Sponsored
REPORT]
How much does it cost to get a lead on Google Ads in 2026? This ungated report [4] from WordStream by LocaliQ lays out benchmarks for key metrics across 23 industries.
Full-text search across article titles and summaries. Refine with sector, year, content type, and reading-time filters.
Sponsored
How much does it cost to get a lead on Google Ads in 2026? This ungated report [4] from WordStream by LocaliQ lays out benchmarks for key metrics across 23 industries.
TikTok is collapsing the traditional funnel as discovery, consideration, and purchase now happen in the same content experience. 64% of users buy after discovering a product on the platform, often without leaving it. TikTok Shop and creator-led formats embed commerce directly into content through episodic storytelling and in-feed purchase loops. Winning brands use native, culturally fluent videos that fit the feed rather than polished ads.
High reasoning mode changes AI visibility and should be treated like a separate search environment. GPT 5.2 increased citation rates from 50% to 68% and nearly doubled citations per response, with 4.6X more fan-out searches. Impact is strongest in mid and lower funnel comparison queries that trigger deep research across pricing, docs, compliance, APIs, and product details. Early funnel content also matters since early citations persist through buyer journeys. Brands need content that answers specific sub-queries and consistent visibility across reasoning modes.
Most captions are added as an afterthought and miss what holds attention. The first subtitle should appear at second 0 and stay under 6 words. Key phrases should be enlarged 1.5 to 2X for emphasis. Background contrast like a semi-transparent bar or drop shadow improves readability. Platform auto-captions should be turned off to avoid duplicate text. Subtitles should sit within the middle 60% of the frame to avoid TikTok, Instagram, and LinkedIn UI overlays.
The Swatch x Audemars Piguet “Royal Pop” pocket watch triggered global queues, resale prices up to 15X retail, and emergency stock warnings. Demand for the $400 co-branded product was extreme. The collaboration shows co-branding as a strategic growth lever, not a tactical promotion. The pocket watch format avoids direct comparison with the $25K Royal Oak. It still borrows its distinctive design cues. AP gains exposure to new audiences, while Swatch gains credibility. Both brands increase demand and long-term brand value.
Sponsored
Wistia's 6th annual 2026 State of Video Report [9] is here! Find out what 13 million video and 1,000 video marketers reveal about budgets, performance, and the impact of AI. Also - get benchmarks that can help you answer the age-old question: is this video actually working? Get
An open-source collection of markdown-based "skills" gives AI coding agents like Claude Code and Cursor structured marketing knowledge across 40+ tasks, from CRO and copywriting to SEO audits and pricing strategy. Each skill acts as a reusable prompt framework that agents reference automatically when they detect a relevant task.
Google Ads performance metrics showed unusual stability this year, with cost per lead declining for the first time in five years to an average of $66.69. Conversion rates improved across 87% of industries, suggesting advertisers are adapting to AI-driven features like Performance Max and AI Max. Meanwhile, CPC averaged $5.42, more than double the $2.32 recorded when the report launched in 2016.
Only 20% of the global workforce feels engaged at work, while 44% of US employees report burnout. Under-support is a structural issue that often hits during role changes, unclear leadership, or AI-driven workflow shifts, which forces employees to be their own coach or feedback loop. The remedy is to build lower friction support systems. This means setting clearer priorities, recurring peer check-ins, and more actionable manager communication.
AI is turning senior marketers and operators into high-impact individual contributors who can ship cross-functional work without large teams. Companies should reward business impact over headcount as one person can now do work that once required a PM, designer, engineer, and marketer. Faster execution changes decision-making by reducing the need for layers of approvals and coordination. Teams should hire people who can connect growth, product, and distribution end to end and give them direct access to business context and core metrics. At Lovable, 90% of time now goes to building instead of meetings and alignment.
Creator ads on Snapchat drive stronger performance, including 66% higher excitement, 29% higher interest in other products, and 25% longer watch time than standard ads.
Marketers should launch their expanded Q4 campaigns now to maximize their in-app opportunities.
Tips include ungating your best info, commenting posts, and posting between 8-11 AM local time.
Sponsored
Everyone's building an agent framework. OpenProse [4] has one: it's called English.
AI will not replace product managers wholesale. The best PMs will use it to improve decisions, alignment, and product impact.
Non-technical PMs need to become builders, not backlog managers. AI tools now let PMs prototype, ship low-risk work, and orchestrate agents while preserving judgment, taste, and product quality.
PostHog keeps product marketing fast, practical, and story-driven. Great launches need clarity, strong writing, smart channels, internal alignment, and real customer stories.
Slowing growth is not always a sales problem. In the AI era, product velocity and competitiveness may be the real growth levers.
Sponsored
Veriff's 2026 Deepfakes Report [9] reveals that the average person's ability to detect a deepfake is just 7% ABOVE THE THRESHOLD OF RANDOM GUESSING. Don't leave your platform's trust and safety to chance. Download the full research report here [9] to stay ahead of identity fraud.
User interviews help teams understand what users need, why they behave the way they do, and where products can improve. This guide shows how to run them well, from recruiting participants to analyzing insights.
PMs should use AI agents to automate routine workflows, not replace judgment. The best PMs will become agent orchestrators who protect their taste, focus, and strategic thinking.
AI is moving deeper into product workflows through Notion Workers, agentic Android features, and vibe-coded tools. At the same time, companies are cutting weak AI features and rethinking how to use AI without adding bloat.
AI frustration is a signal, not a reason to settle. The last 10% of craft is where better tools, stronger judgment, and real differentiation emerge.
Skip abstract strategy debates that only sound smart.
Great products need clear positioning.
Sponsored
Agents are smarter when they can search the web. But Google doesn't offer an API for web search results... so how are AI companies doing it?
Andy Jassy took the role of Amazon's CEO five years ago. He recently placed a series of expensive bets on AI that are audacious even by Silicon Valley standards. In his time, he has killed projects and cut staff, pleasing Wall Street, and now he has to steer the tech giant through its greatest challenge yet. This article tells the story of Jassy's tenure at Amazon.
SpaceX plans to launch the newest version of its Starship rocket on Tuesday during a launch window that opens at 6:30 PM ET. The flight plan will be similar to previous Starship test launches. The most recent test flight was seven months ago. Starship has undergone major redesigns and is now bigger, more powerful, and a step closer to being fully reusable.
CAR T cell therapy was originally designed to target and wipe out cancer by reprogramming patients' immune cells. It is now being offered to patients in clinical trials for autoimmune conditions. There's still uncertainty about how well the treatment works for autoimmunity and how any benefits might last, as well as what long-term side effects might arise. Another major challenge is that the therapy can cost hundreds of thousands of dollars after accounting for hospital stays, cell engineering, and other expenses.
The Vera C. Rubin Observatory is designed to study the universe in greater detail than ever before. It is expected to discover a million asteroids, thousands of comets, and billions of stars and galaxies in its first year alone. The facility has begun collecting preliminary images and astronomers are poring over the initial data. Scientists have already been able to make new discoveries with its first images despite the images still not being as sharp as expected and the observatory still requiring a final tuning.
Sponsored
P50 response time tells you almost nothing about performance. The metric that actually matters [9] has to account for recall, grounding rate, re-query rate, and integration overhead. Fast-and-wrong is
Zero is a systems programming language for agents that provides small native tools, explicit effects, predictable memory, and structured compiler output. The language is still experimental and not stable yet, but it is in a stage where users can try it out and provide feedback. Practical examples are available for users to learn the language with.
Agents have gotten really good in the last fifteen months. They can now be used for real work with light supervision. AI can now be used for writing code changes, investigating and fixing bugs, research in large code bases, manual testing, and more. It still isn't suitable for writing public communications, unreviewed code, or testing UIs, but its capabilities are improving quickly.
Many people try to gain employment through the front door: they find a job advertisement, send their resume, and hope that someone on the other side notices them. Most people don't realize there are other entrances into the building. Conversations at parties, warm introductions, cold emails, and visible proof of work are viable alternatives. Sometimes, it's about creating a signal that attracts the exact person who needs it.
The idea that all exponentials eventually become sigmoids might not be necessarily true for AI. There are several examples of exponentials that don't end up becoming sigmoids (or haven't yet), for example, the decline in birth rates or solar power deployment. Things that are currently happening are more likely to continue happening than not. AI is unlikely to stop progressing on all fronts anytime soon.
TLDR is hiring a VP of Marketing to scale growth, pipeline, and brand across 8M+ subscribers, 13 newsletters, and advertisers like AWS, Google Cloud, Microsoft, and Anthropic. Apply here [15].
Hooks make agent workflows more dependable by implementing tests, policies, logs, and completion gates as deterministic parts of the workflow.
Git has been behind the state of the art for a while now, prompting companies like Meta to develop better in-house systems.
Waymo's rapid scaling to 400,000 rides per week proved that AV operators might not need Uber at all, so Uber is pivoting to asset ownership to avoid being cut out entirely.
Sponsored
The more your product succeeds, the more the invoice hurts? Not anymore. Reveal is the embedded analytics SDK with flat pricing,
It is becoming apparent that understanding the domain and how it connects to the program is valuable, which is why spec-driven development is now in the spotlight.
SwiftUI & Apple's native SDKs are constraints when it comes to building rich text rendering for long-form chats.
Changes are happening so fast that nobody is noticing the underlying architecture decaying, and people are denying that this is an issue because they are focusing on the wrong metrics.
Sponsored
Inngest [5] asked 130 engineers about running AI in production—only 19% were very confident their stack could scale, with gaps in tracing being a key issue. 1 in 5 now spend up to HALF their time on reliability work just piecing together context.
Google is rolling out a new 'Thinking level' option for Gemini. The option has appeared for some users when they select Fast or Gemini 3.1 Pro. Google is also preparing to add more integrations with third-party apps in Gemini. Support for Canva, Instacart, and OpenTable appears to be coming.