(#61) The worst users come from referral programs, free trials, coupons, and gamification; OpenAI's next market; META's envy
Tesla has gained almost $9 billion from selling carbon credits since 2009
Knowledge Partner: EY
Download the latest edition of EY Attractiveness Survey Romania, an in-depth study designed to help businesses make investment decisions.
Findings are based on the views of representative panels of international and local decision-makers. Click HERE.
Strategy
7 trillion USD projects
Sam Altman, OpenAI’s boss, wants to raise 7 trillion USD to build the next generation of (AI) chips in the USA. Getting the money might be easier, but building and operating these fabs can take decades. Just as a benchmark, in 2023 the global CAPEX was $150 bn. Also, you need the (ASML) machinery to build which will take years to build and more years to install.
Lesson: Software is not hardware and software people don’t know how to build it. Sam Altman, ASLM
More on Vision Pro
We discussed last week about some early reviews from top tech gurus. Now, Mark Gurman - an Apple analyst at Bloomberg, has another review concluding the same:
1/ The product will be eventually an iPad killer
2/ Not a productivity tool…yet (but works OK when you are on a plane)
3/ We need more native apps to fully use it to the real value. LINK
Andrew Chen:
“The worst users come from referral programs, free trials, coupons, and gamification”
Why?
1/ Incentivized users vs. Organic users
Users attracted through incentive programs tend to have lower lifetime values (LTVs), worse conversion rates, and lower engagement levels compared to organic users. This trend is attributed to the fact that these users are often more discount-seeking and less qualified.
2/ Challenges with CAC/LTV analysis
Companies often fail to anticipate the negative selection bias of incentivized users, leading to an inaccurate cost of customer acquisition (CAC) to lifetime value (LTV) calculations. This misjudgment can result in attracting users who detract more value than they add, undermining the financial rationale for these incentive programs.
3/ Fraud and abuse
Incentive programs are susceptible to fraud, with users exploiting these systems for personal gain, further diluting the quality of users acquired through such programs.
4/ Cannibalization of the target market
Incentive programs can prematurely attract users who might have otherwise discovered the product organically, thereby unnecessarily increasing customer acquisition costs for users who would have converted without incentives.
5/ Negative impact on product innovation
Focusing excessively on complex incentive schemes can divert resources and attention from product innovation and improvement, potentially stalling broader growth and product refinement.
6/ Varied impact across different user types
While certain incentive programs, like those targeting drivers at Uber, can attract highly motivated and valuable users, others, particularly those aimed at consumers, can attract less desirable discount seekers. This variation underscores the importance of understanding the specific dynamics and motivations of different user segments when designing incentive programs.
7/ Broader implications for tech and gamification
The experience with incentive programs in the tech industry, including web3 and gamified consumer apps, demonstrates that adding game mechanics or incentives to products lacking inherent engagement and retention capabilities does not necessarily improve user acquisition or quality. Instead, these tactics can attract users who are not genuinely interested in the product itself, leading to poor long-term engagement and retention. LINK
Mark Zuckerberg on Quest 3 vs. Vision Pro
After spending $+40 bn on building a platform, Zuckerberg made a video showing that their device is better than Apple’s device and that it “costs 7 times less”. Probably, in the next product iteration Meta will copy some of Apple’s features while not changing significantly the price. LINK
No energy, no money
It is not imaginable to me what happens in Germany on the energy topic. The country prefers to burn coal than to operate the nuclear plants it has remained.
In the meantime, physics Nobel Laureate and former U.S. Energy Minister Steven Chu (under Obama) accuses German Greens of broadcasting "many half-truths" sabotaging German manufacturing and harming the climate by rejecting nuclear in favor of gas.
Who will pay the price for these bad in-chain decisions? The German people. Bloomberg, Die Welt
TikTok
“The TikTok representatives told the president that more than 160 million “fake accounts” had been identified and removed since October 7, including ones spreading anti-Jewish and anti-Israel rhetoric"
You have to see how TikTok covered (though algorithms) to the Israel/Gaza war to understand its real power. Just ban it! (I explained why in my previous newsletters). LINK
Tesla
“The Elon Musk-led manufacturer generated $1.79 billion in regulatory credit revenue last year, an annual filing showed last week. That brought the cumulative total Tesla has raked in since 2009 to almost $9 billion.” LINK
The EV market is having its mini-winter season which can last several years. LINK
Artificial Intelligence
AI tutors
We know that students with AI tutors perform better than those without them. China has purged several key industries and one of those was online education. Now, it seems that parents have adapted by turning to AI tablets for their kids. LINK
on ‘white collar’ jobs
Last year Goldman Sachs predicted that GenAI would affect 300 million white collar workers. I argued that this time is different (vs. the blue-collar industry/automation) because these white-collar workers have a platform and can influence decision-makers (ie. regulation). LINK
Now, it seems that is happening faster than expected:
“As the company grows, we’ll need fewer new hires as opposed to having to do a significant retrenchment,” said Chief Executive Mark E. Newman. LINK
OpenAI is on track to hit $2bn revenue.
“92 per cent of Fortune 500 companies were using OpenAI products, including ChatGPT and its underlying AI model GPT-4, as of November last year, while the chatbot has 100mn weekly users.” LINK
What’s their next market? Software that automates tasks. LINK
The future of education will be personalized with AI. LINK
Things happen
[Must read] How Aristotle created the computer. LINK
McKinsey places 3,000 staffers on review as economies slow. LINK
Roger Martin, a strategy OG, on the relevance of management models. LINK
After 6 years of internal debate, Chanel has a new definition of "fake purse". LINK
Ikea’s AI assistant gives design inspiration. LINK
Data
China is becoming the world’s shipyard. Yes, that means it’s more prone to a future war (over Taiwan). LINK
…in the meantime: China’s population is decreasing too fast. LINK
Median incomes in the West have increased. LINK
Outside interest
If you think World War III is unimaginable, read this. LINK
Alternative headline:
“Taiwanese immigrant builds company valued at more than whole Chinese stock market”
Havard protest
In 1961, thousands of students rioted at Harvard and had to be dispersed by police using tear gas. The cause? The students wanted their diplomas printed in Latin, not English. LINK
When Nissan made self-parking office chairs just for their own offices. LINK
Thank you for being an onStrategy reader!
Interesting newsletters 💡
Excellent curation & commentary, thanks Sorin!