AI search ads will expand quickly but lack clear measurement and brand safety standards, which raises risks for marketers. Platforms like OpenAI and Google are testing ads, but early ROI data will be limited and controlled by the platforms. Experts recommend a “trust but verify” approach as marketers wait for better metrics and third-party validation. There are also significant brand safety concerns due to sensitive topics, misinformation, and the personal nature of chatbot conversations. Performance measurement may shift away from clicks toward actions like being recommended or chosen, though early tests show that clickthrough rates lag behind traditional search.
New research shows that most American marketers lack foundational marketing knowledge. Half couldn't define penetration, over 40% didn't understand positioning, and two-thirds failed to identify a quantitative research method. The strongest variable separating those who passed from those who failed was formal training, which made marketers 6x more likely to demonstrate baseline competency. Formally-trained marketers also reported being more influential, more strategic, and more satisfied in their careers.
Journalists receive 200+ emails a day and ignore most of them, so relevance and targeting are critical for successful media pitches. Small, focused outreach lists outperform mass emails. Strong pitches lead with a clear story idea, headline, and angle instead of long press releases. Relationships increase response rates because familiar and reliable contacts get priority in crowded inboxes, but trust depends on being responsive and delivering on time. Data and timing can improve results, such as using search trends to show demand. Quality versus quantity depends on goals, so it should align outreach strategy to whether it needs broad visibility or one high-impact placement.
AI can automate LinkedIn analytics using Claude and Playwright scripts to pull metrics like impressions, saves, and follower growth into Notion. This removes manual tracking and copy paste work. It also enables faster content scaling. Top posts can be repurposed into blogs, patterns can inform guidelines, and AI writers can be trained on tone.
Lowe's is using its Creator Network to build brand affinity with Gen Alpha well before they become homeowners. Launched last June, the program includes partnerships with creators like MrBeast, who introduced a curated storefront of tools and DIY projects and served as a building partner for season two of "Beast Games". The strategy focuses on making home improvement feel approachable and fun rather than driving immediate purchases. Lowe's sees creators as especially important for reaching younger women who increasingly make buying decisions based on who they follow on Instagram and TikTok.
Carta, Ramp, and Webflow aren't guessing at AI search. See the 4 Plays That Drive Citations and Pipeline [10] from 15M+ queries. This playbook reveals the exact system these teams use across ChatGPT,
This webinar on April 8 covers how to expand from enterprise to midmarket without weakening positioning or splitting go-to-market strategy. It focuses on evolving ICP and personas, adapting messaging and sales motion for midmarket expectations, and keeping a consistent brand across segments. It also explores how AI is changing buyer behavior and how to move quickly without diluting your core identity.
LinkedIn Ads delivered sub-$10 CPLs in a test campaign that spent less than $1,000 promoting a 23-page demand gen playbook to B2B marketing decision-makers. The document ad format, which lets users preview content before downloading, drove a 76% lead form completion rate. Audience research using SparkToro, Reddit, and client meeting notes shaped both the content and targeting before anything was created. Of three ad variations tested, the one with the hook "Steal our best demand gen ideas" captured nearly all impressions and drove the best results.
Sentry puts about 70% of its growth budget into awareness, focusing on podcasts, influencers, and OOH. Performance channels are plateauing, with ~60% of Google searches ending without a click and organic CTR dropping ~40% when AI Overviews appear. At the same time, YouTube drives about 16% of LLM citations. Their approach is phased, starting with niche ICP channels, then expanding to broader audiences and cultural affinities.
AI is shifting from general tools to personalized intelligence that mirrors an individual's mind. Sentience creates digital clones trained on personal data like emails, notes, and conversations, which allows it to recall information, act on a user's behalf, and answer questions with high context accuracy. This approach contrasts with platforms like ChatGPT and Claude, which rely on broad datasets and produce more uniform outputs over time. The system functions as a memory and context layer across apps, addressing fragmented data and missing reasoning behind decisions. Memory and context are emerging as key AI breakthroughs that unlock new use cases, similar to how smartphone features enabled companies like Instagram and Uber. The concept raises privacy and ethical concerns, but it also offers a path to preserve individuality and personal “taste” in an era where AI risks becoming standardized.
Wispr Flow [4] turns your voice into clean, formatted text in any app. Dictate specs, user stories, and project updates at 4x the speed of typing. Flow strips filler words, fixes grammar, and formats everything automatically.
Great products align with human behavior, not just features. Focus on core values, create commitment, and make actions simple, engaging, and actionable.
Build product-market fit by focusing on the right users, measuring what matters, and refining based on real feedback. Prioritize quality, delight, and pricing discipline over growth at all costs.
Building for the future of enterprise software means knowing where it's heading. The Gartner® Predicts 2026 report gives you the insights you need to build products users love.
Becoming an AI PM requires strong judgment in choosing problems, tools, and context—not just technical skills. Build real products, understand core AI systems, and focus on practical application over theory.
Product-market fit comes from disciplined prioritization, not more features. Focus on metrics, rank ideas by impact versus effort, and iterate quickly with tight feedback loops.
Great product intuition is spotting bad ideas early, not picking winners. Use AI as a skeptic to pressure-test assumptions before committing time and resources.
High-quality experimentation requires clear metrics, strong defaults, and shared ownership. Embedding quality into tools and workflows improves reliability and better product decisions at scale.
Apple has complete access to Google's Gemini model in its own data centers. The company is able to edit the model as needed to make sure it responds to queries the way it wants, and also use the answers and reasoning information from the model to train smaller, cheaper models. The distillation process will enable Apple to create models that can run on its devices without the need to connect to the internet. Apple is still working on its own AI models, distinct from the Gemini models.
Meta and YouTube have been found negligent for operating a product that harmed kids and teens and failed to warn about those dangers. The companies have been ordered to pay $3 million to the plaintiff of the trial, with the jury saying that they should pay another $3 million in punitive damages. Both companies plan to appeal the verdict. More than 3,000 similar lawsuits against Meta, YouTube, Snapchat, and TikTok are pending in California courts.
CERN scientists transported antiprotons on a 10-kilometer drive around the campus in a truck. Moving antimatter particles is a major challenge, as when antimatter comes into contact with matter, it annihilates and disappears in a flash of energetic particles. The scientists have found a way to trap antiprotons inside a special ion trap. This will allow them to be transferred to other, quieter facilities where they can be studied with greater precision.
NASA Administrator Jared Isaacman jokingly referred to Carlos Garcia-Galan, the space agency's new Moon base manager, as the 'Lunar Viceroy' during the Ignition event on Tuesday. Garcia-Galan is a long-time NASA employee who has been largely anonymous in the past. He plans to focus on the most relevant things to the critical path and on removing choke points that are slowing the organization down. He says the agency has the budget to support all of the activities he has planned on and near the Moon.
Most teams are stuck in stage one: manually maintaining wikis nobody trusts or reads. Atlassian's new AI Maturity Curve [13] for Knowledge Management charts the path to AI-native systems that keep themselves current and surface contextual answers. Join the WEBINAR ON APRIL 16 for a walkthrough of each stage with live demos [13].
Claude was trained on data, and it developed its own strategies, so the people who built it didn't really know how it worked. Anthropic built a set of tools to trace the actual computational steps the model takes to produce an answer. Its team traced Claude's internal computations across tasks over the course of multiple research papers published in 2025. This article reviews these papers to see what the Claude researchers found.
AI probably won't be able to replace engineers within the next few decades. Businesses currently still need great engineers. AI may generate plausible code, but a human still has to verify it, because people pay for reliable services. This guide shows how to hire great software engineers who get things done, sometimes with AI.
Q Day is the point at which existing quantum computers can break public-key cryptography algorithms. Everyone should prepare by adopting post-quantum cryptography (PQC) algorithms. Google is now giving itself until 2029 to prepare for the event. The beta version of Android 17 will add support for developers to have PQC keys for signing their apps and verifying other software signatures.
The moat around SaaS has weakened. However, a reduction of margins is not the same as being dead. Companies are unlikely to be able to prompt their way to a full functional HR system, accounts payable platform, or enterprise CRM anytime soon. Software businesses are living systems that require maintenance, verification, security, compliance, and ongoing refinement.
Learn from the teams building the next generation of AI with practical tools and hands-on experiences building alongside Microsoft product & engineering teams. June 2-3, San Francisco. Learn more and register today [19].
Google's TurboQuant is a compression algorithm that reduces the memory footprint of large language models while also boosting speed and maintaining accuracy. It reduces the size of the key-value cache so it doesn't have to be recomputed. Early testing shows TurboQuant results in an 8x performance increase and 6x reduction in memory usage without a loss of quality. Compression techniques like TurboQuant could improve the quality of outputs of models for edge devices without having to send data to the cloud.
ARC-AGI-3 was designed to evaluate agentic intelligence via interactive reasoning environments. Beating it will mean an AI system matches or exceeds human-level efficiency on all environments upon seeing them for the first time. 100% of the environments are solvable by humans on first contact with no prior training or instruction. All frontier AI reasoning models currently solve under 1%.
Reflection is a startup leading an effort to create freely available US AI systems. It is one of a handful of Nvidia-linked startups seeking to build a network of open source AI models. The startup is in talks to raise $2.5 billion at a valuation of $25 billion. Investors describe Reflection as the 'DeepSeek of the West' as it offers an alternative to the open source models offered by Chinese companies.