A hacked Meta account, rising shipping costs and ecommerce bottlenecks #112

A very warm welcome to a new episode of Zero to 100. With me as usual is Fredrik Ankarsköld. Today we have so much to go through. New week, new start. How are you feeling, Fredrik?

It feels good. We are going to jump straight into the topic. It has been a couple of weeks since we introduced you as the new podcast host, so we need to do a checkup on how things are going with Core by D, the e-commerce company you run.

What are your biggest challenges right now? The biggest challenge is that there are so many things to work on. That is why I mentioned earlier that I am reading a book Elon Musk and his former president Jon McNeill use all the time: The Goal by Eliyahu Goldratt, if I manage to pronounce his name correctly.

The book is really about there being loads of things to improve. What I am thinking about is: of all the things that can be improved, what should be improved, and why? They talk about the theory of constraints. There is a bottleneck somewhere, and if you optimize in places that are not the bottleneck, you can waste time and money. In his case, it was very much about production when he was alive. Now his son has taken over.

If you optimize a part that is not the bottleneck... Say we can produce even more parts at this station in the factory. But if that is not the bottleneck, you get what they call it "piles up." Yes, you produce more parts, but if it is not the bottleneck, you just get a load of parts accumulating somewhere else in the flow itself.

So instead, you find where the actual bottleneck is. How is that for your e-commerce business? Right now, I am in the phase of trying to learn different methods for identifying the actual bottlenecks.

I usually say e-commerce is fairly simple. You have three main branches once you have nailed the product. You seem to have done that. You have to drive traffic to the site, get it to convert and get the traffic to come back. Acquisition, conversion and retention.

Most often, acquisition is the bottleneck. But you might pour in traffic that does not buy; then conversion is the bottleneck. Or you pour in traffic that converts but never comes back. The product is substandard, the unboxing... You do not respond, you dropship and it takes too long. Then customers do not come back. It becomes difficult to create profitability over time, and retention is the bottleneck. Which of the three is your bottleneck right now?

One of our bottlenecks is that the product is good enough. Or good enough—it is so good that we do not get any returns. But they do not...

It never breaks, so nobody wants to buy again. No, but we have just had the design protection approved in China. Nice.

Europe is on its way, and we are also working on the application for the US. The process is underway for all of them, and lawyers are helping us. Let us pause there. How expensive can a process like that be, getting protection approved in China?

Now we are talking a few hundred dollars. Okay, not too bad. No. Then all the details have to go through. We have developed a new version of the product where you can create different designs when you are not rehabilitating or training with the bands. The current design is based on having to keep the bands on; otherwise, you have the clips flipping and flopping around.

We have developed a new design, but before we launch, we want the design protection in place. What does that mean in practice? That other Chinese factories, at least in theory, should not be able to manufacture the same product and sell it to other brands?

First and foremost, it is also about them not preventing you from having someone manufacture it. On Framtidens E-handel, which I help produce, we recently had another lawyer on, also named Fredrik...

...who explained what I learned a long time ago: there are pirates in China who, when your sales reach a certain volume...

...protect your design, your name and everything. Then they can prevent you from manufacturing in China unless you pay compensation. Clever business idea. Very nasty.

Very nasty, but it is something... I did not know that was a thing. It is a thing. So part of it is preventing others from preventing us from doing something.

Essentially so you can produce as you intended? Exactly. It has been approved for a few hundred dollars. Not too bad.

No, per territory. In total, perhaps 15,000 kronor if you add it up. Not that much. What design protection do we have in Europe, the same procedure?

It is the same procedure here too. A few hundred dollars. Back to the bottleneck: you are doing a lot of work to protect the product, copyright it, so you can first produce in China and then sell in Europe and the rest of the world.

Is that the bottleneck? For us, the bottleneck is really research. A large proportion of the target audience we want to reach consists of veterinarians, equine science professionals and people working with horse training.

Many want research behind it, which is reasonable. That means we are sometimes passed over in favor of competitors. Even though the research there is limited, there is something. We are working on that. Getting a research project through is not...

...super easy. It cannot be super cheap either. No, it is one of the things we are working on a lot. Often, you can get grants from municipalities and county councils...

...or universities if you have a good proposal. Have you submitted it to the University of Gothenburg or something similar? We have, but the response has been that we...

...must have veterinarians selected in advance, already in the funding application, who will be part of the research we want to do.

The group, like the test group? Those carrying out the research, I would say. That has been the response, so we are working on it right now. If the target audience is professionals...

...acquisition is part of the funnel right now. To do that, a prerequisite is needed: research support or research-based evidence to sell the product to them.

Have you asked the veterinarians: if we had research evidence supporting the thesis, would you buy the product then? Yes or no? Then you can get a purchase order.

A confirmation, a nonbinding purchase confirmation by email, saying: when the condition is met, the research support shows X, Y and Z, we will buy in and pay for the products.

That was actually a good idea. What we have done so far when working with veterinarians and equine science professionals is... Most think they are elastic bands, so if it works on one horse and the bands do not differ particularly much, it ought to work on the other.

Over time, we have built trust through people using the product in real life and noticing there is no difference. Okay, reasonable.

But that was a very good idea; I will take that with me. I have also looked at writing in advance. In earlier episodes, we discussed how you find journalists who can write about, for example, Twistboxes...

...and then prewrite articles to send them. For us, there are limitations. We want to reach out in Germany, where there is a limited number of journalists to write to. Many of them also want research.

They want something to lean on. Exactly. Instead of research, you could do surveys. Those are usually something you can lean on a lot, and they do not need the same amount of data or the same... Out of 100 equine science professionals, veterinarians or riders, 89 say it works for their horse.

Then you can express that as a percentage. You do not need to write the number. "89 percent say it works." Do you understand? It becomes support from some kind of study...

...that is self-funded and self-conducted. You do it yourselves. Do you mean contacting 100 and giving products away? You might not need to give them away. If you find a school, you can test there. They might only need three units, or one they test on 50. Then you have a survey they fill out and enter a draw for a unit, so you do not have to provide so many. It costs a lot to do.

That brings the cost down and gives you data collection. They get to say what they think of the product, and you get qualitative data and research. You can publish a study quantified across 100 or 200 people. Then you can publish your own blog article.

Above all, you can send the article to the journalist and refer to the data. They like picking up statistics. We did that at Västerbottenssåpa: we surveyed Swedes' laundry habits.

How was the survey carried out? We asked questions about how often you do laundry, what you wash and which detergent you use. Then we took the survey and published it on our own blog.

We linked it to different sites. I think Svenska Dagbladet took the bait and wrote "This is how Swedes do laundry," linked to the blog article and used the exact statistics from it. It was no harder than that.

How did you do the actual collection, how did people respond? I think what we used was called Typeform, or Google Forms. We sent it to the email list of people we know have bought laundry detergent, because then we know they do laundry. What did the email say to encourage them to respond?

For us, it is enough for a survey to go out. Many people are engaged and want to respond. Of course, you can give incentives, but we do not do that. Not in these particular emails. I cannot quite remember; it was probably six months ago, so I do not remember exactly. But I do not think we gave incentives. Even if you did not, you have something encouraging them. Is it getting to help shape things, or finding out statistics? Nothing.

People just... Of course, it says a blog article will be published, but you will do that whether they respond or not.

People respond regardless. Not everyone, perhaps. If we really want a large proportion of those we email to respond, we can encourage them with a discount code at the end, a gift with their next purchase or something else. Sometimes we want to gather qualitative data...

...about the products: how do you experience a particular or new product? What do you want in the range? Which scents? For that kind of survey, I think we can give something away to those who take the time to respond. It can take 10 to 15 minutes to go through a questionnaire, whereas what I am talking about for you takes two minutes to click through.

So it should not take that long to gather the data. But when you have a survey that takes 15 minutes to answer, what do you usually do? You want to get to know customers better. There is a risk if, for example, you ask for criticism while also saying they get a discount code...

...for their next purchase, but someone is really angry; they do not want a discount on their next purchase. True. Then you attract happy people...

...to respond. That is the downside of that kind of incentive. It could also be something else. At the same time, those who are angry, annoyed or want improvements tend to be overrepresented in forms, and in reviews too. If you have had a bad experience, you are more inclined to share it about the product or company.

They somewhat cancel each other out. It is not entirely scientifically proven, but it is easy to see that the extremes are overrepresented at almost every company, across the board. Either fives or ones. Rarely threes for products: "It is a good product but with some room for improvement. I might buy it again if I do not find anything better." Reviews like that are hard to find, unfortunately.

What bottlenecks do you have right now in the companies you work with? There are any number of them. At Mockberg, we opened a new store at NK Gothenburg, larger premises. We have gone from a small section there to quite a large store for little Mockberg. The company has grown, and we are trying to take greater shares. It has gone well. We have historically collaborated with [unclear in the source: slash], taken over some staff and are working more with in-house staff now. It has gone well...

...and we want to expand. Things like that take time from the organization and management team. You have to think about things and make plans, drive traffic to the store and look after the store staff. It is fiddly, but once you get it right, as I think we have, you get good results.

But it is not a bottleneck in itself. One of the biggest is getting some international countries working. We have got things moving really well in some and not had nearly the same effect in others.

I saw one country increase by 180 percent, while in another we barely have any percentage increase. Why is the difference so large? It ought to be fairly similar, not such a big difference. But you buy Mockberg, a lifestyle brand, largely because of the brand. The longer you have been in a country, the more time you have had to build the brand. That is why it is fun to help legacy brands that have not done digital marketing very well. You see a quick effect when the bottleneck comes all at once.

It is much harder to start from scratch, like Twistboxes or Core by D. From zero. Here you take a brand that is completely unknown in the Netherlands, Germany or Italy and are going to launch there. You have substantial resources, but also an organization focused on other parts of the world, which has launched in other parts and has to continue growing there first...

...before dipping a toe into a new country. That is a bottleneck in itself. At Västerbottenssåpa, it is mainly production challenges again: product and logistics. We have underestimated demand for popular products and have to cap things. Demand might be 2,000 products a day, but we can produce at most 1,200 to 1,300. When production falls behind, we also cannot manage logistically to send things down to the hub in Morgongåva, which gets everything out very quickly.

Then in turn, it takes a long time to ship from Morgongåva to customers. We have to ship directly from Vilhelmina. That is a bottleneck, because we get a backlog of orders that a small team there has to ship. The same team is needed for production, packaging and other things. It is not that small, there are quite a lot of people, but relatively small. In Morgongåva, we already have the plan and have expanded with the logistics team.

It is irritating, because we have to lower marketing costs and adjust quite a lot in marketing. We really always have the challenge of bringing customers into a brand, but at Såpa we have had production and logistics challenges since day one. It sounds as though you could read the book I mentioned, which is about exactly that.

When you were talking, I thought it sounded good for some people on our logistics team. I am not in production or logistics, but we have bottlenecks there. You think: now we have brought in a machine, that should solve the bottlenecks.

Then the bottleneck appears in the next place, also in production. Theoretically, we can produce 3,000 of a product that sells really well. But in the preceding step, only 1,200 to 1,300 can be filled, so the other part becomes the bottleneck. Once that is solved, the cans, bottles or pipettes run out. Something runs out, because we also have relatively poor control of purchasing.

You think: how hard can it be? But it is a difficult operation to coordinate, with many small components and steps. We are a small team that has grown quickly, so it is natural for it to happen. The important thing is to plug the gaps and make sure it does not happen again.

The next season is coming, and we have both the season and growth to take into account. Right now, we went from selling I do not know how many thousand orders a day to having to cap at 1,000 to 1,200 at most. That is what we can produce, pack and ship. Otherwise, we get unhappy customers.

That has been the big bottleneck. Then we have two recruitment processes underway. We have selected a recruitment agency, started screening and got the job ads out. Why did you choose that particular agency?

It was a thorough process. I have a little soft spot for E-commerce Recruit because I know the team behind it. But we did not choose them in the end. A group of us on the management team has met with different agencies. Another agency was better at senior roles than perhaps E-commerce Recruit, which is very narrowly focused and even has e-commerce in the name. They know the e-commerce industry well, but fundamentally Såpa is a production company. A CFO and a Head of Retail are quite far removed from e-commerce. We have that well covered. I might want the person to have some touchpoint with e-commerce, rather than just a logistics specialist coming in.

That said, together we chose another agency we are very happy with so far and hope things will go well. It is difficult, because they all feel same-same. They sell in the same way, cost the same and work the same way. In the end, you are the one interviewing and choosing the candidate. It is not about price, but what network we believe the agency has within the role. Here, it was two senior roles outside e-commerce, so we chose another one.

We will see how it goes; I can elaborate on the podcast. I also checked references. The industry is small, and I know someone at every company where they have reference assignments. So I could call someone, without naming names, a week or so ago: "Hey mate, how are you? All good?" "I am at a bachelor party right now." "Okay, five minutes, I just want to quickly check what the recruitment agency you used six months ago was like."

Do you get a thumbs up or down? Do you get recommendations? Which people on the team did you use? All consultancies depend on individuals, so you want to know which consultant they were happy with. Then we can get the same one, or at least have that person on the team.

Here, it was easy to check references on everyone, so I really did that. I called around and asked, and got very good references for the agency, so we moved forward. That is a typical bottleneck: we need a Head of Retail to expand in retail. We have 180, soon 200 stores. We cannot manage that with just AI, apps and one part-time person who does not have the experience of rolling out broadly across Sweden.

We can link to the books we mention, or list them in the show notes. Easy. I will take that book recommendation too. We talked about the book a few episodes ago, so we must at least skim through it.

I am automating a lot of processes myself to try to build an AI-native company, as they call it. A brain that keeps track of everything. Similarly, an LLM or the ordinary chat does not fulfill that function alone. An entire system is needed for it to work. That is why I am also nerding out on production processes. When so much can be done, I am really trying to figure out what actually moves you forward.

I understand what you mean. This is a whole episode; we need an AI episode. How do you build your own agent? What tools are available and which do you choose today? Token optimization is a whole topic, because building agents quickly gets expensive. Several thousand kronor, sometimes a thousand a day. How...

I probably burned through 2,000 kronor just yesterday on something I wanted to do. You are saying 2,000 just disappears, bam. Is it cheaper to bring in freelancers to build the agents or assist? That is a whole episode; we will have to do it in a few episodes' time, partly when your agent is finished and ours is finished. We have several agents finished, but I have not taken a deep dive into how they were created. People on my team built them. I have assisted with strategy, but have not sat and built. We had a lunch and learn with someone on the team a few weeks ago, where he showed it step by step.

But I felt I would need to sit more with Ivan, side by side, and just get stuck in and see: help me out. Something I am looking at with the system is that anyone should be able to press "Install" and get the same process.

If we build an SEO structure, someone should just be able to press "Install" and gain access to the same thing without spending all the time. When I have spoken to others who build, someone might say: "My friend is an SEO specialist," and then they could talk to him. But not everyone has access to those people.

Typically, it is getting that critique. There are lots of free skills online, but free does not mean they are good.

Far from it. Optimizing them in many cases requires someone who understands what is right and wrong. If you can make templates out of it, we make things easier for everyone.

Love it. That makes sense. Let us pause that discussion, but we should resume it and create one or two dedicated AI episodes. Much more tutorial-based. You can share your screen when creating one. Instead of our lovely faces...

...you will see our screen. We literally do a process step by step. That sounds really good. Those of you who are interested, write to us on LinkedIn or comment on YouTube or Spotify: what do you want to learn about AI...

...and everything around it? We are both doing a lot, or you also have staff working on it. Comment on the post below. It helps us get better and give you better content.

Exactly, one hundred percent. That said, a lot has happened in the wider world. Two, actually three, fairly major things should be addressed. For one, the Riksbank left the repo rate unchanged a few weeks ago, contrary to expectations. First, they said inflation would soar because of the war.

The inflation figures come in at the end of April and are lower than expected. That means the Riksbank does not dare raise interest rates, even though it might initially have wanted to. The banks have already raised their rates as a precaution, which is really shitty.

But many major banks did that, and some followed along. Now they have to backtrack. From an outlook of higher rates this year, it now looks, on the contrary, as though there might be another cut. The most likely outcome is that it remains unchanged. That is a major macroeconomic change in society. Things move fast in hockey, as they say.

How does that affect e-commerce businesses and consumption behavior? I think it is really positive. If interest rates go up, people immediately get terrified. Mortgages are one of the biggest costs, if not the biggest, for individuals and families, so people hold tightly onto their money. The stock market has recovered and reached an all-time high again. That usually means those who are invested have more money in their pockets to spend.

We have interest rates being left unchanged. Both things bode well for the summer, which is otherwise usually slow for many e-commerce businesses. Now it will be very good, especially for seasonal products such as sunglasses, suitcases, beach balls, outdoor furniture and pergolas. Everything to do with summer...

...barbecues and outdoor activities, will do better than if the opposite had happened. The only downside, or there are a few, but the big one right now is that airfares have risen twelve percent just week on week. Fuel prices for airlines have soared, and flights are also being canceled because of it.

Exactly. When flights are canceled, prices rise on those remaining. Pressure increases. More people spend more money on travel and have less for consumption. That will also show in the retail figures for June, July and August.

Does that affect e-commerce itself or the tourist destinations they go to? I do not think tourist destinations are affected that much, because they still get visitors. We heard from Norwegian's CEO or marketing director that travel, or bookings, is unaffected by the price increase. The desire to travel is there, and that is no surprise: Swedes love traveling. We look forward to the winter sports break, autumn break and summer holidays with the children, and happily book something to look forward to and spend money on.

You might cut back on other things, and those other things are what I think get hit. What things? All consumption, most often what is not essential.

Twistboxes could be one example: "I will hold off on a roof box this year," or on outdoor furniture. That is highly speculative, but it is the macro climate now. The other big thing is that China blocked Meta's acquisition of the Manus app. Meta had announced the purchase and already integrated Manus into the platform. It is an AI tool that helps analyze the ad account and write things there. Now China's government is blocking the purchase. Meta never really got access to the algorithms and the smart brain behind Manus. They only had time to pre-integrate the app and release it to the public. Most people have probably seen...

...probably parts of it. Meta has come back and said: it does not matter, we are basing the AI tool Manus on our own LLMs. It is a big deal, a major tech acquisition being blocked by the Chinese government. What do you think? I find it interesting that they are continuing, but also that they put it in before the government approved it. I do not know, or was it approved and then the government came afterward and blocked it? Good question. I do not know. I do not think it had been approved, because larger AI transactions always have to be approved, including by the American government. If we look, for example, at when TikTok was going to be sold to...

...the US, the Chinese government wanted to block the transaction. The US has responded that they will not sell Nvidia chips to China-based companies, or outside the major US companies. First, the big companies get to reserve the chips, then smaller companies, and so on. That has meant Chinese companies have had to develop their own machinery, for inferior hardware to then develop software on.

It is coming from both sides, so it is not really unexpected. I think Meta's management knew and was aware that the risk was extremely high. They still chose to proceed with the transaction and roll it out, because they know AI is so important. That is their vision. Meta wants your budget, not for you to touch the ads or ad account. In the long run, you just say: this is my target audience, my goal and my budget. I want more purchases, leads, visitors or whatever the goal is.

You should be able to give Meta three parameters. Then you enter the website and where you want to drive traffic. It gives recommendations, perhaps that we change the product page, but AI is supposed to manage the ads. That has been their stated goal for a few months. Actually, several years, but more intensely over recent months.

Manus was just the first step. But they already have the LLM Llama, which they have developed for many years and invested many billions in. To me, it was strange that they bought an external tool from China to integrate.

Speaking of Facebook, our Facebook account was hacked recently. No! At Core by D, I mean. I usually never get fooled. When people get hacked, I often look at it and think it is fairly obvious. The problem was that we had run influencer marketing campaigns through an external IRM tool.

What is IRM? An influencer relationship management tool we had bought. What we did not think about when buying it... We enter how much of the budget we want to spend.

Then they find influencers who film, and afterward we approve the material. But we were not allowed to download the ad and put it into our own ad account. We had to connect their ad account as a manager to ours. Are you following?

As a partner, perhaps? Yes, as a partner. If we wanted to advertise, we had to give the money to them, and they would manage the ads.

Wow, incredibly dangerous. If you had asked me, I would never have allowed it. No, we did not do it. Then you do not have the data. No data, nothing.

What happened was that I said I was not going to do that. Good. But I still wanted to see whether we could track anything through them. I had approved the partnership and got an email from Facebook: do you accept this partner's permissions?

I double-checked: it was from Facebook itself, not someone else. But scammers had connected themselves to this company. When I clicked the link, they gained access. They created retargeting ads for...

...Chinese crap products, real scam products. A battery-powered air conditioner you could take with you. I know; I have seen that kind of company many times.

If you clicked the purchase link, you also ended up with viruses. Luckily, we discovered it, or rather Mia did. We monitor comments all the time so we can respond.

Good. There were bots commenting on the ad to create engagement. That is how we discovered it fairly quickly. It only managed to burn 5,000 kronor and things like that.

A bit of luck in the misfortune that she found it. The other thing is that Meta has now put in lots of hard blocks. When this happened a few years ago, in 2020 to 2021, you could easily burn a million before Meta put in a hard block, especially with invoicing or a credit card with a good limit.

Now, if you increase your ad spend from always spending 20,000 a day to suddenly 200,000, Meta will block it because they think something is fishy.

That is smart. Then, if you add a new pixel or connect a new site to drive traffic to, Meta blocks it. Not for a few thousand kronor, but as soon as it burns a little money. Meta refunds the money to advertisers, after all. The 5,000 you burned is not your fault; you were hacked. Report it, and in all likelihood you will get it back.

We got a message saying they expect a refund within 30 days. So it has already happened. It is Meta that loses the money. They do not want it to happen, because they understand you might stop advertising with them, even if it is your fault. The user experience also becomes bad. Terrible; they get junk ads to an audience. Meta has built many processes to prevent it. It is fantastic that they have put so many resources into it, because this is easy to fall victim to. I have seen many people...

...lose their accounts. But it was not your Facebook page? I thought so. The ad account was hijacked. Yes. Fuck. Through a partner account.

That can happen. That is why I am strict about two-factor authentication for all employees. We have access to many partners. All agencies are really a liability, a risk...

...because if you have an agency connected to the ad account, it can advertise on your behalf. All employees there who have access can also be hacked.

That is why you need several levels at the agency: two-factor authentication, not everyone having access to all ad accounts, and very few admins allowed to edit permissions. That alone helps protect against hacking.

I have been saved by also having two-factor authentication. The craziest thing I have experienced was when I had another company name and domain name within the business.

I had canceled the domain name but forgotten to disconnect it from Facebook, where I used it as my login. Suddenly, I got a notification: is that you trying to log in from this email address? Scammers had then, this was one or two years ago...

...bought the domain and tried to reset the password. They did not succeed because I had two-step verification. That actually happens often. But I thought it was crazy that they had managed to identify it as mine, bought the domain name and created an email address to reset my password.

That is next level. The most common thing is passwords leaking on the darknet in large threads that can be bought cheaply. You get ten million email addresses and passwords. Today, people use AI, but previously just scripts, to go through sites and reset people's passwords.

That was the most common and is surely still the most common. But buying a domain to get hold of a password is next level.

Exactly. Completely crazy. But I was saved by two-step verification. That is the tip: have it switched on, especially where you can spend money, like Meta, Google or Shopify with your orders and everything else. One extra click a few times. Fuck it. Install or activate it immediately. I have an acquaintance whose business email was hijacked. They took over the email and deleted everything in it. Then they emailed viruses to everyone on the contact list and translated it into Arabic, so it would take a little longer to get in and restore it.

Oh, shit. Then they deleted it, so when the person logged in, viruses had been sent out, but nothing was left. Great fun. Not.

You mentioned more important news. Yes, the economic changes, the acquisition blocked by the Chinese state and that damn war in Iran that seems to be dragging on longer than they first said. The American president said two to four weeks. I said it was over. Then it was extended to six to eight, and now we have passed eight weeks. The problems we have in the Strait of Hormuz mean energy prices continue to have the same issues. The market has already priced that in.

For e-commerce businesses, in all likelihood that means expensive deliveries from Asia to Sweden, and more expensive shipping within Europe too. We see it in petrol and diesel prices. We have received some subsidies from the government in Sweden, but not when importing from other parts of Europe, and especially not from the east. You need to take that into account and already review your inventory. Should you review pricing and adjust it? People are usually forgiving of price adjustments around transport and fuel surcharges. Do it as a precaution, before it is too late.

I have spoken with several entrepreneurs whose production prices have increased. Screws and parts are getting more expensive, just the other day. In total, perhaps the production cost has risen ten percent. That price usually never goes down afterward.

No, it is damn easy to raise prices, and very rarely do you lower them. Right. I was thinking about food VAT, which was a big thing on March 1 when the government announced a cut to six percent. Initially, all the grocery stores lowered prices across the board by six percent. I was even at my local Coop the day before and saw it. You can see it in the prices: when something cost 25.99, it now costs, what would it be, 22.78 or something like that. You see it. But when you now go into the grocery store to shop...

...you are back to tomatoes costing 49.90. Where did those six percent go? Where is the food VAT? They have priced things the same, at least some products. Not all, I think. Some are surely still reduced. I buy broccoli sprouts every week and see the same price as after the reduction. They probably made some reduction in the system, but products that are continually negotiated—tomatoes, cucumber, fruit and vegetables, meat and dairy—I think have already been adjusted back. Then it becomes difficult for the consumer to see any effect.

Is that because prices have risen so much for the stores too that they need to raise them? I do not think... Purchase prices must have increased because producers' deliveries have become more expensive. Then the food VAT cut gets swallowed up there, unfortunately.

That is a logical explanation, but I do not know; I am not that familiar with it. In all likelihood, it has a major effect on other Swedish consumption. What should those listening do in light of the macroeconomy and the summer you believe in?

First and best: review pricing. Even if you have plenty of inventory, look at the inflow and logistics. Has the price gone up? What do container deliveries and the forecast for the industry look like? Has anything been hit hard? If you work with gold, gold prices have soared over the past year. You constantly have to review whether prices to customers should be adjusted up or down, usually up, so the company is not hit too hard. Then you might need to email customers: we need to raise by X and Y because of this. Usually, customers are very forgiving. You think it will...

...hit us really hard, that no one will want to buy and everyone will go to competitors. That happens very rarely. But does that apply to existing customers? What about new ones? You cannot communicate with them in the same way. They land on the site without a reference point for the price, unless they have seen ads and hesitated to buy. They do not have the same reference point, so it is not a problem.

I am thinking about when you advertise and drive traffic to product pages. If new customers arrive and prices are continually raised, how does that affect conversion?

You should not be raising them all the time. Raise the prices. But it feels as though we have to do it all the time because of everything that has happened in recent years.

I think you raise them that one time and think more carefully in advance, so you know the prices can be maintained for at least six months, a year or two ahead.

So you do not have to raise them all the time. How much margin do you need? A little margin. You cannot raise two percent; you need to...

No, if you raise them, you have to raise them enough to cover the margin. Usually, a few percent is enough. That usually covers it even if fuel prices rise a hundred percent, because that is not the entire cost in GM1. You might only need to raise prices to customers by ten or five percent to cover fuel becoming a hundred percent more expensive, production more expensive and everything else going up. Right? That is usually not the problem.

Cool. We have talked about a lot, as usual a bit of everything. I hope you have picked up tips and tricks. If you have suggestions for other themes or episodes to address, send a DM to me or Fredrik. I promise we read them all. If you comment, we will definitely read and have a dialogue in the comments.

That brings this week's episode to a close. You will hear us as usual every Thursday morning. We will release episodes at least until summer, then we will see whether we take a summer break. Thank you very much. Share it with a colleague or friend if it was educational and you think the podcast should have an outrageously larger number of listeners. That also drives us to create better content.

Thank you so much. Thank you. Bye! Yeah.