el producto #422 🚀
Career decisions, Updated Alexa, Apple's event planning app, New OpenAI and Gemini models, How to choose the right idea, AI defensibility, Big Tech Q4 results & more
Hi folks 👋
Welcome to a new edition of el producto
🎰 The week in figures
$1B: Okta competitor SailPoint’s return to the public markets is set to raise $1B at a valuation of around $11.5B
$300M: Deel, the global payroll platform, sold $300M in secondary shares at its $12B valuation as it gears up for a 2026 IPO
$300M: Payments startup Rapyd is raising $300M, but at a $3.5B valuation—down 60% from its 2021 peak of $9B. Like Stripe and Klarna, Rapyd is facing tougher investor expectations and the end of easy fintech funding. The company plans to acquire a payment processor, betting that more scale will help it survive the slowdown
$200M: The Barcelona business travel platform raised a $200M Series E at $2.7B, at ~2x its previous $1.4B valuation. It also acquired expense management platform Yokoy and plans to further expand to the US
$104M: Alice & Bob, a Paris-based quantum computing startup with an awesome name, raised a $104M Series B to build a fault-tolerant quantum computer with "cat qubits", targeting useful computing applications by 2030
+36%: Video traffic on LinkedIn is booming, increasing 36% YoY and growing at twice the rate of other platforms, according to its Director of Video Product
13%: Retention rates for mobile apps are falling. In 2024, overall day 1 retention rates fell from 14% to 13%. Day 30 retention remained at 3%. Ecommerce marketplace apps had the highest day 1 retention rate of 25%
💸 Q4 results
Alphabet: missed Q4 revenue expectations and announced a massive $75B capital expenditure plan for AI infrastructure. Revenue grew 12% YoY but saw a slowdown in YouTube ads, search, and services
Amazon: Q4 net sales increased 10% to $187.79B vs. $187.30B expected; AWS: $28.8B meeting expectations; advertising: $17.3B vs. $17.4B expected
Spotify: set a new Q4 record and its first full year of profitability. Subscribers grew to 263M, +11% YoY
📰 What’s going on
Amazon is revamping Alexa into a Claude-powered generative AI service to hold complex, context-aware conversations and handle multifaceted requests on its 500M+ Alexa-enabled devices. Amazon is holding an event in New York, on February 26th, where it's rumored it will be launching the long-awaited, AI-powered Alexa
Apple launched the "Invites" iCloud+ event planning app, competing with Partiful. It is also developing "Confetti," an iOS 18.3 Calendar feature with multi-calendar support and photo sharing capabilities
Google Photos, is getting a handy new UX improvement. The ability to “hide clutter from other apps” will remove all non-Photo materials like screenshots, GIFs and attachments from other apps like WhatsApp
Google is making its reasoning model—Gemini 2.0 Flash Thinking—available to all users on the Gemini app, confirming the impending release of Gemini 2.0 Pro Experimental (it was accidentally revealed after it was spotted in Google’s changelogs last week), and introducing Gemini 2.0 Flash-Lite, a cheaper, more efficient model (possibly a challenge to DeepSeek’s R1?). Gemini 2.0 Flash Thinking (launched in December) works through problems, showing its ‘thought’ process, and integrates with YouTube, Google Maps, and Google Search instead of relying solely on its own data
After releasing its first, smaller reasoning model—o3-mini—last week, OpenAI has announced it has just changed how the model responds to queries, meaning it will now show users its “thought” process, before giving its answer. This comes (suspiciously?) soon after Chinese rival, DeepSeek, released its own reasoning model, R1, which shows a similar “chain of thought” experience. Key Points:
All o3-mini users will now be able to see a clear “chain of thought” that shows how the model has arrived at its answer, to “make it easier to understand how the model thinks," giving users “confidence in its responses"
In a preview, when a user asked, “Why isn’t it Friday?” it showed its logical train of thought, but also considered the sentiment of the question, concluding it was humorous, so responded with a mix of logic and fun
Plus, o3-mini will now also search and fetch data from the web, in real-time, providing links to sources, which will enhance the reliability of its answers, and give users greater transparency
OpenAI filed patents for smart jewelry, VR/AR headsets, smartwatches, and smartphone replacement devices, while also launching its Deep Research tool
Snapchat is launching an AI text-to-image research model for mobile devices. It runs 100% locally on-device which makes it much cheaper. Snap says the model can produce high-resolution images in around 1.4 seconds on an iPhone 16 Pro Max and it will be used to power features including AI Snaps and AI Bitmoji backgrounds. This week, Snapchat also shared growth data about its subscription product, Snapchat+ 131% YoY doubling its subscribers to 14M
Tinder is launching an AI feature to help users connect with “better quality matches” to stay competitive, attract more singletons, and reverse its declining user figures (active users are down 10%, YoY)
Oura Ring is implementing on-device AI to deliver health insights through smartphones rather than the cloud, prioritizing user privacy by processing locally
Kraken launched a crypto payments service, enabling merchants to accept cryptocurrencies seamlessly
📚 Good reads
Evaluating company archetypes and making the right career choices. Many people choose jobs for immediate financial needs, while others can be selective. Elena Verna shows how understanding different company archetypes helps to make informed career decisions
How to know if your idea is the right one — a founder’s guide for successful early-stage customer discovery. Jeanette Mellinger, BetterUp's Head of UXR and former Head of UXR for Uber Eats, unpacks her 3-step playbook for early-stage customer discovery so founders can build something users really want
A presentation on the evolving landscape of the AI voice market. The voice agent market exploded in H2 2024. One data point: companies building with voice represented 22% of the most recent YC class. Voice agents are also being added as a capability to more horizontal or multi-modal products. This excellent presentation from A16Z’s Olivia Moore tells you everything you need to know
Defensibility in a world of commoditized AI models. The AI landscape has significantly changed, with many new entrants and increasing traction for AI-first Products. New AI applications are currently benefiting from a novelty effect, leading to impressive initial growth; but as AI becomes commonplace, Products will need effective distribution strategies to overcome market saturation. Andrew Chen argues that network effects and distribution will be king, once more
How Atlassian built a universal navigation for all its products. Atlassian decided to unify its navigation design across its entire product portfolio. The result? With over 30,000 users across 250 organizations using the new navigation, only 2.7% have switched back to the old navigation — far better than the 5% typically expected for changes of this scale. Put another way, 97.3% of users are choosing to stay with the new navigation
The 5 most powerful ways sales teams can use Operator, by The Neuron. Ever wished you had an AI assistant that could actually do your work, not just chat about it? OpenAI's new Operator tool might be exactly that—so The Neuron tested it out on real sales workflows to find out
That’s a wrap for this week! 🌟
I’d love to hear your thoughts—what stood out to you, and how are you thinking about integrating these insights into your product strategy? Reply to this email or drop a comment on Substack to share your take. And if you found this valuable, forward it to a fellow PM or product enthusiast who’d enjoy the read
See you next week! 👋
Angel