Matt Anderson

How to stop micromanaging PMs and teams — Mariah Craddick (The Atlantic) The world's leading product community

Using objective tools like the Reforge PM Competency Model and holding dedicated "Development Conversations" every six weeks removes subjectivity from performance reviews and provides actionable growth plans.

From @randysilver.outofowls.com:

www.mindtheproduct.com/how-to-avoid...

How to actually prove AI agent ROI Outcome-based pricing sounds ideal for AI agents, but measurement and dependency risks break it. Learn how phased ROI unlocks real value.

Both vendor pricing and internal business cases often fail because organizations lack the infrastructure to reliably define and track "outcomes" versus simple usage metrics.

www.mindtheproduct.com/how-to-actua...

Archetypes vs. Personas (Video) Personas and archetypes are different ways of communicating the same user research data. Archetypes describe categories of users; personas humanize those categories to illustrate real impact.

Archetypes vs. Personas www.nngroup.com/videos/arche...

Building Digital Trust: An Empathy-Centred UX Framework For Mental Health Apps — Smashing Magazine Designing for mental health means designing for vulnerability. Empathy-Centred UX becomes not a “nice to have” but a fundamental design requirement. Here’s a practical framework for building trust-fir...

Building Digital Trust: An Empathy-Centred UX Framework For Mental Health Apps www.smashingmagazine.com/2026/02/buil...

Today’s organisations don’t have an AI problem — they have a thinking problem AI has made work faster almost everywhere. But, are many organisations confusing sheer speed for actual organisational intelligence?

Today’s organisations don’t have an AI problem — they have a thinking problem uxdesign.cc/todays-organ...

Getting carried away: When intelligence is replaced by compliance Exploring navigation as something bigger than moving through physical and digital spaces — as a fundamental cognitive act

Getting carried away: When intelligence is replaced by compliance uxdesign.cc/getting-carr...

4 Seconds: How Fast Board Directors Decide You're 'Not Strategic True story: Last month, I decided to record myself presenting to a board member. I thought I’d done well. Strong content. Clear roadmap. Data-backed recommendations.

4 Seconds: How Fast Board Directors Decide You're Not Strategic www.productcoalition.com/p/i-recorded...

What makes generated UI worth keeping? How brand, data, and reuse help ideas move past the demo phase

What makes generated UI worth keeping? uxdesign.cc/what-makes-g...

Anthropic's breakout month Why it may be a while before we declare a winner in the AI race

The AI market is not currently zero-sum, as data shows businesses are not switching from OpenAI to Anthropic, but are instead increasingly paying for both services simultaneously.

econlab.substack.com/p/anthropics...

The Identity Crisis of Design How Design Lost Clarity, Gained Titles, and Forgot Responsibility

The Identity Crisis of Design medium.com/p/the-identi...

4 Proven Methods to Get the Most Out of ChatGPT-5 ChatGPT 5 is the latest version of OpenAI’s AI model. This model shows better performance in several areas, including improved accuracy on…

3 Proven Methods to Get the Most Out of ChatGPT-5 uxplanet.org/3-essential-...

AI-Moderated Interviews: If, When, and How to Use Them AI interviews offer faster feedback at scale, but they're not a replacement for in-depth, human-led semistructured interviews.

AI interviewers are best used to complement human researchers rather than replace them; they excel at collecting structured data at scale but lack the critical judgment and domain expertise required for in-depth, semistructured discovery.

www.nngroup.com/articles/ai-...

Awkward Silences | #180 - What to Fix in '26 with Maria Rosala of NN/g & John Whalen of Brilliant Experience Ben is joined by Maria Rosala from Nielsen Norman Group and John Whalen from Brilliant Experience to wrap up 2025 and make some predictions for 2026. Maria and John share practical insights on ever...

Researchers are transitioning into "orchestrators" of both human and artificial intelligence, requiring them to balance AI efficiency with the irreplaceable value of human insight.

podcast.userinterviews.com/episodes/180...

Time to refresh the accounts you follow by hitting "follow all" on ALL THREE 🦋 starter packs...follow 300+ product managers in 15 seconds

go.bsky.app/9UvJJ6g

go.bsky.app/GCjK7ya

go.bsky.app/HgZVU3P

AI's First Substitution: Freelancers Ramp data shows firms are replacing freelancers with AI…at a steep discount.

Businesses are rapidly shifting their spending away from freelance marketplaces like Upwork and Fiverr toward AI model providers; more than half of the businesses using freelancers in 2022 have stopped doing so entirely.

econlab.substack.com/p/ais-first-...

This is how I use AI for Design Systems at Monday.com www.youtube.com/watch?v=7jeo...

Don’t Outsource Analysis to AI (Video) When you outsource your analysis to AI, you risk more than just bad insights — you risk your credibility. Learn 4 reasons why relying on AI for qualitative analysis can backfire and why critical think...

Don’t Outsource Analysis to AI www.nngroup.com/videos/dont-...

Debunking The Myth That Search Is Dying Graphite partnered with Similarweb to analyze over 40k of the largest sites in the US to evaluate the trend in organic search traffic. We provide substantial evidence that the prevailing narrative tha...

Contrary to myths of a 25% to 60% decline, organic search traffic to the top 40,000 websites decreased by only -2.5% year-over-year, with traffic to search engines actually increasing slightly in 2025.

graphite.io/five-percent...

How AI Literacy Shapes GenAI Use Using generative AI often doesn’t mean using it well. AI literacy requires both prompt fluency and the ability to assess outputs.

Prompt fluency and output literacy do not always develop at the same rate, leading to "naïve power users" who can write sophisticated prompts but often accept inaccurate or hallucinated information at face value.

www.nngroup.com/articles/ai-...

Why your brain rebels against redesigns — even good ones The redesign tested well. Users hate it anyway. Welcome to the paradox that costs companies millions and leaves everyone baffled.

Why your brain rebels against redesigns — even good ones

uxdesign.cc/why-your-bra...

Everybody Wants to Have an API Product API-as-a-Product is a journey and not the final destination.

Everybody Wants to Have an API Product bpedro.medium.com/everybody-wa...

How to plan your year with hypotheses instead of guesses Annual planning should not be about capacity. Instead, it should start out as defining a clear goal that's achievable through in-scope activities and a legible roadmap.

Prioritize Input Metrics over Output Metrics: Instead of focusing on lagging financial indicators like revenue, teams should identify and target "Proxy Metrics"—specific user behaviors that are proven to drive business results.

www.mindtheproduct.com/how-to-plan-...

Software as clay on the wheel Dries is the Founder and Project Lead of Drupal and the Co-founder and Executive Chair of Acquia.

Human value is moving "upstream," where the primary role is to define clear success criteria and validate results, while AI agents handle the iterative labor of coding.

dri.es/software-as-...

How Mastercard turned Token Connect into a $3bn scalable growth engine How Mastercard turned Token Connect into a $3bn scalable growth engine

How Mastercard turned Token Connect into a $3bn scalable growth engine
www.mindtheproduct.com/a-framework-...

Reimagining our front door Using positive psychology to craft delightful sign-in & sign-up experiences

Reimagining our front door
medium.com/microsoft-de...

Practical Use Of AI Coding Tools For The Responsible Developer — Smashing Magazine AI coding tools like agents can be valuable allies in everyday development work. They help handle time-consuming grunt work, guide you through large legacy codebases, and offer low-risk ways to implem...

AI can "hallucinate" or produce "slop," so developers must act as critical reviewers, testing all output and verifying AI suggestions against official documentation before shipping.

www.smashingmagazine.com/2026/01/prac...

AI Won’t Fix Your Broken Delivery System. Here’s What Will. You’re solving the wrong problem. And now you’re solving it faster.

AI Won’t Fix Your Broken Delivery System. Here’s What Will. by @paulisthrivving.xyz

medium.com/thrivve-part...

Ideas Aren’t Getting Harder to Find—Asterisk For half a decade we’ve been worrying that ideas are getting harder to find. In fact, they might just be harder to sell.

While it is often assumed that scientific "low-hanging fruit" has been picked, evidence suggests that the real bottleneck is a decline in "allocative efficiency," where the market fails to reward innovative firms with increased market share.

asteriskmag.com/issues/12-bo...

Beyond Generative: The Rise Of Agentic AI And User-Centric Design — Smashing Magazine Developing effective agentic AI requires a new research playbook. When systems plan, decide, and act on our behalf, UX moves beyond usability testing into the realm of trust, consent, and accountabili...

Agentic behavior is categorized into four distinct modes—Observe-and-Suggest, Plan-and-Propose, Act-with-Confirmation, and Act-Autonomously—allowing users to choose the level of control they maintain over the system.

www.smashingmagazine.com/2026/01/beyo...

Despite the need for constant supervision, developers report high levels of enjoyment and significant productivity boosts, viewing agents as essential collaborative partners rather than autonomous replacements.

arxiv.org/pdf/2512.14012