KIMP’s Picks September 12th: Weekly Highlights In Design, Marketing, AI, & More

As we wrap up another week, it’s time to talk about the latest social media buzz. From major AI: news to buzzworthy social media algorithm changes and shareworthy posts we came across, here is a quick roundup to close the week on a productive note. Are you ready for the recap we have curated for you? Then let’s get started with the KIMP’s Picks September 12th Edition. 

In the Marketing Realm 

Google’s updating Asset Studio 

Google is upgrading its Asset Studio with new generative AI tools to streamline the creative process for marketers. The new features, powered by Imagen 4, will help users create high-quality, on-brand creative assets more efficiently.

The idea is to simplify the process of creating visuals for ads. Marketers can now transform simple product shots into realistic lifestyle images using text prompts. They can also apply changes to up to 100 photos at once, saving significant time. The new tools also allow for a style reference feature, ensuring all new visuals align with a brand’s unique look and feel.

Google Ads mistakes that SMBs should avoid 

Search Engine Land shared a resourceful post delving into some common Google Ads mistakes that many small business owners tend to make and some tips and tricks to avoid them. 

The post highlights how many small businesses struggle with Google Ads because the platform is complex. It explores common pitfalls that often waste ad spend and prevent growth. The biggest mistakes include a lack of a clear strategy, poor campaign structure, and not focusing on the user experience.

So if you are wondering if you are spending too much or spreading yourself too thin when it comes to Google Ads, this is a great post. 

AI’s impact on marketing operations 

AI is changing how marketing teams work and in a recent post, MarTech shared an insightful discussion on four significant ways in which the integration of AI is changing marketing operations. 

The post discusses some common implications like the enhancements that AI brings to data management and how teams handle insights. It also talks about the use of hyper-personalization to boost customer experience. Additionally, the post covers predictive analysis, forecasting and automation of marketing workflows. So, if you are wondering how AI is changing marketing workflows around the world and how your team can keep up, this post is for you. 

Amazon & Netflix join forces to boost streaming ads 

In a surprising move, Amazon and Netflix are teaming up to let advertisers buy Netflix’s programmatic ad space through Amazon’s demand-side platform (DSP). This partnership follows similar deals, like Amazon’s agreement with Disney and Netflix’s partnership with Yahoo.

The goal is to simplify TV ad planning by offering marketers a wider, smarter pool of viewers. Programmatic ads, which use algorithms to target audiences, will now be more streamlined. Advertisers hope this helps reduce repetitive ads and deliver more precise campaigns.

For marketers, this opens up a bigger and smarter ad marketplace. And the collaboration comes as a reminder that advertising is evolving! 

Contextual targeting in a privacy-first era 

With privacy rules tightening and ID-based tracking fading away, contextual targeting is back and smarter than ever. ADWEEK discusses the evolving trends in contextual targeting in the post here. 

Unlike traditional methods that simply matched ads to a page’s topic, modern “contextual audiences” analyze broader content consumption patterns to identify user interests and intent. This allows for a blend of audience and contextual targeting without relying on cookies or personal data. 

This approach ensures compliance and also boosts the consistency of performance across channels. By embracing these smarter, privacy-first tools, marketers can future-proof their strategies while driving better results today.

Tapping into the power of video ads on LinkedIn 

LinkedIn recently shared some insights highlighting the significance of video ads on the platform and some strategies for B2B brands to make the most of this format. 

Recent data shows 70% of buyers rely on video to understand business challenges, and video ads on LinkedIn get three times more engagement than static ads. With more than a billion professionals on the platform, marketers can target decision-makers actively searching for solutions.

In addition to this, LinkedIn now offers multiple video ad options like in-feed videos, pre-roll BrandLink ads, Connected TV ads, and event promotions. These formats help brands tell stories, build trust, and reach audiences where they’re most attentive. Partners like Adobe and QuickFrame also support video creation, campaign management, and performance tracking.

In the Content Marketing Realm 

Writing a blog post 

Blogs remain one of the crucial formats in content marketing. The Semrush post here breaks down some key strategies to write blog posts that convert, posts that bring traffic. It breaks down the process into 12 actionable steps any content marketer can tackle. 

From choosing the the topic to performing keyword research, and drafting an outline all the way up to writing the post and publishing and distributing it, the post details the steps and useful tips for each. 

In the Design Realm 

Heinz’s Looks Familiar campaign 

Heinz recently unveiled its “Looks Familiar” campaign. The global push cleverly links the brand’s iconic keystone logo to the shape of French fry takeout boxes. It suggests fast-food packaging mimics Heinz’s design, tying ketchup to fries.

The campaign’s genius lies in leveraging a simple design element to spark recognition. Brands can learn to anchor campaigns in unique visual identities, use everyday cultural moments for relatability, and add playful storytelling to deepen impact. Heinz shows design can transform a logo into a global conversation.

NFL kicks off 2025 season with AI-enhanced “You Better Believe It” campaign

The NFL has launched its 2025 Kickoff campaign, “You Better Believe It” blending live-action, CGI, and generative AI. This high-energy spot features a surreal, music-video-style parade float packed with fans from all 32 teams.

While there have been mixed responses to the campaign, the creative spectacle that it is demonstrates the magic that can happen when you combine AI with human creativity. Instead of resisting AI, now is a good time to embrace its capabilities to create something spectacular. But the smart move is to enhance the work of AI with human creativity to create on-brand visuals that accurately represent your brand. 

Tootsie Pop leverages nostalgia 

Tootsie Roll Industries recently launched a refreshed version of its iconic “How Many Licks” campaign, reviving the beloved 1969 TV spot featuring Mr. Owl and Tootsie Pops. The ad, a staple of American pop culture for over 50 years, follows the classic storyline – a curious child asks the wise but impatient owl how many licks it takes to reach the Tootsie Roll center.

This design-focused revival modernizes the original 2D animation with brighter colors, high-definition clarity, widescreen formatting, and subtle updates. As more and more brands lean into surreal visuals and AI-generated imagery, Tootsie Roll’s decision to go back to retro 2D style and its use of nostalgia stand out. 

In the AI Realm 

“Select & Ask” feature in Gemini Canvas 

Gemini Canvas has introduced a new feature called “Select and Ask” which allows users to visually edit their web applications without writing any code. This update is designed to simplify the web development process and make it more accessible.

Canvas is a feature in Gemini that opens a collaborative workspace for coding, content writing and more. 

This new feature allows users to simply select any component of their web app on Canvas and then describe the change they want to make. The edits are then applied instantly, and users can see the modifications live in a preview. This “no-code” approach removes a major barrier for those without technical coding skills.

Google launches AI Quests to teach kids about AI 

Google Research, in partnership with the Stanford Accelerator for Learning, has launched “AI Quests” a new educational program for students aged 11 to 14. This initiative uses a game-based format to teach young people how to apply artificial intelligence to solve real-world challenges in areas like climate, health, and science. 

AI Quests is part of Experience AI and is available globally for educators and organizations. The program encourages hands-on learning, critical thinking, and creativity, showing students how AI can address real problems while building foundational skills in AI literacy.

Tencent announces HunyuanImage 2.1 – an open-source text-to-image model 

Tencent recently announced an open source text-to-image model, HunyuanImage 2.1, marking a significant advancement in AI-driven creativity. This model is notable for its native 2K image generation, which produces high-quality visuals efficiently.

A key feature is its Advanced Semantics, allowing it to handle complex prompts of up to 1000 tokens and accurately generate multiple subjects in a single image. The model also excels at Precise Chinese and English Text Rendering, making it ideal for creating designs that seamlessly integrate text, such as posters and covers.

Moreover, the model also supports the generation of images in multiple styles like photorealistic portraits, vinyl figures and more. The strengths of this model come from the integration of a multimodal large language model (MLLM) and a multi-language character-aware encoder. This ensures superior text and image alignment as well as better text rendering.  

Anthropic adds file creation and editing capabilities to Claude 

Anthropic has introduced a new feature to Claude. With this new update, Claude can now create and edit files like Excel spreadsheets, Word documents, PowerPoint presentations, and PDFs directly within its chat interface. This feature transforms Claude from a conversational assistant into an active collaborator, capable of handling complex, multi-step projects.

Instead of just providing text responses, users can now upload data, describe their needs, and receive ready-to-use files in return. Claude can perform tasks such as turning raw data into an analyzed report with charts, building financial models with working formulas, and converting a PDF into a slide deck. The feature is available as a preview for Max, Team, and Enterprise plan users.

ElevenLabs’ Voice Remixing feature – now in alpha

ElevenLabs has introduced Voice Remixing in alpha. This new feature lets users transform their existing voices or designed voices in creative ways. You can adjust gender, age, or accent to craft unique characters and more natural, localized speech.

This feature is ideal for storytelling, agent design, and expanding businesses across regions. For example, a UK company can remix a British English agent’s voice to sound American, giving a local feel without re-recording. Voice Remixing works with any cloned or custom-designed voice on ElevenLabs.

Google expands AI Mode to five new languages 

Google’s AI Mode has gained a lot of traction and has changed the way people search and find answers. Now Google is expanding this immersive AI search experience in five new languages: Hindi, Indonesian, Japanese, Korean, and Brazilian Portuguese. This expansion is designed to provide a more nuanced and locally relevant search experience, going beyond simple translation.

With the introduction of new languages, this custom version of Gemini 2.5 powered search feature allows users in these regions to ask complex questions in their native language. According to Google, this technology enables a deeper understanding of local information, delivering more useful and contextually relevant search results.

Google’s Veo 3 upgrades 

Google has announced major updates to its Veo AI video generation models, Veo 3 and Veo 3 Fast. The company has reduced pricing for both models, making them more affordable for developers and creators. Veo 3’s price has been cut from $0.75 to $0.40 per second, while Veo 3 Fast is now just $0.15 per second, down from $0.40.

In addition to the lower costs, both models now support new output formats. Users can generate videos in the 9:16 vertical aspect ratio, making content creation for mobile-first platforms like TikTok and Instagram Reels much easier. They can also create videos in 1080p HD resolution for higher-quality visuals. These models are now stable and ready for scaled production use via the Gemini API.

Stability AI’s enterprise sound production model 

Stability AI has released Stable Audio 2.5, its first audio generation model designed specifically for enterprise-grade use. The new model focuses on speed, quality, and creative control, addressing a significant market gap as only 6% of creative campaigns currently use a unique sound identity, despite custom audio making brands up to eight times more memorable.

This new version delivers faster generation, producing up to three-minute tracks in under two seconds, and smarter compositions that respond better to mood and genre prompts. It also supports audio inpainting, allowing users to extend or refine their own audio clips seamlessly.

Replit launches Agent 3 for a more autonomous app building 

Replit just rolled out Agent 3, their most advanced app-building assistant yet. According to Replit, this new model is 10 times more autonomous and 3 times faster than before, helping you build, test, and fix apps without constant supervision. Additionally, Agent 3 can even create other agents and automations, opening up new ways to streamline workflows.

A major highlight is its App Testing feature. The agent tests your app in a real browser, checking buttons, forms, APIs, and data sources, then automatically fixes issues it finds. With Max Autonomy, it can run for over 200 minutes without manual input, managing tasks and supervising itself. Agent 3’s ability to build other agents lets you create tools like Slack bots, Telegram assistants, and task automations using simple natural language. 

Baidu’s new AI reasoning model – ERNIE X1.1

Baidu unveiled ERNIE X1.1 at Wave Summit 2025, and it’s a serious step forward in AI reasoning. Compared to its predecessor, ERNIE X1.1 delivers major improvements where it matters most – factual accuracy is up 34.8%, instruction-following is up 12.5%, and its ability to act autonomously (agentic capabilities) is up nearly 10%.

In benchmark tests, ERNIE X1.1’s performance is on par with leading global models like GPT-5 and Gemini 2.5 Pro. The model is now available to developers and enterprises through Baidu’s platforms. 

Ralph Lauren x Microsoft announce “Ask Ralph” 

Ralph Lauren has launched “Ask Ralph” a new AI-powered styling companion within its US mobile app, built in partnership with Microsoft. The tool is designed to replicate the personalized experience of an in-store stylist.

Using natural language, users can ask for outfit recommendations for any occasion. “Ask Ralph” responds with curated shoppable outfits from the Polo Ralph Lauren brand, pulling from real-time inventory. The tool is powered by agentic AI, which can reason and act to deliver tailored recommendations. 

Collaborations and tools like these show how brands can use technology like AI to build deeper, more meaningful connections with consumers.

Social Media Feature Updates and Algorithm Changes  

Reddit’s Pro Tools for publishers 

Reddit is stepping up its game for publishers. To help media outlets connect with its 110 million daily users, Reddit is rolling out Reddit Pro tools in beta. These new features let publishers track how their stories are shared, sync articles via RSS feeds, and get AI-powered community suggestions to post content where it’ll get the most traction. 

According to Reddit, big names like The Atlantic and NBC News have already tested the tools with positive results, often seeing Reddit become a top source of traffic.

Reddit is also making it easier for users to read and discuss articles. With a new link viewing experience, users can swipe up to see comments and interact with publishers without leaving the app.

TikTok x Fandango partnership for in-app movie ticketing 

TikTok and Fandango have announced a unique partnership that allows users to purchase movie tickets directly within the TikTok app. This collaboration is powered by TikTok Spotlight, an in-app solution for entertainment promotion.

The new feature adds a “Get Tickets” button on movie-related content hubs. From there, users are seamlessly connected to Fandango’s ticketing flow to select seats and complete their purchase without leaving the TikTok platform. 

Threads expands character limit with a text attachment option 

Recognizing that 500 characters aren’t always enough, Threads now lets users attach up to 10,000 characters of text to a post. This feature makes it easier to share excerpts from books, articles, newsletters, or podcast transcripts, all viewable directly within the app.

Users can also add links to full content, helping followers explore their work across other platforms. The idea is to allow creators to share more descriptive stories that help build anticipation and boost engagement while also sparking meaningful conversations. 

View on Threads

Snapchat’s Imagine Lens – AI-powered Snaps 

Snapchat is taking creativity to the next level with its new Imagine Lens. This is the platform’s first Open Prompt Image Generation Lens, allowing users to create, edit, and recreate Snaps by simply typing prompts. More than 400 million users have already interacted with Gen AI Lenses on the platform over 4 billion times, and this new feature builds on that momentum.

Available exclusively to Lens+ and Snapchat+ Platinum subscribers, the Imagine Lens offers access to advanced AI tools. Users can easily pick or edit prompts, generate unique images, and share their creations with friends, on their Story, or beyond Snapchat.

YouTube expands multi-language audio to help creators go global 

YouTube is rolling out multi-language audio to millions of creators, letting fans around the world watch videos in their native language within moments of release. This lets creators reach diverse audience segments in various regions at the same time. 

According to YouTube, early trials showed impressive results, over 25% of watch time came from viewers using non-primary languages. Big names like MrBeast and Mark Rober are already tapping into global audiences, with some videos dubbed into 30+ languages. Chef Jamie Oliver’s channel saw views triple using this feature.

Additionally, YouTube is also testing multi-language thumbnails, making it easier for creators to localize visuals based on viewers’ language preferences. Creators like Chef Nick DiGiovanni are expanding into new markets like Spanish, Turkish, and Arabic-speaking regions.

To Conclude…

That’s a wrap on the KIMP’s Picks September 12th Edition! Social media evolves fast and it’s easy to feel buried trying to keep up with all the latest trends and tech news. But you don’t have to spend all your time scrolling to stay in the loop. We’re here to make it simple, bringing you the key highlights every week, so you’re always in the know without the hassle. Stay tuned, and we’ll be back next week with more updates!