FastLane, Gmail polish and an AI ban
Plus: YouTube fights TikTok with a new strategic Chrome extension
The Weekly Briefing is a product-led perspective of what’s happening in tech - and why it matters to product teams. Paid subscribers also get access to in-depth DoP Deep dives to learn from the strategies and new features released by top tech companies, The Knowledge Series for sharpening your tech skills and more.
Hi product people 👋,
Could the future of SaaS spending be powered by autonomous AI agents? One startup has raised $8.5 million to bet that it will. Skyfire is a new product that allows businesses to assign their own digital wallet to an AI agent, where the business can deposit a set amount of money they want the agent to spend. The agent can then log into various third-party services and make payments on the company's behalf. Skyfire doesn't build the agents, but rather the infrastructure that allows the agents to make payments, with a dashboard for users to monitor spending patterns.
For product teams working in SaaS, this could ultimately become the way SaaS spend is managed which could mean SaaS businesses optimizing for agent monetization through APIs built specifically for agents as well as humans. But given that the company only announced its new funding this week, we’re not quite there yet.
Another new payment update worth knowing about this week is the latest new update from PayPal. This week the company announced that it is partnering with European payment giant Adyen to bring PayPal’s new Fastlane to the US. Fastlane is a new guest checkout experience that PayPal says helps shoppers convert 80% of the time at a 32% faster rate than traditional guest checkouts.
Meanwhile, the folks at development platform Vercel have released a new beta version of their text-to-UI generative AI tool. It allows product teams to do things like generate pricing table components from scratch, simply by describing them. But one of the most interesting new features is its ability to create architectural diagrams and visualisations of technical concepts from prompts or code snippets (e.g. “create a diagram to explain the lifecycle of a HTTP request”). Super helpful for less technical members of the team who need to get their heads around specific concepts.
Finally, speaking of visualisations, if you find it difficult to keep up with various new AI announcements, this neat tool takes the most important ones and transforms them into visual mind maps.
Enjoy the rest of your week!
Key reads, resources and tools for product teams
New from the Department of Product Substack this week:
Deep - How are companies driving engagement and retention?
A carefully worded notification was found to single-handedly drive a significant percentage of DuoLingo's retention and engagement numbers. And a retention tactic from Spotify even managed to increase engagement for an under-used feature by a massive 20,000%. Find out more in this in-depth analysis on the tactics top tier tech companies are deploying in 2024 to drive engagement and retention. (Department of Product)
UX - Why pop up toasts are bad UX
Real world examples of pop-up toasts in action on products like YouTube and Gmail with some suggestions on what to do instead. (Max Schmitt)
Skills - How to build a product prototype with generative AI in 30 minutes
From Idea to Clickable Prototype in 30 minutes with Generative AI. This guide shows you how to use tools like ChatGPT and others to build a prototype quickly. (Rolf Mistelbacher)
Case studies - What Figma learned from Atlassian’s culture of “developer joy”
“We routinely missed almost everything on the public roadmap,” says VP of product Matt Schvimmer. “It was just, ‘We’ll change Q2 to say Q3.’” Beyond a productivity hit, developer employee satisfaction scores were less than 50%. (Figma engineering blog)
Tools you can use
Sparkle - organize the files in your downloads, desktop and document folders automatically
ThinkPost - diagram and brainstorm on a split screen
Codegen - analyse and optimize your entire codebase
Interview - GitHub’s CEO on AI and the future of programming
New product features and innovation
Gmail is getting a new “polish” feature that will refine email drafts on web and mobile.
Luma AI has released Dream Machine 1.5 - a new version of their flagship text-to-video model. It offers enhanced realism, improved motion tracking and more intuitive prompt understanding. One of its biggest new capabilities though, is its ability to render text in videos from prompts. See some examples of it in action here.
Peloton is adding a new feature which will allow users to read Kindle books while exercising. “No more awkwardly holding a book or Kindle ereader while trying to maintain proper form” says the press release from Amazon. But while this is a neat new feature for users who predominantly use Peloton for fitness activities, it reflects a wider strategy by Peloton of offering multiple entertainment options for users who don’t necessarily want to use Peloton screens for exercise.
OpenAI is still debating whether to release its AI detection software publicly but Grammarly is one step ahead. A new AI detection feature called Grammarly Authorship is coming to Google Docs and other document editors and is aimed at educators who want to spot students who have used AI. Use of similar tools in the workplace is probably inevitable, too.
LambdaTest has launched a new end-to-end testing product called KaneAI, designed to automate and speed up the QA process.
YouTube is launching a new Chrome extension that will allow creators in the US save a product while browsing a retailer's website. An interesting example of a highly targeted use case for browser extensions which tie to a product’s broader strategy (in this case, boosting Shopping purchases to compete with TikTok Shop). PS, did you know that YouTube implemented a glowing button whenever a creator says “smash that like button?”
📈 Product data and trends to stay informed
Are we entering the AI backlash era? Procreate’s CEO has pledged to ban generative AI from its core app and more companies may follow suit. For now, earnings calls show more tech leaders are mentioning AI but how long will this last?
7 of the 10 top software companies use the sparkles emoji to represent their AI functionality according to new analysis. “This is the biggest example of an entire industry effectively saying, ‘OK, this symbol is our representation of this kind of feature,’” said the editor of Emojipedia. One company hasn’t adopted it though: Apple Intelligence instead uses a brain icon.
A new software engineering assistant called Genie from Cosine has achieved a score of 43.8% on the new engineering benchmark SWE-bench test, beating the previous score of 19.27% - the largest ever improvement in this benchmark.
The Federal Trade Commission has banned AI-generated reviews - a move which could have implications for product teams who have built genAI capabilities in ecommerce products.
The average headcount of a seed stage tech startup has dropped from 6.9 in 2021 to 5.3 in 2024.
Waymo is now giving more than 100,000 paid robotaxi rides every week across its three main commercial markets in Los Angeles, San Francisco and Phoenix.
📅 Upcoming live programs
If you’re looking to level up your technical skills as you return from the summer, the live Web Technologies program kicks off September 28th.
Set over 4 weeks, you’ll get the chance to work with an instructor and other participants from all over the world to get a deeper understanding of the technical aspects of product development.
The Web Technologies program is designed to help you become more technically confident - not transition into engineering. Early bird pricing gets you $150 off.
Other product news in brief
Cohere’s CEO has warned that selling access to AI models via APIs is quickly becoming a “zero margin business”.
OpenAI has signed a multi-year deal with Condé Naste.
Ex-CEO Eric Schmidt said WFH impacted Google’s ability to compete in AI.
The Department of Product Weekly Briefing is written by Rich Holmes. If you enjoyed it, hit the ❤️ like button below!