Hey👋,
I'm Giacomo

Breaking news!‍

Apple hit by its first-ever EU fine over anti-competitive behaviour.

The €1.8bn fine was triggered by a formal complaint raised by Spotify against Apple Music in 2019.

The Background

Investigations into Apple's anti-competitive behaviour in the EU have been ongoing for years.

The App Store on iOS takes a 30% cut from all in-app payments and subscriptions for third-party apps, such as Spotify.

The issue is that Apple doesn't let third-party apps advertise alternative payment methods besides the Apple Store. Therefore, even if third-party apps wished to offer lower prices for users subscribing directly and not through the app store, they are unable to do so.

This creates a problem of transparency towards the end user and is seen as an abuse of dominant position.

The Trigger

As if that wasn't enough, Apple began launching its own apps, directly competing with some of the most popular ones on the store. For instance, Apple Music was launched to rival Spotify, and Apple TV to compete with Netflix.

It goes without saying that Apple Music doesn't pay the 30% cut, which allows them to potentially offer lower subscription prices.

Spotify was not the only one to be upset by Apple's practices. Meta was also among the “victims”.

You may recall the story of Apple's Tracking Transparency in 2022:

Apple provided its users with the option to opt out of tracking by third-party apps, which is essential for effective ad sales for companies like Meta.

Problem is, Apple's own apps don't provide the same option, and users can still be targeted by personalised advertising sold directly by Apple through its Apple Search Ads service. Once again, this was seen as anti-competitive behaviour.

The Future

‍Daniel Ek, CEO and founder of Spotify, published a video on Linkedin yesterday saying he’s happy about the EU’s decision but he’s still skeptical about whether the fine will actually change Apple’s behaviour.

He believes that Apple will just ignore the ruling. Maybe they will apply some small changes here and there, but will not fundamentally change the way they operate. He mentions how similar fines Apple received in other countries like South Korea or The Netherlands also didn’t achieve their ultimate objective.

The Underlying Problem

There is a fundamental problem over the openness of the internet.

Today’s internet is practically owned and operated by a handful of American tech behemoths. Ai doesn’t seem to be solving this problem. In fact, it looks quite the opposite.

From this perspective, the recent lawsuit raised by Elon Musk against OpenAI seems reasonable.

Last week, Universal Music Publishing Group released a statement on their Instagram account.

The background

Universal Music Group and TikTok clashed over royalty payments for songs affiliated with the music label and played on the social media platform.

Negotiations failed because TikTok reportedly wanted to pay only a low-single digit percentage of advertising revenue, in contrast to the 20% that YouTube, for example, pays to Universal.

Therefore, Universal withdrew their license, forcing TikTok to stop playing its music catalogue, which includes artists like Taylor Swift and Drake.

The Impact

Over the past few weeks, TikTok has muted all recordings owned by Universal. But it didn’t stop there. It also muted songs written or co-written by artists affiliated with the major.

Some analysts predict that 60-80% of all the popular music on TikTok will be muted, although the Chinese giant claims the figure is closer to 30%. Still, it’s a huge deal, impacting at least one-third of all songs played on TikTok.

The Future

Reality is, TikTok can live without mainstream hit songs.

Why?

Because regular users can now create catchy tunes using AI. In fact, TikTok plans to pay an advertising cut to AI-music creators as well.

AI-generated music is perfectly suited for accompanying TikTok videos. In fact, it may even perform better as it can be tailored specifically for this purpose. In near the future, we may have tools to create music directly from existing videos.

Like every industry, the music industry is also being disrupted by AI. The only way to survive is by improving the quality of its output. The formula for mass-produced summer hits won’t work anymore, at least from a streaming point of view.

Instead, TikTok should be used as a marketing tool by labels and artists to generate revenue elsewhere, such as live shows. Taylor Swift demonstrated this model with her Eras Tour.

As a marketing professional myself I know that marketing is a cost center, not a profit one, at least not directly. Therefore, I don’t see how music companies can expect significant revenue from a marketing too like TikTok.

‍

Americans do it bigger!

If you've ever skied in Andermatt, Switzerland, you might have not realised that you were a guest of the Americans. The same would be true if you recently skied in Crans-Montana.

Indeed, both ski resorts are operated by the American giant Vail Resorts.

Forget about local flavour and traditional food.‍ Vail Resorts owns and operates more than 40 resorts across the US, Canada, Australia, and Switzerland. Their acquisitions of the Andermatt and Crans-Montana ski resorts in March 2022 and November 2023 respectively, marked their entry into the European market.

In Andermatt, they quickly injected millions to upgrade lifts, improve snow-making, enhance hospitality and more.

But more importantly, they included the Swiss resort into their signature international format, the Epic Pass.

With the Epic Pass, skiers can choose to ski at any of the 40+ resorts owned by the company. This allows them to literally follow the snow across continents!

Given that day-passes at American resorts are priced nearly $300, a season Epic Pass costing around $900 looks like a bargain. For a New Yorker, flying 6-7 hours to Zurich or 4 to Colorado plus the driving doesn’t make a big difference.

With the number of overnight stays by North Americans quadrupling in the last year, it’s now common to hear Yankee accents on the slopes of Andermatt.

The entry of the Americans has driven up real estate prices, by 12% in the last year, reaching CHF 18-19k per square meter. International companies, such as the Egyptian-Montenegrin-Swiss Orascom Development, are undertaking major real estate projects in the area. Crans-Montana will probably follow suit.

The model of large public companies operating multiple ski resorts is new to Europe.

In fact, most European resorts are managed by local entities. For example, the Dolomiti Superski area in Italy and the Arlberg area in Austria, among the largest of the Alps, are managed by local companies bearing the same name.

An exception is the French Compagnie des Alpes, a public company listed in Paris which operates some of largest ski resorts in France.

Due to climate change, running ski resorts is becoming an expensive endeavour. It requires increasingly larger capital, something Americans don’t lack of.

However, this might come at the expense of the local alpine favour, which I believe is essential for an enjoyable ski holiday in Europe.

History repeats.

NVIDIA is leading the Ai boom of the 2020s, while Cisco was the champion of the internet infrastructure race in the late 1990s.

The fate of Cisco and its peers is well known. Will it happen the same to the Ai industry?‍

The Economist has compared the proliferation of Ai apps to the invention of tractors, while the Financial Times likens it to the early days of the internet.

Both tractors and the internet took a long time to replace their predecessors and establish themselves as labor and economic revolutions.

The Dot Com bubble didn't halt the internet; quite the contrary. However, it took at least another decade for today's tech giants to emerge, many of which didn't even exist in 2000.

Same could happen to the Ai industry and its champions.

Today, chips are king. But they quickly become commodities.

Modern chips, currently in high demand, will become cheaper and in large supply in maybe a couple of years, potentially threatening Nvidia's revenue.

The shortages experienced in 2021 and 2022 were promptly fixed, resulting in overcapacity.

Among others, Samsung Semiconductor had to decrease production last year due to a growing chip surplus, while its Japanese counterpart, KIOXIA Group, reported a record loss of $1.7 bn. Meanwhile, global silicon wafer shipments fell 14.3% last year.

Oversupply, fuelled by too much cash too soon, was also at hearth of the Telecom crush in 2000-2002.

High valuations and hype don’t always (almost never) reflect the reality of an industry!

Facebook turns 20 this year, but it’s a walking dead.

Or, is it?…

Facebook remains the most popular social media app worldwide, with over 3bn active users as of January 2024 (up ~2% YoY), ahead of YouTube and Instagram. Not bad for a walking dead. 🤔

However, the social network is dead.

I’ve written before about the switch between social networking and social media, but now something even deeper is happening.

We used to network with our friends and acquittances, then with strangers, eventually with fans. Then we stopped networking all together, spending time on social media mostly for entertainment, consuming content (mostly video) made by professional companies or individual creators, much like TV.

Now, we’re back to networking, but not with humans. We’re “networking” with an Ai algorithm.

When we open Facebook or TikTok, we’re talking to an Ai that chooses content for us. It is an Ai algorithm that really chooses how long we should spend on the app, what ads we should watch and what creators we should “network” with.

However, the need for online social networking remains. According to a Morning Consult study, only 30% of people interviewed said they would share a movie recommendation on social media, behind the 43% who chose text or email and the 35% who opted for group chats. Same happens for news and politics.

This explains the increasing popularity of chat apps. WhatsApp is the third most popular social media app worldwide, Facebook Messenger and Telegram the 7th and 8th respectively, ahead of Snap Inc. and X.

This has massive implication for digital marketing.

On the one hand, marketers need to realise that they’re not marketing to people anymore, but rather to Ai algorithms, who in turn will talk to real people. Even though some platforms claim they prioritise quality content vs Ai-optimised one (Google search, for example), currently it doesn’t seem to be happening much.

On the other hand, marketers need to find other ways to talk to real people. WhatsApp and Instagram channels are good candidates. You won’t get rid of Meta platforms easily!

Many have declared the premature death of Facebook before, even as early as 2009 when The New York Times titled "Facebook Exodus".

Reality is, long live The Facebook!

Social media are more popular than ever and advertising revenue is flowing in. 🤑

Meta Ads revenue grew by 16% in 2023 YoY, representing a whopping 98% of total revenue! Net income (after tax) increased by 69% in 2023, at a margin of 29%. 😱

Facebook couldn't have entered its 20s in a better way!

6 Ai tools I use on a daily basis:

1. ChatGPT-4, premium version ($20/month)
   This is a must. Virtually any Ai task can be performed with ChatGPT, especially since the launch of custom GPTs.
2. Custom GPTs (part of ChatGPT premium).
   I created two custom GPTs that I use on a daily basis:
   >> Improve Text (public)
   >> Thumbnail Generator (private, for now)
   With Improve Text I simply paste a paragraph in the chat and without any prompting it automatically proofreads and improves my text. It generally uses a better grammar and vocabulary. I'm still blown away by the quality of its writing. Nowadays, anything I write goes through that first.
   Thumbnail Generator creates two thumbnails for my blog posts. I gave it specific instructions to create images always in a certain size and style. If you check my blog you'll realise that all my thumbnails look similar, as if they were made by the same artist.
   It doesn't require complex prompting. A few words are usually enough to obtain a great result. For example, "Netflix wins the streaming war" was the prompt for my latest blog image
3. ‍Perplexity Ai (free)
   Lately, I find myself searching on Perplexity Ai more than on Google. I use it especially for researching the topics I write about on Linkedin and on my blog.
4. ‍Stable Diffusion (free)
   I use its image-to-video generator. It's free, it's quick and generally good quality. Usually, I create images with the Thumbnail Generator, then use Stable Diffusion to create a short animation out of it. Then, I create a GIF that I sometimes post here on Linkedin.
5. ‍HeyGen ($29/month + add-ons)
   I used it to create my own avatar for some work-related videos. I am now exploring also library avatars for more impersonal videos. Excellent quality, easy to use.
6. ‍Midjourney ($10/month)
   This is by far the best image-generation tool (in my opinion), but since I've created my custom GPT, I stopped using it to save money 😄.
   However, we use it at SMG Swiss Marketplace Group for marketing work.
   
   That's it! Ai doesn't need to be complicated.

When will Ai replace human labour? Probably later than you think.‍

The Economist compares the introduction of ChatGPT and other Ai applications to the invention of the tractor at the end of the 19th century.

The tractor revolutionised agriculture, allowing farmers to be more productive and efficient. Such improvements contributed to raise American GDP by 8% by the mid-50s.

Yet, it was not an explosive revolution, but rather a slow incremental change.

The chart shows how the tractor took over 50 years to establish itself as a better substitute for horses and mules.

Why?

Essentially, for several decades human and animal labour remained cheaper and more reliable than the tractor.

Early tractors were extremely expensive and still not capable of most farming tasks. On the other hand, labour was cheap and plenty available.

Plus, farms were still not organised to accomodate the tractor in an efficient way. For example, early tractors were most suited to some specific terrains and crops. For all the rest humans and animals were better.

Something very similar is happening with Ai.

Although we’ve already been blown away by the capabilities of GPT-4, its real-life applications are still to be proved.

Real wages have hardly risen in the past years, so hiring humans instead of machines will still be more efficient for another while.

It might take decades for Ai to establish itself as a labour revolution. But rest assured, it will happen! Same as with the tractor.

In a conversation with The Economist in Davos, Sam Altman stated that “when we’ll reach AGI (artificial general intelligence), people will freak out for a couple of weeks and then they’ll go back to their normal lives”.

In the end, in over one year since the launch of ChatGPT, our lives haven’t changed much.

Probably, there won’t be any “boom” or explosion moment, but rather a gradual yet irreversible change over decades.

‍

‍

Have you seen the Ferrari movie?

It was released at a perfect timing, right at the close of a record year for the sports car company.

The movie is excellent and provides an accurate portrayal of Emilia, the Italian region where Ferrari is based, and its people. There's even a scene in a restaurant where Enzo Ferrari is served the boiled meat tray, a peculiar speciality of the Emilia region. As an "Emilian" myself, I can confirm its accuracy!

Unfortunately, the movie mainly focuses on racing and a specific episode in Enzo's life, without addressing Ferrari as a luxury car manufacturer.

2023 was a record year at Ferrari. Some facts:

• Car sales were up 3% YoY to 13’663 units.
• Revenue were up 17% to almost 6bn. Ferrari sold higher-priced cars and more personalised features.
• Sales were mostly driven by the new SUV-like Purosangue.
• The average price of a Ferrari car in 2023 was €397’000, a new record!
• Net profit climbed 34% surpassing 1bn for the fist time in its history.
• 10% of revenue came from sponsorship and merchandise. store.ferrari.com sells Ferrari-branded clothing and accessories, from $200 baseball hats to $5000 leather suits. Ferrari is a luxury brand, not just a car maker.
• It is the first luxury group to have eliminated the pay gap between men and women at a global level.
• Lewis Hamilton, one of the most successful F1 drivers in history, will join Ferrari in 2025.
  
  Not bad for a company out of the small town of Maranello, Italy 🇮🇹

Is Netflix the winner of the streaming war?

In recent years, traditional studios like Disney with Display+ and Warner Bros. Discovery with HBO Max have entered the streaming industry, significantly intensifying the competition in the space.

The streaming business requires huge investments to produce high-quality original content, which is key to "stealing" users from competitors. But with monthly subscriptions starting from as little as $5, streamers need an exceptionally large user base to turn a profit.

Profit hasn’t happened at Disney yet, whose streaming unit lost $420 million in FY23 (down from a loss of 1.4bn the year before). Warner Bros. Discovery does better, finally scoring a profit of $111 million in 2023, up from a loss of $634 million in 2022.

But Netflix does significantly better!

In 2023, it increased users by 13% to 260 million, against a worrying drop the year before. It’s been consistently profitable for over 15 years, delivering almost 5.5bn net profit (after tax) in 2023 at a net margin of 16%.

On top of successfully cracking down on password sharing and launching an ad-supported subscription tier, Netflix has a secret weapon: content licensing.‍

Competitors, under pressure to deliver profit, are (re)starting to license their original content to third-parties, including and especially to Netflix. While this generates easy money for the likes of Disney, it also enriches the Netflix’s catalogue, for the delight of its users.

Netflix is so good at servicing its subscribers that when NBCUniversal licensed “Suits” in June 2023, the show went straight to the number-one most watched, despite being relatively old (originally aired on cable TV in 2011). In fact, according to Nielsen, in November 2023, nine of the ten most streamed programs were licensed content.

With around $17bn spent on content in 2023, Netflix might reduce this expense while still expanding its catalogue with premium, although old, content.

Beware, this is not reciprocal! According to Netflix’s executives, there’s more value in acquiring licensed content, than to sell it. So don’t expect “Ozark” to stream anywhere else.

Also, ad revenue is increasing and projected to represent up to 22% of total revenue in 2027. With this, Netflix won’t have to fight for new subscribers, but rather will extract more value from existing users.

If we add to the mix the recent $5bn deal to livestream WWE, we get the perfect recipe to declare Netflix as the winner of the streaming war.

However, streaming remains a costly and hit-driven business. When compared to YouTube, which spends virtually nothing on content, streaming looks way less appealing.

Stay tuned!

‍

‍

‍

‍

‍

‍

‍

The search war is on!

A recent study by Adobe reveals that 2 in 5 Americans now use TikTok not only for entertainment but also as a search engine. While Gen Z leads this trend, Millennials and Gen X are also on board.

‍

‍

This marks a historical shift in Google's dominance in the search engine market.

In fact, TikTok isn't the only emerging competitor. The study indicates that over 1 in 10 consumers used ChatGPT to search for information they would typically find on Google.

Ai chatbots, especially those capable of web searching like Perplexity or Microsoft Copilot, are becoming extraordinary competitors to traditional search engines.

Even Bing itself is likely to increase its relevance since ChatGPT often refers to it.

‍

‍

Although Google remains the clear leader, its market share is visibly declining in Europe, particularly on Desktop. At the same time, Bing's market share on Desktop rose from 8.5% in January 2022 to 11% in December 2023.

‍

‍

Keep in mind that in the huge search engine market, every percentage point is a big deal!

This shift represents an incredible challenge, but also an opportunity for marketers quick enough to adapt to the new changes. SEM will inevitably evolve, but how exactly it’s still too early to tell.