Today's Guide to the Marketing Jungle from Social Media Examiner...
presented by

The weekend is almost here, Alluser! Here's a recap of the most important insights, trends, and updates from the week. Catch up in minutes and go into next week prepared.

In today's edition:

Instagram CTA engagement performance
Testing for ineffective ad placements
LinkedIn Strategy: Person profiles vs.company pages
The Value Ad Framework for videos, with examples to model
2 ways to create quality video with Veo 3
Facebook Updates: Utility Messages, Partnership Ads, and More
Testing for and adjusting for 6 types of bias in your AI tools
Worried AI is dulling your critical thinking? Try this.
🗞️ Industry news from Anthropic, OpenAI, Pinterest, and more

Instagram Growth: Likes vs. Reposts vs. Shares

If you're still prioritizing Likes and Reposts on Instagram, you might be missing out on the platform's latest growth tactic. A simple experiment from Brock Johnson echoes what Instagram chief Adam Mosseri has hinted at for months.

Brock ran two nearly identical posts, each asking followers to engage differently. One CTA focused on Reposts, the other on Shares to Stories. The results revealed surprising disparities in reach, engagement, and algorithmic performance.

For marketers, this insight offers clarity and a strategic edge. Read more here.

Are Your Meta Ads Placements Delivering Results or Just Appearing To?

Surprisingly low CPMs and high click-through rates can feel like wins, until you dig deeper.

For advertisers using Meta's automated placements, misleading performance metrics often signal a larger issue: your ads may be over-delivering in low-value environments.

If you're investing in Meta campaigns focused on traffic, video views, or awareness, this post offers a clear framework for identifying problematic placements using data you already have access to.

You'll learn how to diagnose the root cause of inflated results, interpret placement-level performance, and make informed adjustments without compromising overall campaign efficiency. Read more here.

A Combined Strategy for LinkedIn

If you're a solopreneur or running a small marketing team, chances are LinkedIn feels like a juggling act. Should you prioritize your personal profile, your Company Page, or both?

Overwhelmed by where to start?

Michelle J Raymond breaks down the Power of Two strategy, showing why it's not about choosing one over the other—but how to use them in tandem to build credibility and drive results.

This episode cuts through the noise and gives you a lens to prioritize based on your unique business goals. Watch more here.

Were You One of the Lucky Ones, Alluser?

Earlier this year, thousands of smart marketers descended on Southern California with one purpose in mind…to give their marketing a special boost that you can't find anywhere else.

"This was my first time at Social Media Marketing World, and it exceeded every expectation I had. The value of the sessions and networking experiences cannot be overstated," said Maddie Young.

"OMG, what a conference that was! This was beyond anything I was expecting. I'm already having Social Media Marketing World withdrawals 😂," said Rehema A.

Whether you are trying to figure out how to get more reach, exposure, and sales—or you're trying to understand how AI tools can help you become a better marketer—this is the place you wanna be.

Don't be left behind. Transformation happens when you attend Social Media Marketing World.

Get your All-Access ticket for 50% off.

Positioning Strategy: How to Ensure Your Message Connects

This positioning strategy is a comprehensive process that involves formulating testable hypotheses, conducting thorough market research, and analyzing results to validate or invalidate your marketplace assumptions.

Conduct Multi-Source Research to Validate or Invalidate Your Hypotheses

Once you've established foundational hypotheses, you need to gather comprehensive information from three distinct sources. Brand strategist Stephanie Roland developed the following research methodology to ensure a complete understanding of the market landscape.

Each research component serves specific purposes and generates unique insights unavailable through alternative approaches.

The combined results will provide comprehensive market understanding, audience insights, and competitive intelligence.

Research Public Forums

The first research component involves an online investigation of platforms where authentic conversations occur that may never appear in Google alerts or conventional research sources.

Reddit threads and Quora discussions represent invaluable repositories of authentic market insights. These platforms reveal where target audiences actively discuss the problems you solve and how they evaluate solutions they're considering.

For example, one of Hermes's clients sells solutions to insurance brokers. Research revealed a Reddit community containing twenty-six thousand brokers actively discussing quoting processes and systemic inefficiencies. Neither the client nor the research team previously knew this community existed, yet it provided extraordinary insights into authentic pain points and market frustrations.

You can post questions when conducting this research on these platforms, but your questions should exclusively support information gathering. No promotion should happen in these environments.

Interview Internal Teams

The second research category involves holding structured group conversations with internal stakeholders to uncover your team's knowledge regarding market processes and their perceptions of audience dynamics, market positioning, and competitive landscape.

The questions you ask will vary based on your objectives.

One particularly effective technique, borrowed from GoodQuest, involves asking team members to compose a company eulogy. How would the organization prefer remembrance if it ceased operations? This emotional exercise reveals deeper organizational purpose beyond profit generation and uncovers authentic motivations and values that drive your organization's mission and values.

Another creative approach is to ask your team to consider the company as an actor. What would define its signature roles?

During these interviews, pay close attention to everyone's facial expressions and reactions, particularly during video conferences. Understanding enthusiasm levels, neutrality, and potential resistance to a hypothesis helps you prepare for the final presentation of a strategy shift and ensures proactive rather than reactive concern management.

Interview External Stakeholders

The third research component is often the most valuable yet challenging to obtain.

These conversations should encompass current customers, lost customers, and customers who selected competitive solutions.

Deciding who to interview depends on your research focus. Underwriting product research might target senior underwriters, CIOs, and innovation officers. For a marketing event, you'd approach marketers through LinkedIn, offering incentives like gift cards for participation. Interviews with prospects who selected competitors help you understand why they rejected your solution.

Start your interview by focusing on relationship building and understanding the daily challenges faced by your consumer. Many organizations consider this relationship development time-wasting, but it establishes a level of comfort that proves essential for participants to share authentic information.

Be sure to ask about work challenges, potential product assistance, current solution inadequacies, competitor failings, and category overpromising tendencies. The objective is to understand the market perspective comprehensively, not merely solicit opinions about your specific solution.

Analyze Your Research Data

Once you've compiled information from all three sources, systematic data analysis will help validate or invalidate a hypothesis.

Research documents and transcripts create substantial manual analysis challenges. AI can help you process this information more efficiently while identifying patterns you might overlook.

For example, compile all your transcripts into a single document and ask ChatGPT to identify hypothesis-related themes.

  Here is one hypothesis, and here are all the interviews. Extract commonalities or elements related to this hypothesis.

Repeat this process individually for each hypothesis.

You can also prompt the system using multiple approaches for different data perspectives to request theme extraction related to specific topics, or identify common pain points or desires mentioned across interviews.

Google Veo3: Creating Pro-Quality AI Video Without a Film Crew

Google Veo 3 is not a free service; a credit system operates as the currency for video generation. Different video types and quality levels require varying credit amounts.

When using Google Veo 3, you can choose between the fast model, which produces high-quality results for about twenty credits per eight-second clip, or the quality model, which costs one hundred credits per clip.

Samuel recommends using Google Veo 3 via your personal Gmail account coupled with one of Google's AI plans. Each of these plans gives you access to other AI tools he uses in his video production process, such as Google Flow and Google Whisk.

With the Pro plan, you can generate fifty video clips using the fast model. This allocation allows for substantial experimentation and content creation while maintaining cost efficiency.

The Ultra plan provide significantly more generation capacity, making it suitable for heavy users or professional content creators who need to produce large volumes of video content.

How to Create Quality Video with Google Veo 3

When you're ready to generate your first video clips, access Google Veo 3 from within Google Flow, and choose your video creation option.

Text-to-Video Generation

Most creators will find text-to-video generation sufficient for their needs, particularly when starting with Google Veo 3. This method provides excellent results while maintaining straightforward workflows that don't require additional image creation or complex setup procedures.

Start by assembling your pre-production elements into a comprehensive prompt that includes the scene description, character DNA (detailed visual description), voice DNA (detailed audio characteristics), and specific dialogue or actions for the scene.

The scene description should include environmental details, lighting conditions, camera angles, and any relevant context.

You should also be explicit about audio requirements. While Google Veo 3 can generate background music and sound effects, Samuel typically instructs the system to focus only on character dialogue and essential scene audio to prevent jarring transitions between eight-second clips that might have different background audio elements.

The reasoning behind this instruction relates to post-production control. Rather than accepting whatever background elements the AI generates, Samuel prefers adding music and sound effects during post-production with your favorite video editing tool, ensuring smooth transitions and consistent audio throughout the final video.

  Here is the scene: [detailed scene description]. Here is the character description: [comprehensive visual details]. Here is what I want the character to say: [specific dialogue]. Here is the voice description: [detailed audio characteristics]. Create this scene with character dialogue only, no background music, no sound effects, no ambient audio.

Next, select the generation mode. You have your choice of Fast and Quality.

The choice between Google Veo 3's fast and quality models affects both cost and output characteristics. Samuel consistently uses the Fast model because the quality difference doesn't justify the five-fold cost of the Quality model.

Frames-to-Video

For creators seeking maximum control over their video output, Google Veo 3's frames-to-video feature allows you to use photographs, AI-generated images, or any visual content as the foundation for video creation. The AI analyzes the provided images and creates movement, animation, and visual storytelling based on your additional prompts.

Samuel considers this method "the ultimate when it comes to character consistency" because it starts with exactly the visual you want rather than hoping text descriptions will generate appropriate imagery. When you control the input image, you dramatically improve the consistency and quality of the output.

The pre-production work begins with creating precise images using Google Whisk. Samuel designs individual characters, backgrounds, and complete scenes in Whisk, iterating until he achieves exactly the visual style and character appearance he desires.

He then develops prompts describing how the image should be animated, what characters should say, and what actions should occur within the scene.

This method enables complex scene creation where Samuel can design specific character interactions, environmental details, and visual compositions before animation begins. He might create an image showing two characters in a particular setting, then animate their conversation and interactions through the frames-to-video process.

While frames-to-video requires additional preparation time compared to text-to-video generation, the results justify the investment for creators prioritizing visual quality and consistency.

Once satisfied with the static images, Samuel brings them into Veo 3 via Google Flow and selects the frames-to-video option then generates his clips.

Pro Tip: Use Frames-to-Video With ElevenLabs to Create Video From Images of Real People and Voice Cloning

Samuel has experimented extensively with using personal photographs of himself as input for frames-to-video generation, allowing Veo 3 to create appropriate lip-sync and facial animations for the dialogue. The initial audio doesn't sound like him, but the visual elements are correctly synchronized.

He uploads the final video to ElevenLabs and extracts the original audio track from the generated video and processes it through ElevenLabs' voice cloning system to replace the original AI-generated voice with a cloned version of his voice, maintaining the same timing, cadence, and dialogue content.

This workflow requires training ElevenLabs on the target person's voice characteristics beforehand. Once trained, the system can swap voices while preserving the original timing and synchronization, resulting in videos that feature both the person's actual appearance and their authentic voice.

For businesses wanting to create personalized content featuring actual team members or spokespersons, this methodology provides a pathway to professional-quality video production without traditional filming requirements.

Utility Messages Transform Business Communication

Facebook has introduced utility messages for Messenger, creating a new category of business communication that operates independently of traditional marketing message limitations. This feature functions similarly to transactional emails in email marketing, where businesses can send essential operational updates even to customers who have unsubscribed from promotional messages.

Utility messages enable businesses to send timely, important updates about order confirmations, appointment reminders, shipping notifications, and account alerts. These messages bypass the typical Messenger marketing restrictions, ensuring that critical business communications reach customers regardless of their marketing message preferences.

The implementation uses pre-approved templates, which significantly simplifies compliance with Facebook's platform policies. Previously, businesses struggled with complex and frequently changing rules around Messenger marketing. The template-based approach eliminates much of this complexity by providing clear, approved formats that businesses can customize for their specific needs.

This feature is particularly valuable for e-commerce businesses, service providers, and local businesses that rely on appointment scheduling. For example, a restaurant can send reservation confirmations, a retail store can provide order updates, and a medical practice can send appointment reminders – all through Messenger, regardless of whether customers have opted into marketing messages.

The utility messages feature is available for developers and businesses in the United States, Philippines, Thailand, and Vietnam, with a broader rollout expected in the coming months. This represents Facebook's replacement of the existing message tags system, streamlining the process and making it more accessible to businesses of all sizes.

For businesses currently using Messenger for customer communication, this update provides a more reliable and compliant way to ensure important operational messages reach customers. The pre-approved template structure also reduces the risk of message rejection or account penalties that could occur with the previous, more complex system.

Professional Dashboard and Content Management Enhancements

Facebook has significantly upgraded its professional dashboard, bringing key performance insights and management tools to the forefront of the user experience. This refresh addresses many of the pain points that business users have experienced with Facebook's creator and business tools over the years.

The updated dashboard prominently displays crucial metrics, including earnings and engagement data, making it easier for businesses to assess their content performance and monetization results quickly. The interface has been redesigned to mirror the mobile experience more closely, reducing frustration when switching between desktop and mobile management of Facebook business accounts.

New metrics have been added to provide deeper insights into content performance, including views over time, view rate, and retention analytics. These metrics allow businesses to understand not just how many people saw their content, but how engaging that content was and where viewers might be dropping off during video consumption.

Two significant features are currently in development and will be available soon. The first is data export functionality, which will allow businesses to extract their performance data for analysis in external tools or for record-keeping purposes. This addresses a long-standing request from businesses that need to integrate Facebook data with their broader analytics and reporting systems.

The second upcoming feature is bulk uploads, which will dramatically improve the content publishing workflow for businesses that regularly post multiple videos. Currently, businesses must upload videos one at a time, which can be time-consuming for content creators and social media managers handling multiple pieces of content daily.

The comments manager tool has also been updated to maintain consistency between desktop and mobile experiences. This seemingly small change has significant practical implications for businesses that manage customer interactions across different devices throughout the day.

These dashboard improvements reflect Facebook's broader commitment to supporting business users and content creators with more professional-grade tools. The updates make Facebook's business tools more competitive with other platforms that have historically provided better analytics and management interfaces.

More Features, Tools, Changes, and Updates Discussed Include:

AI-Powered Translation Expands Reach for Facebook Reels
Advanced Advertising Features and Partnership Enhancements
Brand Protection and Community Management Updates

Watch it on YouTube

How to Remove Bias in Your AI Tool Outputs

You trust AI tools to save you time, spark ideas, and scale your marketing, but what happens when those same tools reinforce the very biases you're working to overcome?

This eye-opening episode from Katie Robbert and Christopher Penn shows you how to put AI models like ChatGPT, Gemini, and DeepSeek through a real-world bias benchmark test.

From subtle shifts in tone to outright disparities in feedback based solely on gender-coded names, you'll see firsthand how bias creeps into everyday outputs. More importantly, you'll learn a replicable framework for running your own bias audits—no PhD required. Watch more here.

How to Sharpen, Not Dull, Your Critical Thinking Skills With AI

Too many marketers hand over thinking to AI, treating it like an answer machine instead of a strategic partner. The real edge isn't in what AI gives you, it's in how you question what it gives you.

Liza Adams shares a practical framework for using AI to sharpen your critical thinking, not dull it. From social media content planning to pricing decisions and ABM strategy, you'll see how just one extra layer of questioning can flip generic prompts into high-impact insights.

You'll learn how to ask better questions, see clearer patterns, and stress-test assumptions before your next campaign or client pitch. Read more here.

Instagram Adds Retention and Skip Rate to Reels Insights: Instagram is rolling out a new retention chart in Reels insights, offering creators a visual breakdown of viewer engagement over time. The interactive chart highlights exactly where audiences drop off, helping creators fine-tune content to maintain viewer interest. Alongside this, Instagram will soon replace the view rate metric with a more actionable skip rate, which shows the percentage of viewers who exited the Reel within the first 3 seconds. Together, these tools aim to help creators improve their hooks and retain audiences through the end of their videos. Instagram via Threads

Meta Licenses Midjourney: Meta has struck a licensing deal with Midjourney, a leader in generative AI imagery, to embed its acclaimed aesthetic technology into future Meta models and products. The move enhances Meta's ability to compete with rivals like OpenAI and Google by improving visual features across its platforms and potentially lowering content creation costs. Notably, Midjourney will maintain its independence, operating without external investors even as it scales its impact. Meta Engineering via Threads

Caption Length and Instagram Reach: Instagram chief Adam Mosseri clarified that using long captions has no significant impact on how far a post reaches in the app. Responding during his weekly Instagram Stories Q&A, Mosseri encouraged users to write longer captions if they want to, noting that while not essential, they can enhance storytelling and engagement for those who choose to use them creatively. Social Media Today

Reddit Pro Goes Mobile with New Tools for AI-Era Discovery: Reddit has rolled out a major update to Reddit Pro, now available on iOS, equipping businesses with mobile tools for real-time community engagement and brand monitoring. The update includes new performance insights that track the impact of brand comments, helping businesses understand where and how they're building trust. Reddit also introduced its first-ever Organic Playbook to guide authentic participation and brand strategy on the platform. With Reddit now the top source cited by AI platforms, building an organic presence has become essential for brands seeking visibility in AI-generated recommendations. Reddit

YouTube Rolls Out Templates, Expands Hype, and Boosts AI Summaries: YouTube is enhancing mobile video creation with the launch of templates in its YouTube Create app for Android, giving creators customizable video structures and royalty-free music to streamline production. The platform's Hype feature—designed to elevate smaller creators—is expanding to 17 new markets and will soon include topic-based leaderboards and community-driven sharing tools. Viewers can post hyped videos directly to their feeds, and established creators will be able to spotlight rising talent. Meanwhile, YouTube is broadening access to AI-generated video summaries, offering quick content overviews that complement but don't replace traditional descriptions. YouTube

GPT-realtime and Realtime API Launches: OpenAI has launched gpt-realtime, its most advanced speech-to-speech model, alongside the general availability of the Realtime API. Designed for developers building production voice agents, the API now supports image inputs, SIP phone integration, and remote MCP servers. The new model delivers more natural and expressive speech, better instruction adherence, and higher accuracy in tool usage. Developers also gain access to two exclusive voices—Cedar and Marin—and a 20% pricing reduction. OpenAI

Claude for Chrome Enters Pilot Phase: Anthropic has launched a pilot for Claude for Chrome, a new extension that enables the AI to interact with web pages by clicking, filling forms, and reading content on users' behalf. Available initially to 1,000 Max plan users, the tool introduces site-level permissions and action confirmations to enhance security. The company's red-teaming tests revealed vulnerabilities to prompt injection attacks, but newly implemented mitigations have significantly reduced risk. The pilot aims to refine safety systems through real-world usage before broader release. Interested users can join the waitlist at Claude.ai. Anthropic

Gemini App Gets Major AI Image Editing Upgrade: An advanced image editing model within the Gemini app, offers enhanced control and creativity. The update focuses on maintaining consistent likeness in edited images, allowing users to transform photos with costume swaps, scene changes, and multi-image blending. New capabilities include multi-turn editing, style transfers, and realistic alterations while preserving subject identity. All generated images are visibly marked and include SynthID watermarks to ensure transparency. Users can start using the new features in the Gemini app today. Google

Meta Launches REST API Framework: Meta has unveiled a new REST API framework aimed at simplifying and standardizing developer integrations. Built on the OpenAPI standard, the new system addresses longstanding feedback about the complexity of the Graph API and offers a more streamlined, contract-driven approach. Developers can expect broader adoption of the REST framework beginning in late 2025, with it becoming the default for all new APIs by 2026. Existing Graph API users can maintain current integrations without immediate changes. The REST framework is already being used in tools like the Llama API for generative AI. Meta

Marketing API to Default Dynamic Media to Opt-In for Catalog Ads: Beginning September 1, 2025, the Marketing API will default the Dynamic Media setting media_type_automation to OPIT_IN for all new Advantage+ Catalog ads in Single Image, Carousel, and Collection formats. This update brings the API in line with Ads Manager's default settings, aiming to boost creative performance through greater use of video assets. Developers should modify their API configurations by September 1 to avoid unintended media activation, with full rollout concluding by October 20. Advertisers are advised to review catalog video readiness and use preferred_video_tags where needed. Meta

Pinterest Academy Rolls Out Two New Badges and Multi‑Language Support: Pinterest Academy announced two new digital badges: a gold badge for completing the Media Planner Certification and a silver badge for the Performance Essentials skills badge. In a major expansion of accessibility, the Performance Essentials training is now offered in seven languages (English, Italian, French, German, Spanish, Portuguese, and Japanese), while the Media Planner Certification is available in French, German, Spanish, Portuguese, and Japanese in addition to English. Pinterest

Snap Introduces App Power Pack for Enhanced App Campaigns: Snap has unveiled the App Power Pack, a comprehensive suite of advertising tools designed to boost app installs and engagement on Snapchat. This package includes dynamic tCPA bidding, immersive Sponsored Snaps in Chat, auto-generated App End Cards, and interactive Playables now in alpha. Supporting both SKAN and non-SKAN campaigns, the tools are delivering at least a 25% lift in installs and seamlessly integrate into new and existing campaigns. Snap

X Adds View Quotes Button to Post Interaction Bar: X is testing a new View quotes button to streamline how users access quote posts. X

What Did You Think of Today's Newsletter?

😠 😞 😐 😃 🎉

Did You Know?

Only male crickets chirp, and the frequency of chirping increases as the temperature rises. You can estimate the temperature outdoors by counting cricket chirps for a given time period.

Michael Stelzner, Founder and CEO

P.S. Add michael@socialmediaexaminer.com into your contacts list. Use Gmail? Go here to add us as a contact.

We publish updates with links for our new posts and content from partners. Your information: Email: tukangpostoemel@gmail.com Opted in on: 2021-09-06 17:20:47 UTC.