Briefing: Google Gemini, Twitch's shut down, Meta voice whispers
Plus: How to manage stakeholders, GraphQL explained, Reddit’s redesign
Hi product people 👋,
here. If you’re one of the hundreds of new subscribers who has recently joined us, welcome! The DOP’s Weekly Briefing is sent every Thursday and is a hand curated collection of what’s happening in tech - and why it matters to product teams.This week, Google Deepmind announced its GPT rival, Gemini. It is described as the most powerful model ever built by Google and benchmarks shared by the company ahead of its release claim that it outperforms OpenAI. One of its more intriguing features is a new concept Google calls Bespoke UI which goes beyond the traditional chat interface to present specific bits of information in different formats. With Bespoke UI, Google is betting on chat interface fatigue kicking in some time soon as users tire of having multiple conversations with bots throughout their workday (coincidentally, another startup called Visual Electric this week demo’d its product which does away with chat interfaces, too).
But as impressive as all of this looks, Google risks running into some branding and product positioning issues here if it’s not careful. Deepmind, Google, Gemini Nano, Gemini Pro, Gemini Ultra, Bard and Assistant are some of the names Google has used in recent months to showcase its AI efforts. Contrast that with Microsoft’s simple Copilot branding or OpenAI’s embrace of ‘GPT’ and it’s not difficult to imagine consumers getting confused. Despite this, Gemini looks very impressive and it will be fascinating to see how its launch early next year shakes things up.
Other products worth exploring this week include this new tool for Figma which allows you to transform designs into animated gifs or videos - ideal for when you need to put together a quick product demo or showcase to a stakeholder or potential new customer.
Meanwhile, product team favorite Asana reported its earnings this week and the company reported a net loss of $8.2 million with an increase in revenues to $166.5 million. Despite a challenging economic climate Asana recently unveiled a new AI-infused product roadmap at the Work Innovation Summit which you can explore here.
Enjoy the rest of your week!
Essential reads for product teams
Freshly published from the Department of Product this week:
🧠Knowledge Series - What is GraphQL? - What is GraphQL? Who is using it and why? A closer look at how PayPal, Netflix and Shopify are using GraphQL, what the future holds for GraphQL and its potential downsides, too.
📈Chartpack - Product Marketing Trends worth knowing in Q4 2023 Building products is one thing, but if nobody knows about your product, what’s the point? In this Chartpack: important trends to know about including data and privacy laws, omnichannel experiences and AI. Plus, practical steps product teams can take to work effectively with marketing teams.
(Department of Product)
Skills - How to manage stakeholders with competing visions
When multiple stakeholders in an organization have competing visions and different success criteria for completing a project, it’s almost inevitable that conflict will ensue. (Harvard Business Review)
Design process - How to use the IMPACT framework for product design
Each component of IMPACT - Illustrate, Mobilize, Prototype, Articulate, Cultivate, and Teach - represents a critical facet of how you can leverage design thinking to make a tangible difference, no matter the constraints of your corporate structure. (Better by Design)
UX - Address validation in form design
47% of ecommerce sites have no address validation at all. In this piece of analysis by Baynard Institute, find out how to introduce address validation properly. (Baynard Institute)
Strategy - Usage-based pricing models explained
Is a usage-based pricing model appropriate for your product? In this piece, investors at A16z argue that usage-based pricing generally works best for SaaS products whose end user is other software, while subscription-based models tend to work best for SaaS products with human end users. (A16z)
Tools you can use
Vercel v0 - build UIs instantly from scratch, using AI prompts
TableFlow - an open source CSV importer, alternative to Flatfile
Hitab - transform your homepage into a dock with smart icons
New product features, launches and announcements this week
Microsoft’s Copilot is getting a bunch of new updates to celebrate its one year anniversary. Some of the ones worth knowing about for product teams include Code Interpreter, which will help engineers perform more complex tasks including writing code, visualising data and more. Deep Search is another new feature which promises to deliver a new, optimised search functionality for complex topics.
Discord has launched a major refresh of its mobile app. The new redesign has been inspired by user feedback and is designed to help declutter the UI. “Over time, the amount of *stuff* that Discord can do has grown exponentially.” Discord Group Product Manager Francesco Polizzi wrote in a post describing the changes. It’s always encouraging to see product teams willing to take a step back and admit their UI has become overly complex with time - and to cut back and organise features where necessary.
Meta has unveiled v2 of its AI translation model, SeamlessM4T. The new version includes a feature called “SeamlessExpressive” which is designed to mimic the variations in human voices. Listen to an example of a human whispering:
TikTok has expanded its concert ticket booking functionality to users outside the US. The feature is powered by an integration with Ticketmaster and allows artists to sell concert tickets directly through the app. The partnership is lucrative for Ticketmaster but there’s a strategic risk that TikTok might do to Ticketmaster what it to Shopify and ultimately build its own native experiences feature.
Reddit has unveiled a new logo and brand design language. Product designers are wondering: does this mean the flat design trend is over?
📈 Product data and new trends to stay informed
iOS beats Android for 30 day retention rates and at day 30, most apps can expect to lose over 90% of users who installed the app. Knowing this before launching an app should hopefully help you set expectations internally; it is not necessarily a disaster if your app hits ~10% retention after the first 10 days - and to some extent, it is to be expected.
88% of Microsoft Copilot users said they would choose the tool over a free monthly lunch provided by their employer. 77% say they would choose it over a weekly lunch.
WhatsApp is secretly becoming a major platform for publishers. The New York Times has over 4 million WhatsApp followers, The Atlantic over 2 million and the FT 100,000+.
“Human favoritism” is a new concept emerging thanks to genAI. Humans prefer human generated content - but only if they know it’s produced by humans. When they don’t know, humans prefer the AI generated content instead. Fascinating study by MIT.
The European tech startup ecosystem has rebounded in 2023. Capital invested in startups has grown YOY to $45 billion, but the continent still lags significantly behind the US and China. Full State of European Tech report.
How to boost your product’s retention rates
If you’re looking for ways to boost retention for your product, here’s some tactics you can deploy:
Define activation - how quickly a user can get to the value your product offers is referred to as activation. Don’t burden users with cumbersome sign up screens and onboarding - get to them to experience your value proposition, quickly. Defining what ‘activation’ means in this context, will help you to measure it. For example, for an ecommerce business, it might be to add an item to a cart. For a gaming app, it might be to play one game.
Understand your value and retention window - what value does your app actually offer and what impact does this have on engagement and retention? News or social media apps should expect to be used regularly (daily), whereas a tax return SaaS product might only be used once a month. Extending your window of retention means not assuming a 30 day period is enough to measure retention. Instead, consider the unique value your app offers and the appropriate retention window for that.
Smart, value-driven notifications - Instead of sending generic push notifications, send contextually relevant and value-driven alerts. For instance, if you have an e-commerce app, instead of just sending a "Sale is on!" notification, you could send "Your wishlisted item is now on sale". By making notifications timely, relevant, and beneficial, you're more likely to see users returning without feeling bombarded.
Build retention boosting features - building features which incentivise users to return can boost your retention rates. When DuoLingo introduced its ‘Streaks’ feature for example, it boosted its DAUs.
📈Get the full Chartpack on Mobile app trends and benchmarks
Other product news in brief
Spotify is cutting 17% of its staff as it continues to push for profitability.
Klarna’s CEO says the company has implemented a hiring freeze, thanks to AI productivity gains.
Twitch is planning to shut down its operations in South Korea as it was unable to make it operating model work. Streamers are not happy.
The DOP Weekly Briefing is a product-led perspective of what’s happening in tech - and why it matters to product teams.
If you want more than the Weekly Briefing, paid subscribers also get access to DOP Deep dives, The Knowledge Series and Product Databases.
If you enjoyed this week’s Department of Product Weekly Briefing, hit the like button below! ❤️
Thanks for sharing the IMPACT framework! 🙏