Wolt går efter dine kunder, men nu går AI efter Wolt

I denne uges nyhedsbrev skriver jeg om:

At the bottom of the newsletter, you can read an English-language version.

Kære læser

For fem år siden havde jeg en sjov oplevelse.

Familiens dovne søn lå i sin seng en lørdag formiddag og bestilte en burger på sin iphone. En halv time senere kom et Wolt-bud med burgeren.

Jeg var rystet, og jeg har fortalt anekdoten flere gange, men oplevelsen lærte mig noget vigtigt.

Jeg lærte, at tidens vilkår i erhvervslivet er at digitale platforme tager magten.

De tager en position mellem en producent (restauranten) og kunden (min søn), og skubber på den måde producenten ned i en sårbar position som en underleverandør, der skal konkurrere på lav pris frem for på høj kvalitet.

Det er blot fem år siden. En evighed.

Fordi: Nu har Wolts ejer, amerikanske DoorDash, købt Instacart for 17 mia. dollar.

Hvad er Instacart?

Instacart er en platform, hvor forbrugere bestiller dagligvarer fra supermarkeder, og den leverer også teknologi til supermarkeder.

Med opkøbet kan Wolts ejer både udvide sin dagligvareforretning og styrke sin teknologi.

Men jeg har en hypotese: Opkøbet er et svar på den trussel, som AI-agenter udgør mod de digitale platforme.

Wolt har opbygget sin magt ved at være det sted, hvor kunderne finder og bestiller deres varer.

Og hvad sker der, hvis kunderne bruger en AI-agent til det?

Forestil dig, kære læser, at min søn ikke længere åbner Wolt, når han vil have en burger.

I stedet beder han sin AI-agent om at finde en god burger, der kan leveres hurtigt og billigt.

Agenten tjekker restauranterne og gennemfører bestillingen. Min søn er ligeglad med, hvem der kommer med burgeren.

Derfor er Instacart-købet interessant.

Mit bud er, at DoorDash med købet af Instacart forsøger at gøre sig uundværlig, selv om AI-agenter overtager relationen til kunden. DoorDash har stadig teknologien og cykelbudene og gør sig dermed svære at undvære.

Da jeg mødte DoorDash-stifteren Tony Xu i København i 2023, sagde han, at hans ambition er, at Wolt skal kunne levere stort set alle slags varer fra lokale butikker.

Nu får den ambition en ny betydning. En AI-agent kan godt finde og bestille en burger, men den kan hverken tilberede eller levere burgeren. Derfor kan Wolts netværk blive en styrke i konkurrencen med AI.

Der er imidlertid en afgørende forskel på at levere burgeren og at have relationen til kunden.

Hvis en AI-agent overtager kunderelationen, kan den vælge mellem forskellige budtjenester, og så risikerer Wolt at ende som en underleverandør - ligesom restauranterne frygter at blive det over for Wolt.

Så hurtigt kan det gå:

Restauranten laver stadig burgeren, og Wolt kan stadig levere den, men en AI-agent bestemmer, hvem der får ordren.

PS: Da jeg talte med Mærsks topchef, Vincent Clerc om det med platformene i 2023, sagde han, at kunderelationen er mindst lige så vigtig som hans skibe og terminaler. Selv kæmpestore Mærsk ser dette som en trussel mod deres kunderelation.

Derfor nørder jeg så meget ned i det her kaninhul:

Jeg ser budtjenesterne som et laboratorium for resten af erhvervslivet.

Ugens billede: Fra den litterære verden

Et smukt billede, ikke sandt? Min nye bog ”Tænk som en topchef” har fået en fornem placering på det lokale bibliotek.

Men ak, se så, hvad der står på ryggen af bogen. En voldsom ydmygelse af den store forfatter. Den er hård.

Ugens træf i tech-miljøet: Tre AI-læringer

Jeg var til et sjovt møde i aftes. Tech-miljøet holdt et træf på Københavns Universitet.

Jeg ser på tech-miljøet lidt ligesom jeg ser på budtjenesterne. Det, der sker i tech-miljøet, det kommer til resten af erhvervslivet lige om lidt. Et kig ind i fremtiden.

Nå, men før at de gik i gang med at drikke fadøl, så var der en paneldebat.

I panelet sad Rolf Pedersen, True Collective, tidligere GoLearn, Ken Villum Klausen, tidligere Lunar, nu Repodo og Rasmus Møller-Nielsen, LAB08.

De tre kom langt omkring, især om AI, og jeg lyttede med dette perspektiv:

Når mere og mere arbejde udføres af ikke-mennesker, hvad sker der så med vores syn på en virksomhed?

Et skævt spørgsmål, jeg ved det, men sådan er det her i nyhedsbrevet. Vi tænker skævt.

Torsdag aften i Lundbeck-auditoriet på Københavns universitet. Fra venstre: Rolf Pedersen, True Collective, tidligere GoLearn, Ken Villum Klausen, tidligere Lunar, nu Repodo og Rasmus Møller-Nielsen, LAB08.

Jeg noterede tre budskaber.

1. AI er mere end et værktøj
Vi er vant til at tænke på AI som noget, medarbejderne bruger til at løse opgaver hurtigere, men her hørte jeg om noget andet.

Om at gøre AI-agenter synlige i virksomhedens organisationsdiagram.

Om en fælles digital hukommelse - en company brain - hvor viden fra møder og processer kan bruges af medarbejdere og agenter.

Teknologien kan på den måde blive en integreret del af virksomhedens hukommelse.

Min læring: AI ændrer ikke bare vores arbejde, men udfordrer vores forestilling om, hvad en organisation er.

2. Behovet for mennesker ændrer sig
AI øger produktiviteten nærmest absurd meget, men en virksomhed har stadig brug for medarbejdere.

Den har brug for mennesker, der forstår kunderne og kan sælge løsningerne.

Læg dertil, at flaskehalsene flytter sig. Når det bliver lettere at udvikle et produkt, bliver udfordringen i stedet at finde kunderne og at få en løsning til at fungere hos kunderne.

Det er blevet nemt at udvikle software, men det er stadig svært at vinde i et marked.

Min læring: AI øger ikke kun produktiviteten, men ændrer også grundlaget for, hvordan en virksomhed konkurrerer.

3. AI kræver et nyt syn på ledelse
Alle virksomheder bruger AI, men det er langt fra alle, der har gentænkt deres organisation.

Investorer spørger ikke kun tech-iværksætterne om, hvad de udvikler med AI, men også om, hvordan de selv bruger AI i deres arbejde. Det siger noget om, at evnen til at organisere arbejdet med AI også er ved at blive en målestok for god ledelse.

Det er ikke nok at bruge AI til at sætte strøm til en organisation, du bør også overveje, hvordan beslutninger skal træffes, og hvordan mennesker og AI-agenter arbejder sammen.

Min læring: Den største udfordring for virksomheder er ikke at bruge AI, men at ændre den måde, virksomheden er organiseret på.

Lars Mørch har aflivet havkatten i Jyske Bank

Vi har taget hul på de nye regnskaber for tredje kvartal – i denne uge kom blandt andre Carlsberg, Danske Bank, ISS og Jyske Bank – men de var ret forudsigelige.

Så jeg kiggede efter noget andet, og jeg lagde mærke til noget sjovt hos Jyske Bank.

Havkatten er væk.

Den aggressive fisk, som var Jyske Banks symbol i mere end 40 år, kan stadig ses nogle steder internt i banken, men set udefra er havkatten død. Aflivet af topchefen Lars Mørch.

Lars Mørch er en pæn mand og ikke en havkat. Og fremtidens Jyske Bank skal ikke forbindes med en havkat.

Det er slut med at være en provokatør, nu skal Jyske Bank være forudsigelig. Mere almindelig. Måske endda kedelig.

Havkatten var bankens identitet. Et synligt alternativ til de traditionelle banker i en branche, hvor produkterne er nærmest ens.

Lars Mørchs forgængere, Anders Dam, Kaj Steenkjær og Poul Norup, var helt anderledes aggressive profiler end kollegerne i bankerne i København. Frækhed som en slags jysk konkurrencefordel.

Så: Kan Lars Mørch gøre det modsatte? Kan han gøre forudsigelighed til noget, Jyske Bank kan konkurrere på?

Jeg forsøger at kloge mig på forudsigelighed som princip i den nye Jyske Bank her.

Ugens citat: »Vi kneppede dem igen«

Niels Frederiksen, cheftræner i den polske fodboldklub Lech Poznan, gav den gas i omklædningsrummet efter sejren over Legia Warszawa.

Klippet er gået viralt. Ikke underligt.

Selv en pæn mand som Niels Frederiksen med en fortid som leder i Danske Bank, kan slippe håndtaget. Dejligt at se. Du skal fejre dine succeser.

Du kan se klippet her.

Disclaimer: Vi rødmer her i vores noble nyhedsbrev. Vi kunne aldrig finde på at råbe sådan noget.

God weekend!

This is an AI-generated translation of the newsletter. It has been proofread by me for errors after the translation.

In this week’s newsletter:

Dear reader,

Five years ago, I had a funny experience.

The family’s lazy son was lying in bed on a Saturday morning, ordering a burger on his iPhone. Half an hour later, a Wolt courier showed up with the burger.

I was stunned. I’ve told that story several times since, but the experience taught me something important.

It taught me that one of the defining realities of business today is this: digital platforms are taking power.

They insert themselves between the producer (the restaurant) and the customer (my son), and in doing so, they push the producer into a vulnerable role as a subcontractor — forced to compete on price rather than quality.

That was only five years ago. It feels like an eternity.

Because now Wolt’s owner, the American company DoorDash, has bought Instacart for $17 billion.

So what is Instacart?

Instacart is a platform that lets consumers order groceries from supermarkets, and it also provides technology to supermarkets.

With the acquisition, Wolt’s owner can both expand its grocery business and strengthen its tech platform.

But I have a hypothesis: this deal is also a response to the threat AI agents pose to digital platforms.

Wolt built its power by becoming the place where customers find and order what they want.

So what happens if customers start using an AI agent for that instead?

Imagine, dear reader, that my son no longer opens Wolt when he wants a burger.

Instead, he asks his AI agent to find him a good burger that can be delivered quickly and cheaply.

The agent checks the restaurants and places the order. My son doesn’t care who delivers the burger.

That is why the Instacart deal is interesting.

My view is that by acquiring Instacart, DoorDash is trying to make itself indispensable — even if AI agents take over the customer relationship. DoorDash still has the technology, the courier network, and the operational infrastructure, which makes it harder to bypass.

When I met DoorDash founder Tony Xu in Copenhagen in 2023, he told me his ambition was for Wolt to be able to deliver virtually any kind of product from local stores.

Now that ambition takes on a new meaning. An AI agent may be able to find and order a burger, but it can’t cook it or deliver it. That is where Wolt’s network could become a real strength in the competition with AI.

There is, however, one crucial difference between delivering the burger and owning the customer relationship.

If an AI agent takes over the customer relationship, it can choose between different delivery services. And in that scenario, Wolt risks ending up as a subcontractor — just as restaurants fear becoming one in relation to Wolt.

That’s how quickly things can change:

The restaurant still makes the burger, and Wolt may still deliver it, but an AI agent decides who gets the order.

P.S. When I spoke with Maersk CEO Vincent Clerc about platforms back in 2023, he told me that the customer relationship is at least as important as his ships and terminals. Even a giant like Maersk sees this as a real threat to its relationship with customers.

That’s why I keep going down this rabbit hole:

I see delivery platforms as a laboratory for the rest of the business world.

Image of the week: From the literary world

A beautiful image, isn’t it? My new book, Think Like a CEO, has been given a very prominent spot at the local library.

But alas — look at what’s written on the spine. A brutal humiliation of a great author. It’s tough.

This week’s gathering in the tech scene: Three lessons on AI

I went to a fun event last night. The tech community had organized a gathering at the University of Copenhagen.

I look at the tech scene much the same way I look at delivery platforms. What happens in tech tends to reach the rest of the business world soon after. It’s a glimpse into the future.

Anyway, before people got started on the draft beers, there was a panel discussion.

The panel featured Rolf Pedersen of True Collective, formerly GoLearn; Ken Villum Klausen, formerly of Lunar and now at Repodo; and Rasmus Møller-Nielsen of LAB08.

The three covered a lot of ground, especially on AI, and I listened with one question in mind:

As more and more work is carried out by non-humans, what happens to our idea of what a company actually is?

A slightly odd question, I know. But that’s how this newsletter works. We like to think sideways.

I came away with three takeaways.

1. AI is more than a tool

We are used to thinking of AI as something employees use to get work done faster. But what I heard here was something different.

I heard about making AI agents visible in the company org chart.

About building a shared digital memory — a company brain — where knowledge from meetings and processes can be used by both employees and agents.

In that sense, the technology can become an integrated part of the company’s memory.

My takeaway: AI is not just changing how we work. It is challenging our very idea of what an organization is.

2. The role of people is changing

AI is driving productivity to almost absurd levels. But companies still need people.

They need people who understand customers and know how to sell solutions.

And on top of that, the bottlenecks are shifting. When building a product becomes easier, the real challenge becomes finding customers and getting the solution to work for them.

Software has become easier to build. It is still hard to win in the market.

My takeaway: AI is not just increasing productivity. It is also changing the basis on which companies compete.

3. AI demands a new view of leadership

Every company is using AI. But far fewer have rethought their organization around it.

Investors are no longer just asking founders what they are building with AI. They are also asking how those founders themselves are using AI in their day-to-day work. That says something important: the ability to organize work around AI is itself becoming a measure of good leadership.

It is not enough to use AI to turbocharge an organization. You also need to rethink how decisions are made and how humans and AI agents work together.

My takeaway: The biggest challenge companies face is not adopting AI. It is changing the way the organization itself is structured.

Lars Mørch has killed off the catfish at Jyske Bank

We are now into the third-quarter earnings season. This week brought results from the likes of Carlsberg, Danske Bank, ISS, and Jyske Bank — but most of it was fairly predictable.

So I looked for something else, and I noticed something interesting at Jyske Bank.

The catfish is gone.

The aggressive fish that served as Jyske Bank’s symbol for more than 40 years can still be found in places inside the bank. But viewed from the outside, the catfish is dead. Killed off by CEO Lars Mørch.

Lars Mørch is a proper gentleman, not a catfish. And the future Jyske Bank is not meant to be associated with one either.

The days of playing the provocateur are over. Now Jyske Bank is supposed to be predictable. More conventional. Maybe even boring.

The catfish was the bank’s identity. A visible alternative to traditional banks in an industry where the products are more or less identical.

Lars Mørch’s predecessors — Anders Dam, Kaj Steenkjær, and Poul Norup — were far more aggressive public figures than their counterparts in the Copenhagen banks. Boldness as a kind of Jutland competitive advantage.

So here’s the question: can Lars Mørch do the opposite? Can he turn predictability into something Jyske Bank can actually compete on?

I try to think through predictability as a strategic principle in the new Jyske Bank here.

Quote of the week: »We fucked them again«

Niels Frederiksen, head coach of Polish football club Lech Poznań, really let loose in the locker room after the win over Legia Warszawa.

The clip has gone viral. Not surprisingly.

Even a proper gentleman like Niels Frederiksen — with a past life as a manager at Danske Bank — can lose himself in the moment. Wonderful to see. You should celebrate your successes.

You can watch the clip here.

Disclaimer: We blush a little here in our noble newsletter. We would never dream of shouting something like that.

Have a great weekend!