

Do you usually have the logo link to the article too, or just display the logo? I never usually link out from the product page. I try to, once you are on the product page, you should stay on the product page. So you do not want to send people away to other pages.

Mm. So the logo itself will not link out. But it has happened that if I get a really good article, say on Breakit or DI or, well, [unclear in source: svensk närings alltså någonstans Sv], ads to that article, long, somewhere I think they read the ad and then it is about the product or they gain more awareness of the product. So just driving traffic relatively cheaply to the newspaper site works too. But then I am thinking of, I think it is Fisher Investments, as an example. They target people who are well off.

Yes, I get their ads all the time. And then I look at how they advertise in Aktiespararna, and almost all their social posts do not link to their own site but to their native ads on Aktiespararna.

Yes. And that is next level. What you do then is buy a native article, so they usually pay for it, or you buy a sponsored article where it clearly says it is sponsored by Breakit or something like that.

It does say that. Yes, but it clearly says it is sponsored, and you see article after article after article. Yes. When you have already bought an article that might cost 5, 10, 15 or 20,000 kronor.

Mm. But nobody reads it. What is the point? Yes. Drive traffic to it. Mm. Spend at least as much on driving traffic so people become aware of it as you did on writing and publishing it and everything that involves.

Mm. So that is, yes, but you can also create that kind of article yourself through listicles, like creating your own article on your own sales page, as it is called, another site. It might, if you sell saddle pads, it could be called [unclear in source: bästabraken.se]. Then it is a test of the different best saddle pads on the market, and your saddle pad happens to have a great deal of information, clear text, so that even if it perhaps ends up at the bottom, people click through to it.

Smart. But would you use that for, because I am also thinking you can have influencers say: five tips for the best skincare. Yes. Then the third one happens to be a product you sell.

Mm. As an example. But now I am thinking about Twistboxes, because you also have an expensive product and need to educate people about it. How do you do that? Because you have different markets. One might be the family with children who simply do not have room in the car. It does not have to be that they are traveling any particular distance. Or you have people who surf a lot, or people who ski.

Yes. What is the plan for reaching them, then? We have done some thorough work first on the backend structure in Shopify so we can easily launch different frontend sites, and we will have different ones. Right now we have a couple, but in the longer term there will be lots of different sites. So the frontend will differ depending on the target group we want to reach. If it is families with children in a big city, there will be a specific site for that, because we know that page should appeal to that target group. Partly copy that should appeal directly to the mum of small children in the city, or there will be different pictures, so you have lots of small children in cars, lots of cramped living in the city. You have very little room in the car because you have to pack two strollers. There is no room for anything else after that. Even if you have a

big Tesla [unclear in source: modell i], you cannot fit more than two strollers and a little air. Okay, then perhaps you want that extra space on the roof, with 520 extra liters. Especially if you are going away for a bit. You can make a great deal of content about that. The problem, the problem is, now there is AI help, but it is still very difficult to get AI to generate this kind of content, and to do a photo shoot with a family with children on a particular day when it is warm or cold, depending on the weather, because you do not want the same content all year either. It is not that easy. Mm.

So you need to find a surfer, a camper, have a family, a family with small children, and so on and so on and so on. So there is content work that has, that takes more time than you think when you get into how much you need to do. Then you want different things depending on which market you are in. If you are in Sweden, perhaps you want more people who, yes, look as though they come from Sweden

because people relate to people, and it is much more credible: okay, this looks like me, I will order it. And if you are in Asia, you want another kind of content. We sell globally, so first of all the focus was to launch in English, which we did in January, at the end of January. Then translations for local markets take a couple of weeks, because even with AI you need to proofread and look at policies and cookies, all sorts of things.

Yes. Do you use Weglot, then? Yes, Weglot. Mm. So we run Weglot connected to ChatGPT, and then you need to localize certain markets, because you know Swish works really well in Sweden, but it does not work in Italy.

Mm. Okay. It does not really work in Germany either. So what payment method should we have in Germany? What payment method should we have in Italy? What payment method should you have in Asia? Then the next step after that is: okay, now we have that, what about shipping methods? It is really fucking dull if you only offer one shipping option. Ideally you want a local shipping option too. And when you start adding the complexity that suddenly you have lots of different frontends, lots of different markets, lots of different, it gets very difficult.

But that is the structure we built at the start, so that we can, it is not easy, but relatively easily duplicate and redo this for different parts of the world. Mm. But how do you handle warehousing? Do you have third-party warehousing or your own? No, one of the investors, co-investors, is called [unclear in source: Packit], and they have their own warehouse here around [unclear in source: Eskel mellanskolan Norrköping].

Mm. Now I do not know 100%, I think it was called Kungsör. Mm. And they pack for the Nordic and European markets. So now we have a factory and our own warehouse in China, which Stefan [unclear in source: Körling] manages and has. Then we have a prospective partner for the American market. We will have our own warehouse there. And then we now have distributors around the world. So the American market can be handled by a distributor. Then he has bought X number of Twistboxes, or our subcategory brand called [unclear in source: Weoll it]. Because we do not want to, we should not flood the market with Twistboxes either. If a distributor wants the same product or an equivalent product, we have another brand called [unclear in source: Weoll It], and they can distribute it through their channels, while we sell Twist

Boxes globally on our own website with advertising. What have you done in terms of design differences, then? There are very small design differences, but you can see that it is [unclear in source: en weit]. It is a different logo. There are small design differences, but fundamentally the same product, or at least an equivalent product. So it is not that the product differs hugely; it differs in design and execution, then sites, then how it will be communicated, and price-wise it should be roughly the same.

Mm. But I was thinking, because you mentioned when we last saw each other that you have also been focusing on recruitment lately. What are you doing there now? Which key people are you looking for, and have also found?

It is always recruitment that lets you down. But you fall, and not that you fall, but things go well or badly depending on the people you have around you. So those, the people you surround yourself with are somewhere the sum of what you then become. And recruitment is incredibly difficult. I want to put a great deal of time and love into it, because I know that when you do that, the outcome is usually better. But it is often something you do well for a while, then you find, then the next thing comes along that is some [unclear in source: fire gray], where you have to put out a fire somewhere, and you let it go for a few weeks, then go back, then let it go for a few weeks. So I want more time for it, and it is almost the most enjoyable thing, to

work with the people you have around you. But we have recruited, we recruited two key people to Såpa in the form of [unclear in source: avan] and Andro, who have been really good and worked all of Q1 now with Västerbottenssåpa. Then we have a content person in mind, and that is easier said than done. It is fucking hard to find people who can both stand in front of the camera and edit, and be a bit natural so it does not feel rehearsed or scripted or fake, so they are a bit of a natural in front of the camera. Fucking hard.

How have you handled it so far, then? So far we have not found anyone, so we are recruiting through contacts. I have not even posted about it on LinkedIn. It is easier to find, yes, now we have a growth manager. That is quite a standardized title, but saying: we are looking for a creator, whose role is quite vague and hardly exists in any job title description anywhere, yes, much harder. But that is what we are looking for right now. Then growth managers are always welcome to apply to us, because we are always looking for generalists. We talk a lot about this on the podcast: I think the future is very much for people who can connect different data points.

The generalists who have found it easy to see different systems and connect them. Even though we know an AI will be able to connect different systems and deliver an outcome, it will be a very long time before it can handle it in a case the way a human can. Among specialists, whatever you do, whether you are a media buyer or copywriter or SEO specialist or content editor or social media management or influencer marketing manager, whatever title you have had in e-commerce, it is easier to be replaced by an AI agent or at least automated to a very high degree.

Mm. And when that happens, the role changes in a way. So at Porobic Group, at the agency, we require people to take three steps forward every year now, roll up their sleeves and say: okay, which tools can I use to amplify my work so that I do not get replaced or become replaceable?

Mm. And everyone works with AI. I think the difference some people imagine, because I have heard some talk about AI being seen as cheating, that it is not you doing the work, so to speak. What do you think about that? No, I think it is very strange to think it feels like cheating.

How, what does the reasoning usually look like? But I think it comes from school, where you were not allowed to get help from others; you had to memorize and understand the knowledge yourself.

Exactly. So I am guessing that is where it comes from. Whereas when you then enter working life, it is more about: how quickly and cheaply can you create the best possible results?

Yes, true. And that is actually a slightly different view of the whole thing. If your perspective is that of the school's [unclear in source: old school screens], mm, then I understand why they see it as cheating, because it is not you doing the work. But then the question is also, is it, because I was thinking about this the other day while prompting: am I prompting the AI, or is the AI prompting me? Because sometimes I can ask it to do research, then it comes back with suggestions for me, and then I get going, looking for new software and connecting new things.

Like this weekend, I sat reading about ISO 27001. Because AI gave you the suggestion, or? Yes, simply because I said I want to work more on security within AI in the platform I am building

and then it could go through: then the code must comply with ISO 27001. The new one within AI that has just come out is 42,000. I think it was 42,001 too, and I have forgotten right now what the American version is called, because the EU and the US differ a little depending on which companies you target, whether they are American or European.

Oh, right. But I would not have spent my weekend researching it if it had not been for, and listening to podcasts about it, like Vanta, with a certifier, they help certify it. I would not have sat listening to that if AI had not got me into that field.

Yes, I actually have not even thought about that. This is, like, next-level meta, a meta level where, okay, you prompt AI, but AI thinks one step ahead and gives you, because often it then asks you a question back.

Mm. Do you want me to do this, this and this next, as the next step? Yes, I do. And sometimes it is right, basically. That is what you were going to follow up the question with. So it is very relevant nowadays

and of course it can find areas you have not even thought about. Mm. And then prompt you to, yes, do everything. Yes, because then I have to start reading up on rules and understanding and so on, in collaboration with it. Listen to podcasts so I understand the difference, even database structures and things. But it is the same. I think bad content can come from when you just ask it to do something and do not give it access to any previous context, and then it becomes AI slop, as they call it. Some generic mush you cannot really do anything with. But just like you with health, I have a health project. I use, I use Claude, Gemini and ChatGPT, and sometimes I let, I have connected them so they have cross-

conversations with one another, because the three of them get to discuss back and forth to reach the most optimal suggestions. Because what Claude has thought of, perhaps Gemini has not thought of, and ChatGPT has not thought of, so they give different suggestions. I can spend many hours going back and forth between them. But then in my health project, one of the main instructions is that you need everything, because in the beginning it was always: this is the ultimate answer. This is what will change everything. But then I put in that you are a functional medicine doctor and constantly ask questions to understand me better, so you do not just blurt something out. And that has also given me, through it being programmed to

make me explore things, I learn a lot more. It does not turn me into a headless chicken who does not think for himself; I develop and learn faster. Thank you for that.

I like it. That was a, I usually have that too when I start prompts: okay, I want to do X, Y, Z. Mm. This is the goal. This is really the route there. Ask questions so you learn, or so I can fill things in before you spit out an answer. Then it asks, like, ten questions. Who is the target group, and what does the budget look like, and

okay, is there a particular market? And it learns. But if you had not asked that, it would spit out an answer. Mm. Because that is what it is programmed to do.

And that answer is usually very generic. But once you ask the question, ask me questions until you feel satisfied, it will ask questions. Then suddenly it is no longer a generic text; it becomes very specific.

Yes. And you can also set guidelines. You can program AI, which I have worked a lot with, because I work with things like database structures. Then there was a bug that arose, which it managed to create itself when it was changing something, and I was, by the end I had been sitting there for 14 days, almost 15 days, and it was completely: this is the right answer, now I have finally solved it, now it will work. And I sat there like: what?

No, you will not have done it, because you have said this day in, day out, day in, day out, day in, day out. And I was like: will I ever get out of this? But then I learned, I sat discussing among the three AIs why they behaved the way they did. They gave me the answer that I used to write a handoff, so if I end one chat and want to start the next, because they have a limited context window. That is, the amount of data it can hold in memory at once before it starts compressing and forgetting things. The problem is that we do not know what it forgets.

That is what means that sometimes, if someone, if you chat too long in the same chat, you can start, you need to repeat yourself over and over again. But what I did was always ask it to write, I thought that was my hack. Like: yes, I will write a handoff that I can just copy and paste to start the next chat. But what I did not know, the mistake, was that I thought, with Claude, when you have the Max plan you have a one-million context window, whereas ChatGPT has 200,000. Think of it as really the amount of data it can remember, the amount of text, simply. Because if I sat saying the numbers 7, 8, 6, 5, 4, 3 and asked you to remember them, then kept going with new numbers, new

numbers, new numbers. At first you will remember the first numbers, but the more numbers I add, unless you have used Jonas [unclear in source: Fessens]' memory techniques, you will start forgetting.

Yes. I understand, and that is how it is. But then I thought: yes, write a handoff letter. But what I did not know was that there is a token limit on the letter it can write, so it had not mentioned all the things it had tested before. That meant it tried to repair it using the same thing it had already done, because even though it has a large memory, the letter it gives to the next chat is limited. The solution was really to ask it to keep regularly saving to a memory file: its reasoning, what it had tried, what works, what does not, and all the connections it had made, for diagnostics so it could understand. In that way it could develop itself further and start setting its own rules. So then it was that

the rules must not be too loose, because then it starts hallucinating. But they must not be too strict either, because then it starts making mistakes because the rules are too strict. That was what I used, I had to use all three AIs together to solve it and move forward.

Fuck, that is insanely interesting. We almost need to do half an episode just on this, to see how you have built it. You can demo it for me before the episode so I can ask good, relevant follow-up questions.

Yes, we can do that. But then suddenly now I can ask AI, and I am developing [unclear in source: smoothly] further, because I have built AI to build itself. So I have had fucking great fun, you, I showed this music player it had built. It was not me who built it, and not me who prompted it; it is the platform itself that has developed itself further so it builds things and tests what breaks.

So I had fucking great fun, and that was when I showed, I sat watching the screen and it had just built a weather app at random. Then it turned out it wanted to animate the sun in the weather app so it rotates and pulses, just to test whether, if it makes this kind of animation code, the build will crash or not. Those are the kinds of rules I have made, so now it builds and develops itself further. I have it on a side screen, which is fucking awesome.

Okay. Sounds bloody expensive, so the credits must just be flying. Yes. They burn pretty quickly. Oh, right. But that is why I am training it to then be able to build something else, because then I teach it all, it learns the design rules, the code to build according to the design rules without anything breaking. Eventually it will have, what is it called, secured most of the cases where something can go wrong, which guarantees it creates a better result. Or guarantees, but increases the probability, at least.

Mm. Fuck, that is crazy. Okay, we will wrap up soon, but before we wrap up the podcast. Mm. There has been a lot in this first episode together with you, Fredrik. Mm. Nice to get to know you better and look forward to more episodes where we dive into AI, Core by D, e-commerce and everything in between. But before we do, when I told you Jacob was leaving, it was immediately like: ah, could you not, could we two do the podcast together? Why did you think, or what were you thinking then? Was it not a little too hasty?

I thought: fuck, running a podcast again is a big commitment. You did Lyckopodden for 100 episodes. Mm. Why do you want to be part of 0 to 100? For me it is more that, first of all, it is really fun spending time together, and second, I am extremely interested in the field, and it has the advantage, and I also feel that even though perhaps I have not built everything myself, I can share those things. I meet incredibly many entrepreneurs in all the different places where I help produce things and learn a great deal. So I still feel I have, it is more like perhaps I prompt you, so I can bring questions for you and you can answer them too, but also ideas and that, and

I also share as much as I can, openly, about AI if people want to learn about it. Yes, we do. That is about the hottest field for me right now. So I would gladly take a crash course, and we can talk more about it. Because I think one thing we would like to take the podcast towards is more guides, tutorials, operations, how to do things live. You would almost want to share a screen so people can pause, rewind, see how we built this thing.

Mm. That would be a cool episode to see. I have followed another American podcast that has gone a long way in that direction, and I think it would be nice just to do that for one episode. Then we have other things going on. Another track we have thought about a lot that has not happened is that we ought to post more reels on TikTok, YouTube Shorts, Instagram

so we get the podcast out to an even bigger, a larger audience. But above all, the clips spread outside the e-commerce sphere, because right now there are lots of you who listen very devotedly to every episode, and you are all from the industry-ish, or around it, but we reach very little outside the industry. I think as soon as we start posting on these platforms, which you have done very well with Björn's podcast, I think it will get wings outside the sphere. Then it would be fun to talk a bit more about it and geek out even more about personal health, because I have had all sorts of feedback saying people want to hear more about that. If you are listening to this episode now and hearing Fredrik's journey, ask questions in the comments or send a DM to

me and we will take it privately, because I know commenting on something posted can sometimes feel a bit scary. I do not understand, I cannot relate, because I am very open myself. I have no problem leaving comments on other people's videos here or when I have questions. But I understand, because I have talked to lots of people close to me who watch everything, consume a great deal of content, but do not want to comment and send a DM instead.

So send a DM if you are that kind of person. And thank you for listening to another episode, episode 102. We have just turned a new page, and we will be back again next week. Thursday morning, same time, 0 to 100. Now we take the next journey and try to take that all the way too, to 100-plus million SEK. We will see you together with Fredrik Ankarsköld. Thank you, Fredrik.

Thank you very much. I was thinking, instead of Filip and Fredrik, it is Fedja and Fredrik this time. Yes, I had not thought of that, but yes, Fedja and Fredrik. Yay. Bye

bye. Mm.