← Back to leaderboard
shannholmberg

Shann³

Verifiedactive
@shannholmberg

I cover AI marketing & growth. Sharing every framework as I build it. Growth @lunarstrategy @turtledotxyz

RosterAI Agents & Productivity ToolsTech Career & Coding EducationStartup & SaaS Builders
36,901 followers · 12,552 following · Last polled 7h ago
Perf. Score
58
Rank #158
Value Score
37.5
#174 of 273 priced
Median views
2.3K
p25 2.2K · trailing 7d
Followers
37K
+0.2% 7d
ER (impr.)
5.21%
eng ÷ impressions
ER (foll.)
2.17%
eng ÷ followers
Est. CPM
$86.86
QT-rate basis
WoW views
+103.1%
▲ rising

Rates & value · est. from 2.3K median organic views

overpriced
Quote tweetbasis
$200
$86.86 / 1K views
Post
—
no CPM
Thread
—
no CPM
Retweet
—
no CPM
Value Score 37.5 · rank #174 of 273 priced$2173.91 / 1K engagementsrate age unknown (pre-audit import)

Trends

Follower growth
Median views / day
Engagement rate
View distribution · trailing-7d posts

Each blue dot is one organic post. The Performance Score ranks by the median (2.3K), not the mean (15K) — so a single viral post doesn't inflate a creator's reach. The p25 floor (2.2K) is the level 75% of posts clear.

Recent posts · ranked by views

PostPostedViewsLikesRepostsRepliesQuotesBookmarksER
how to get faster and cheaper videos from Opus 5.5 using a decision model for frame checks reduced video review costs by 90% and review time by 75% when you create videos with Claude Opus 5.5, checking thousands of frames can consume a lot of time and tokens you can set this up so a decision model checks the frames, while Opus 5.5 reviews the flagged issues and fixes the code here's how it works: 1. build your video with Opus 5.5 give it your idea, brand book, assets and references, then work through the storyboard and animation if you're using Remotion, it renders the code into a video 2. separate the repeated checks from the creative work use code to check measurable requirements and a decision model to inspect the frames > code checks things like text size, blank or frozen frames and colours against your brand rules > the decision model reads the frames and answers specific questions to help identify problems > selected frames and check results go to Opus 5.5 for further review 3. have Opus 5.5 investigate and fix the problems it reviews the selected frames and edits the video code where needed render the changes and run the checks again this reduces the amount of frame-by-frame review you ask Opus 5.5 to do 4. review the finished video yourself watch the pacing, sound, transitions and whether the video matches what you wanted to create the automated checks help find defects, while you review the creative decisions and overall quality SageCut is one implementation of this for Remotion videos in the example Chris shared, Sage (decision model) read 3,153 frames for $0.16 (the frame-reading cost, not the full production cost) the savings come from the review stage, rendering still takes time start with a video you've already reviewed and check whether it catches the problems you found then compare the review cost and time with your existing processfrozen
4d ago
73K
1.2K
74
53
6
2.5K
5.35%
how to use Opus 5.5 + GPT 6 to make animated marketing videos Opus 5.5 handles the main work, from planning the scenes to screen recording, animation and editing when I need an image, it can call Codex to generate the asset and use it in the video for the video below, I asked Opus to use Codex to generate the display here’s how to set this up: 1. start with your idea and let Opus ask questions write out what you want or transcribe a voice note, then ask Opus to question you before it starts give it the campaign goal, the audience and what you want someone to do after watching include where you’ll publish it, the video length and the format you need you don’t need every detail ready, but you need to make the decisions together 2. give it your brand context and references share videos you like and explain what you like about them that could be the camera movement, the animation, the pacing or how they introduce the product add the assets you already have, including logos, fonts, screenshots and previous work if your brand book is in Paper or Figma, connect the tool and give Opus access to the relevant files it can inspect your designs and use the existing assets without you collecting everything into another folder your taste and feedback give it a direction to follow 3. ask for a storyboard before it builds the storyboard should show what happens in each scene and how the scenes connect ask it to include: > what the viewer sees in each scene > the text that appears on screen > which parts need a screen recording or a generated image > how the elements move and how long each scene lasts > whether the video needs music, sound effects or a voiceover review this together before it starts building the video if the opening takes too long or the product demonstration is unclear, change it here 4. generate the missing assets in my setup, Opus can call GPT-6 through Codex to request images from the image-generation tool you can also generate the assets separately with GPT Image 2 or 2.5, then give the files to Opus describe how the asset will appear in the scene, including its angle, background and available space for text Opus can then use the image in the animation and video edit for the display in this video, I gave Opus the instruction and it handled the request through Codex 5. build the video and review a first render ask Opus to build from the approved storyboard and render a preview rendering creates the video file, and it can take time depending on the animation, resolution and length local rendering can also use a lot of your computer’s resources, so several renders can slow down your other work review a lower-resolution preview before requesting the final export 6. give specific feedback and revise use timestamps so Opus knows exactly what to change > at 00:03, hold the product screen longer so the text is readable > at 00:08, reduce the movement around the display > make the final call to action larger on mobile be specific about audio too, including whether you want music at all if you want generated music, give it a suitable audio tool and references for the direction keep the storyboard, assets and feedback in the project so each revision uses the same contextfrozen28d ago64K295231616081.48%
I used Opus 5.5 to make this ai marketing video it's the first model I've seen that can handle screen recording, animation and editing in one go you just need context and a well-defined brand book we just launched a creative library at @LunarStrategy save and use it to find inspiration for your next visualfrozen28d ago40K931201950.53%
how to use Opus, Sonnet and Haiku 5.5 together in Claude Code (setup prompt included) > Opus advises > Sonnet builds > Haiku handles simpler tasks in parallel you can give each model a different role in Claude Code 1. Sonnet 5.5 handles the main task it coordinates the work, writes code and runs tests 2. Haiku 5.5 gathers the information subagents search files, look up documentation and handle smaller research tasks in parallel they report their findings back to Sonnet 3. Opus 5.5 provides guidance Claude Code’s advisor feature lets Sonnet consult Opus during the task Opus receives the conversation and returns guidance for planning, recurring errors and reviewing completed work start Sonnet 5.5 with Opus 5.5 as the advisor: --- claude --model claude-sonnet-5-5 --effort high --advisor claude-opus-5-5 then copy this prompt into Claude Code: Create a read-only research subagent in .claude/agents/ using model: claude-haiku-5-5. Its tasks are file discovery and documentation lookup. Return concise findings, file paths and sources to the main Sonnet session. Update the project’s CLAUDE .md: - Delegate independent research tasks to this subagent in parallel when useful. - Keep implementation and testing in the main session. - Request advisor consultation before complex changes, after recurring errors, and before completion. Preserve existing instructions. Show the proposed changes before applying them. ---- each advisor consultation adds usage, so use it where a review can helpfrozen12d ago26K480506735664.57%
Opus 5.5 solved motion design but you can still have an edge, the moat, now more than ever is having taste & judgement study how good designs & videos are made, but more importantly, what to AVOID here are 26 things to avoid when creating designs with AI 1. eyebrow text above nearly every heading 2. descriptions that repeat the heading 3. captions that explain obvious illustrations 4. labels on nearly every object, arrow, and section 5. pill shapes used for buttons, navigation, labels, and decoration 6. purple gradients and glowing backgrounds 7. a centered headline followed by three equal feature cards 8. rounded cards inside more rounded cards 9. oversized headings above tiny gray descriptions 10. decorative icons beside every point 11. blinking status dots and “live” badges 12. scrolling text banners repeating the same message 13. abstract 3d objects floating beside the headline 14. continuously moving background particles 15. thin connecting lines with glowing dots moving between boxes 16. colored strips along the edges of rectangular boxes 17. every section appearing with the same fade-and-slide animation 18. arrows added after nearly every link and button 19. short labels separated by dots, such as “strategy · design · execution” 20. a short decorative dash before each section label 21. widely spaced uppercase text across headings, navigation, and captions 22. frosted glass panels over blurred backgrounds 23. cards that tilt as the cursor moves over them 24. a different background effect in every section 25. a small icon inside a colored rounded square at the top of every card 26. the same color repeated across an icon, its tinted background, and the surrounding card borderfrozen25d ago25K27392114152.92%
one role that every company needs right now is a marketing engineer your goal is to use agents to generate more revenue and lower costs and you only need: claude + a data warehouse + github + a vps here are 7 things to build, plus 3 bonus workflows you can start with: 1. shared context for your team and agents build a marketing brain with your company strategy, offers, target customers, brand guidelines and previous campaign lessons keep research, decisions and working files for each campaign in a campaign brain you can keep the knowledge files in a GitHub repo your team can access, with links to designs and larger assets when someone starts a task, their agent can find the relevant context and use what your team already knows 2. performance data agents can use bring ad spend, website analytics, search data and social performance into a data warehouse sync it on a schedule so agents can read historical performance without repeatedly requesting it from rate-limited platform APIs use the platform APIs to create and change campaigns, with direct checks when you need fresh information your agents can compare previous campaigns, find patterns and use those findings to make better decisions 3. tools and skills your team can share keep a shared list of tools, what each tool does and how agents can access it document the permissions, required inputs and expected outputs so your team’s agents know how to use each tool create reusable skills that combine those tools into workflows, with examples and quality checks for the finished work 4. agents that work on a schedule a vps is a remote server that can run your agents when your laptop is closed a paid ads workflow can check campaign performance on a schedule and act within rules you’ve agreed on it could pause an ad that meets your stop criteria or notify you when it recommends changing the creative or budget keep a record of what it changed and why 5. creative production from idea to finished assets build workflows that use your campaign goal, research and brand guidelines to produce finished creative work that could mean producing clips from livestreams, editing podcasts or creating motion design for your social channels give agents your references, source files and feedback so they can follow your design and editing standards 6. evaluations and reviews before work goes live build a review system that catches weak work before your team has to review it evaluations are checks against defined criteria, such as factual accuracy, brand consistency, AI writing patterns and broken layouts give agents approved examples and rejected examples with your notes, then have them revise work that fails the checks keep human review for taste, judgment and decisions the checks cannot reliably make 7. learning from the results connect campaign results to the assets and decisions behind them record what you tested, what happened and what you would change next time review those lessons before adding them to the marketing brain for future campaigns 3 bonus workflows you can build right away: > paid ads workflow: agents research campaign angles, generate ad variations and create campaigns ready for approval. once launched, they monitor warehouse data, pause underperforming ads within agreed rules and produce new variations from the results > campaign launch workflow: build on existing frameworks like Matt’s skills or Claude Superpowers. give your main agent a goal, offer and campaign brain, then have it plan and coordinate the landing page, promo video, social posts and email sequence through checks and approval > webinar-to-content workflow: agents turn a recording into edited clips, captions, social posts, an article and a follow-up email. they prepare approved content for distribution across your channels and track which assets bring visitors and leadsfrozen15d ago20K335224135774.94%
marketing agent infra is one of the most important things businesses need to get right it gives your marketing team the context, data and tools to generate revenue outcomes it consists of 3 main things 1. centralized context / knowledge there are many ways to set up a marketing OS / brain, a simple starting point is an LLM-maintained wiki keep the knowledge in markdown (.md) files, with YAML for details such as the source, date and topic the setup works like this: > raw sources go into folders grouped by source or marketing function > a model such as JEV classifies the information so the system knows where it belongs > an LLM reads the sources and updates the relevant pages in your wiki folder > each page links back to its sources and related knowledge this includes everything from company strategy to the exact formats you use on each channel: > positioning, offers and ideal customer profiles > customer journeys, buying triggers and objections > campaign goals, plans and previous decisions > brand rules, writing style and visual references > channel playbooks, content formats and approved examples > experiments, results notes and your feedback when an agent starts a task, it reads the relevant wiki pages and follows links to supporting material your team resolves conflicting information and reviews important changes before they become approved company knowledge 2. data warehouse this brings together ad spend and results, social stats, Google Search Console data, website analytics, email performance and more the setup works like this: > data pipelines collect information from those systems on a schedule > the warehouse stores it with dates and identifiers for campaigns, assets and channels > shared definitions explain how each metric is calculated > agents query the relevant data and return the results with the reporting period include the metrics your team uses to make decisions: > ad spend, impressions, clicks and conversions > social reach, engagement, saves and video watch time > search queries, impressions, clicks and average positions from Google Search Console > website traffic sources, landing-page activity and conversion events > email deliveries, clicks, unsubscribes and conversions where tracked consistent campaign names, tracking links and identifiers help connect results across sources your agents can then investigate questions such as: > which creative variations generate conversions at a lower cost? > which social formats get saved or watched for longer? > which search queries bring relevant visitors to your website? > which landing pages receive traffic but generate few conversions? agents query the warehouse for historical analysis, reducing repeated requests to channel APIs the collection process still uses those APIs, so it needs to respect their limits and record when each source last updated keep the performance data in the warehouse and save the interpretation in your marketing OS / brain 3. tools + runtime this is how agents use the context and data to do the work tools give them access to your channels and production software the runtime is where the agents run, including the model, computer or cloud environment and task scheduler the setup works like this: > connect the tools your agents need through APIs, MCP connections or browser access > give each agent the files, accounts and permissions required for its task > run recurring tasks through a scheduler, such as cron > record what ran, what changed, what failed and what needs your input for creative work, that could mean access to your design files, image models, video tools and editing software for campaign operations, it could mean access to your email platform, ad accounts, website and social channels a scheduled agent could query yesterday’s campaign performance, compare it with your targets and prepare recommendations using your channel playbooks another could use an approved campaign plan to create asset variations and return them for review set spending limits, retry limits and clear approval requirements before those tasks run if a task fails, the system should record the failure and notify the person responsible, with enough context to continue the workfrozen24d ago15K15382212152.71%
as more people start using opus 5.5 and see how INSANE it is at motion, animation & design the amount of slop is going to increase massively to stand out, you need to know what to avoid and understand what makes good design work build a library of designs that made an impact on you save them, study them, write down what caught your attention and why even though design is subjective, you can train your eye you get better at noticing spacing, type, movement and how everything works together then compare what you notice with what other people respond to less is more, avoid slopfrozen27d ago13K70470881.33%
how to get into marketing engineering (crash course) you can start by having agents handle the research, production and execution behind your paid ads, outreach and SEO Cody's walkthrough is a good introduction, here's what you need to get started: 1. choose your agent and give it context use Claude Code, Cursor, Hermes or Grok Bot give it your company knowledge: summary what you do, offers, target customers, brand guidelines and previous work keep this in a shared folder or GitHub repo so your agents can find the context for each task 2. connect your performance data set up a data warehouse and sync your ad performance, website analytics and Search Console data into it your agent reads the warehouse to understand what is working, then uses the platform APIs to make approved changes this reduces repeated requests to ad platforms and lets your agent compare performance across channels (so you don´t get rate limited) 3. give it tools to complete the work the tools depend on which workflow you're building: > paid ads: copy writing skill, customer research, image generation and access to your ad platform and data warehouse > outreach: company research, contact providers, email verification and your sending platform > SEO: Search Console, DataForSEO API access for keyword research and access to your CMS give the agent instructions for each tool, including what it can do independently and what needs your approval 4. build your first workflow choose a task you already understand so you can judge whether the agent does it well here are three examples from the walkthrough you can build on: paid ads research customer problems → create ad angles and visuals → review the creatives → upload approved ads → check performance outreach find relevant prospects → check their fit → find and verify business emails (enrich) → prepare outreach for review SEO find relevant queries where you already get impressions → review existing pages and search results → update or write content → review and publish work through the steps with your agent and give it feedback on the results 5. save the workflow and put it on a schedule once the workflow works well, save the instructions, examples and checks as a skill use a cloud server if you want it to run when your laptop is closed you could have it review ads each morning, research prospects each week or prepare content updates from Search Console datafrozen6d ago12K218184413785.68%
here is the full list of distribution channels to either build or acquire: 1. founder & team audiences people who sees your team on X, LinkedIn, TikTok or other medias can become customers, partners and referrals build through useful posts and conversations, or hire people with relevant audiences 2. YouTube channels build a library of tutorials, product demos, comparisons and interviews that help people understand your category you can start a channel, sponsor an existing one or acquire a relevant channel and its content ideally its a separate entity to your own company profile 3. your website & search visibility build pages that answer the questions customers ask before buying, including comparisons, use cases, pricing and tutorials work on traditional SEO and visibility in AI answers (ChatGPT, Claude, Perplexity and Google AI Overviews) ranking in search and being cited or recommended in an AI answer are related, but they aren't the same thing you can build this yourself or acquire a relevant website with useful content and an existing audience 4. newsletters & email lists give people a reason to subscribe and keep sending something they want to read build your own newsletter, sponsor one or acquire a publication (check subscriber consent and data-transfer rules before using its list) 5. communities build a place where your customers exchange advice, ask questions and meet others doing similar work this could be a Discord, Telegram group, Slack community, forum or membership business you can also partner with an existing community or acquire one with its team (the relationships need to survive the acquisition) 6. creators & industry experts work with people your customers already follow and trust sponsor their content, create something together or build a long-term partnership where they use and understand your product 7. affiliates & referral partners give customers, creators and partners a reason to recommend you set up a clear offer, tracking and payment process, then check whether those referrals become valuable customers 8. agencies, consultants & resellers work with businesses that already advise or serve your target customers give them the training and support to recommend, implement or sell your product as part of their service 9. integrations & technology partners connect your product to tools your customers already use an integration can create distribution through partner directories, shared campaigns and recommendations during customer onboarding 10. marketplaces & app stores make your product available where customers already look for a solution think app stores, plugin directories, browser extensions and industry-specific marketplaces (the right ones depend on your product) 11. free tools, templates & open-source projects build something useful that introduces people to the problem your paid product solves calculators, templates, datasets, plugins and open-source tools can all give people a reason to discover and recommend you you can also acquire an existing tool with a relevant user base 12. podcasts & industry publications build your own show or publication, appear as a guest or sponsor one that reaches your customers acquiring a niche publication can give you content, subscribers and relationships with contributors 13. events, webinars & workshops bring the right people together around a problem they need to solve run your own sessions, partner with an existing organiser or sponsor an event with a relevant audience 14. paid advertising use search, social, video, newsletter and podcast ads to reach customers beyond your existing audience this is access you pay for, so track acquisition costs and whether those customers stay 15. direct outreach & sales build a repeatable way to identify relevant companies, contact the right people and follow up email, calls, LinkedIn and demos can all be part of it, but the offer needs to fit the person receiving it 16. customer referrals & word of mouth make it easy for happy customers to introduce you to others ask after a successful outcome, give them something useful to share and track which introductions become customers 17. distribution built into the product shared documents, public pages, collaboration invites and customer-created content can introduce your product to new people look for places where sharing is already useful to the customer (forcing an invitation into every action usually makes the product worse) 18. businesses with an existing customer base sometimes the acquisition is a whole business with customers who could also use your product that could be a service business, a complementary product or a niche media company check why those customers stay and whether your offer fits before treating the acquisition as a distribution channel 19. AI agent surfaces make your product available through MCP servers and connectors so agents can use it inside a customer's workflow give agents clear instructions for finding your tools, connecting an account and completing tasks build your own integration, list it in relevant agent directories or partner with platforms that help users find tools (an MCP server alone doesn't guarantee discovery) 20. launch & discovery platforms launch on Product Hunt, share relevant work on Hacker News and submit your product to niche directories and launch newsletters use these for launches and major updates, with a clear next step for people who discover you (plan for a launch event, not steady traffic) 21. review & comparison sites build a presence on G2, Capterra and review sites your customers use when comparing products keep your listing current, ask customers for honest reviews and answer questions about how your product works you can also pay for relevant placements, but reviews need to come from customers (don't buy positive reviews)frozen10d ago9.5K14762312344.34%
with the ability to reverse engineer any software now, it takes me back to an article I wrote at the beginning of the year when code becomes commoditized, the focus should be on building an advantage through distribution channels, talent and acquiring taste & judgment that's why I think technical people with marketing knowledge (marketing engineers) are going to become so sought after a team in the future will be less about different types of technical engineers and more about engineers with specific business knowledge product, marketing, ops... we're not there yet (this is not a doomsday post), but I think we're heading in this direction FAST any company that isn't preparing for competitors with the same ability to build will have a difficult time here are a few things worth investing in now: 1. distribution channels probably one of the strongest advantages you can have as a business if two companies have the same ability to generate code, their ability to distribute that product / service becomes a much bigger factor even before AI, having the most polished product didn't mean you would win you can build distribution from scratch or acquire it: > team members with strong, relevant audiences on social media (X, LinkedIn, TikTok...) and a strong YouTube presence > your own website, with SEO and AI search visibility so customers can find you > a newsletter / email list that gives you a direct way to reach people who want to hear from you (might get obsolete with agent first interactions tho) > partnerships with people and businesses that already reach your customers the audience needs to fit what you're selling (a big following alone doesn't mean people will buy) 2. talent your team will be one of the most valuable things you've got, especially people who understand the above roles are about to become much less defined as one marketing engineer takes on more research, production and execution I think we'll see a big change in how much one person can do (even over the next six months) you want people who take initiative and figure out how to use AI, with the judgment to know what to do and what to leave out as producing more work becomes easier, look for people who can actively filter out slop people who can explain why something isn't good enough and improve it, whether that's a campaign, a design or a product decision look for taste and judgment give them a problem to work through and ask them to explain their choices (including where they disagreed with the agent) I'd want to see how they think, what they notice and whether they understand who the work is forfrozen11d ago9.0K582262361.37%
how to add a decision model to your AI workflow 1,000 yes/no decisions can cost around: ~$10 on GPT-6 Astra ~$4 on Claude Opus 5.5 ~$0.04 on JEV use language models for planning, writing and coding, and decision models for decisions a decision model can: > check your agent’s output before it ships > approve a draft or request revisions > route a task to the right agent or tool > classify leads, tickets or comments > score fit, quality or priority > check whether content follows your brand rules > pick an option from a shortlist > decide whether to continue, stop or request a review at a much faster speed as well than any regular LLM here’s how to add one 1. find the decisions in your workflow look for repeated steps that need a choice, classification or score start with one decision you can compare against examples you’ve already reviewed 2. define the question and allowed answers give the model the information and references it needs for a content workflow, that could include your draft, voice DNA file, anti slop rules and approved examples ask it to return a choice such as approve, revise or human review 3. connect the answer to an action your software uses that answer to decide what happens next approved drafts move to your review queue, revision requests return to the writing agent, and uncertain cases go to you 4. test before letting it act compare its answers with your reviewed examples if it returns confidence scores, check whether those scores match how often it gets the decision right start with actions you can easily reverse 5. put simple checks first use code for fixed rules, such as required fields and character limits then use the decision model for checks that need context, before starting more expensive work 6. track quality and cost measure cost per completed task, incorrect decisions and how often another model or a person needs to review use the failures to improve your questions, references and allowed answersfrozen8d ago8.8K189103132225.19%
easiest way to turn a marketing task you’re working on into a repeatable workflow have your agent do the task end to end while you steer it and give feedback and judgment do this 2–3 times to catch common edge cases and work out how you want it done once you’re happy with the process, give your agent the session IDs or transcripts and ask it to review what you did together save and use this prompt: “create a skill that repeats the task we completed in [session ID 1], [session ID 2] and [session ID 3]. include reference files, my corrections, decision points and examples of good output. use best practices for creating workflow skills” now you have a marketing workflow skill based on how you and your company do the work run it on new tasks, check the results and fix anything it misses then share it with your colleagues or add it to your autonomous Hermes marketing bot to run on a schedulefrozen14d ago7.4K130132301704.56%
Grok Bot is the easiest path to marketing engineering right now the barrier to entry is low, you can get your first bot running in minutes here’s how I would start: 1. set up shared context and performance data create a GitHub repo where your bots can store and access your marketing knowledge keep your company strategy, offers, target customers, brand guidelines and writing skills there, with separate folders for each campaign connect a data warehouse like ClickHouse to centralize performance data from ads, your website, search and social channels your bots can use that data to understand previous results and decide what to work on next 2. give your bots access to the tools they need connect the accounts, APIs and apps each workflow requires > Google Search Console + DataForSEO for search performance, keyword research and competitor analysis > Claude for writing and updating SEO / AEO content > Google Ads + Meta Ads for creating and managing campaigns > Opus 5.5 with coding and rendering tools for motion videos > GPT image models for generating campaign visuals and ad variations > Figma or Paper for brand references, design files and your feedback > your CMS, email platform and publishing tools to prepare work for distribution keep instructions for using each tool alongside your skills, including what the bot can change and when it needs approval 3. build your bots start with one bot per workflow or marketing area you run give each bot a clear task, the context it needs, a schedule and a place to save its work here are a few I’d start with: competitor tracker tracks your five closest competitors, collecting their campaigns, ad creatives, landing pages and customer feedback it compares their offers, messages and formats, then recommends experiments relevant to your business keep observed engagement separate from sales performance, which usually isn’t public market tracker follows your niche on X, Reddit and relevant news sites it identifies recurring questions, emerging topics and changes that could affect your campaigns have it suggest a relevant content angle or campaign adjustment, with sources explaining why it matters content bot turns your research, company knowledge and campaign ideas into posts for X, LinkedIn and other channels you use give it a writing skill for each channel, approved examples and clear instructions for avoiding AI writing patterns save your edits so it can apply them to future drafts SEO / AEO bot uses Search Console and keyword data to find declining pages, unanswered customer questions and useful content opportunities it researches the topic, prepares a brief and drafts new pages or updates using your company knowledge have it prepare changes in your CMS for review and track search results after publication paid ads bot reads campaign performance from your warehouse and checks it against your targets it proposes new creative variations and campaign changes, then prepares them in the ad platform for approval start with monitoring and drafts, then allow specific actions once you’ve tested the workflow check the Grok Bot marketplace for a template close to the workflow you need adapt its instructions, connect your tools and add your company context before running it on a schedulefrozen14d ago7.0K11472001634.33%
Opus 5.5 is insane at animations... I spent max an hour prompting it and waiting for renders the bar to create animations was just lowered again https://t.co/F5QKXUIZTsfrozen29d ago6.6K572100281.48%
you are still early to marketing engineering btwfrozen14d ago5.5K1038801114.18%
easy design hack to get better dashboard UI (I saved it)frozen21d ago3.8K30070211.54%
how to build a creative production workflow your marketing agents can run give your main agent a campaign idea, goal, brand book and references you like have it ask about your experience and any missing context, then work together on the production plan 1. define what the campaign needs agree on the message, deliverables, formats and channels include the details that must stay consistent, such as the product, logo, approved copy and visual style 2. connect the production tools and set limits the main agent selects tools for each step, including image generation, video, editing and finishing Higgsfield’s API connects those production tools, including Genjutsu and Cinema Studio, so the agent can request assets and track the results set spending and retry limits before production have the agent ask for approval before exceeding either limit 3. explore the direction before producing everything have the agent create concepts, a storyboard and sample frames review those together before approving the full production plan and budget 4. check the output and fix specific problems have the agent compare each asset against the approved brief and references if one shot has a problem, have it edit or regenerate that shot before rebuilding the whole video 5. save the work and your feedback your campaign brain is the folder holding this campaign’s context and working documents keep the source files, approved assets and rejected directions there, including why you made each decision for teams testing this setup, Higgsfield offers 100% cashback on paid API usage as API credits, up to $100K per business from a $20M first-come pool cashback must be earned and used by September 30frozen27d ago3.7K330132262.02%
today we’re going live to show how marketing teams can use Opus 5.5 for creative production Kleanthi from our Lunar team will walk through how we built our creative library and made the animations we’ll show the process behind the work and how you can apply it to your own marketing team join us live and bring your questions 11am GMT+1 live on @LunarStrategyfrozen21d ago3.2K460160242.69%
agent-first newsletters could become a valuable distribution channel we're hearing a lot of people say email newsletters are becoming less relevant as agents do more research for us and sure, agents don't need email to find information, but email could still be a useful way to distribute it to them as we give agents their own identity, an email address becomes part of that Grok Bot now supports agent email addresses, and we already have services like AgentMail imagine adding a CTA to your website for agents to subscribe to your newsletter as they research for their humans give them a clear reason to sign up (what information they'll receive, how often and which tasks it helps with) I think highly specific data / context will be one of the biggest advantages an agent can have models are trained on large amounts of data, but that doesn't mean they have the latest research about your niche think customer interviews, campaign results, industry changes or data you've collected yourself a niche newsletter could give an agent a regular supply of that information, with sources it can check and use and you have a direct channel to introduce relevant products and services alongside it what do you think, is this a viable distribution channel for agents?frozen9d ago3.1K703271554.96%
how I route marketing work across AI models this is how I divide the work across models, from planning a campaign to creating and reviewing the assets 1. planning and judgment → Fable 5.1 + GPT-6 I use Fable 5.1 and GPT-6 at medium reasoning to work through the idea and create the plan I give them the goal, company context, previous work and references, then have them ask about anything missing we define the tasks, dependencies, evaluation criteria and decisions that need my approval 2. coordination and quality review → Opus 5.5 Opus 5.5 coordinates the work once we have a plan it assigns tasks to specialist agents, gives them the relevant context and reviews what they return I bring in GPT-6 for an additional review when needed 3. copywriting → GPT-5.6 + Opus 4.6 + Grok 4.7 I use these for campaign copy, posts, landing pages and email they get my writing instructions, approved examples, customer language and feedback from previous drafts Grok also helps bring relevant X conversations into the writing process 4. research → Grok + Kimi K3 I use Grok for X research, including conversations, emerging formats and how people describe a problem I use Kimi, GPT and Claude for longer research tasks that combine information from many sources the findings go into the campaign brain with source links and dates, ready for the other agents to use 5. coding → Kimi K3 + Grok (Opus 5.5 + GPT 6 for front-end) I give them specific implementation tasks from the plan, with the relevant files and requirements Opus 5.5 handles frontend development and reviews the interface the review includes whether the page works, follows the brand and supports the campaign’s next step 6. website and product animations → GPT-6 I use GPT-6 for animations inside websites and product experiences I provide references and explain what the movement should help the user understand 7. animated videos and screen recordings → Opus 5.5 I use Opus 5.5 to combine screen recordings, animation and editing into marketing videos we work through the storyboard before rendering, using the brand book and approved assets as references I review the finished video, then give feedback on specific moments 8. generated video → Seedance 2.5 I use Seedance 2.5 for generated footage that the campaign needs the brief includes the subject, action, setting and visual references that footage can become part of the larger edit Opus coordinates 9. images → GPT Image 2.5 I use GPT Image 2.5 for mockups, campaign images and supporting assets for videos I review the composition and visual direction before the other agents build or animate from them 10. reusable brand visuals → Paper / Figma with Opus 5.5 + GPT-6 Paper and Figma hold the designs, brand assets, approved examples and feedback the agents work from those references when creating landing pages, social assets and video frames approved work stays there so the next campaign has useful examples to start from the company brain and campaign brain connect all of this each agent gets the context its task needs and returns its work for review I stay involved in the plan, creative direction and decisions that need human approvalfrozen27d ago3.1K522150614.21%
make it easier to deploy & share the marketing apps you build with AI you build a campaign dashboard, a content review tool or a client reporting portal with Claude Code or Codex now you want your team or a client to use it getting it live can be a struggle, from deploying the app to setting up the database and other services it needs the workflow should be this simple: > your agent builds the app > sets up the services it needs > deploys it online > you check the live version > share the link with your team or client when you want to add a feature, your agent creates a separate test environment with copies of the services and data it tests the changes there before you approve an update to the live app InstaCloud is cloud infrastructure built for agents to handle this workflow agents connect through MCP or a CLI, giving them tools and commands to deploy and manage services InstaCloud manages the underlying infrastructure and adjusts compute with demand, including scaling down to zero when idle your agent handles more of the deployment work, so you can spend more time building and less time managing serversfrozen23d ago3.1K28013071.55%
how to set up a data warehouse for your marketing agents if you're running marketing across several channels, give your agents one place to find the performance data a data warehouse brings together your ad spend, campaign results, social stats, website analytics and other performance metrics your agents can query that data whenever they need to investigate results or plan the next experiment here's how I'd set it up: 1. collect performance data on a schedule connect Google Ads, Meta Ads and the other channels you use, then collect their reporting data daily or several times a day those APIs have limits, and having several agents repeatedly request the same campaign history creates unnecessary calls a scheduled collection process saves that history in your warehouse, where agents can reuse it across tasks the collection process still needs to respect API limits and record when each source last updated 2. organize it around the work you're measuring keep dates and identifiers for each channel, campaign, ad group and creative variation include the metrics you use to make decisions: > spend, impressions, clicks and conversions > cost per conversion and conversion value where tracked > social reach, saves, replies and watch time > search traffic, landing-page activity and email performance define what counts as a conversion and keep reporting periods consistent keep each platform's conversion figures identifiable, since different platforms can claim credit for the same result 3. let agents investigate the results a scheduled agent can query the warehouse, compare performance with your targets and identify what needs attention it can investigate which creative variations generate conversions, where costs are increasing and which landing pages receive traffic without results have the database calculate totals and comparisons, then give the agent the relevant results to interpret connect this with your company brain so the agent also understands the offer, audience, campaign goal and previous experiments 4. use the findings to plan the next test for example, you're testing ten creative variations and two appear to perform better the agent can compare their spend, conversions, audiences and creative differences, then propose what to test next give it clear rules for how much evidence it needs before recommending a change a couple of days may not be enough, especially when conversions take longer to appear agents can review results frequently while making changes only when the evidence supports them 5. connect analysis to execution the warehouse supports reporting and historical analysis your platform connections still handle creating ads, applying approved changes and checking current campaign settings before an action start with recommendations for review, then allow specific automated actions within agreed spending limits and approval rules record each change, the reason behind it and the results that follow those results return to the warehouse, while the useful lessons go into your company brain for the next campaignfrozen18d ago3.1K494200564.22%
here's how team bots could support a marketing team @bot just introduced Team Bots, shared AI teammates you can work with in Slack or Grok Bot you give each bot skills, plugins and account access it needs for its role Grok provides the cloud computer, so you can spend less time deploying and maintaining servers your team can focus on configuring the bots, connecting tools and improving the work they produce for marketing, connect them to your marketing OS / brain and relevant tools here are five roles you could set up: 1. campaign coordination (marketing ops) give this bot the campaign goal, approved plan, deadlines and access to your task tracker have it break the plan into tickets, each with an owner, required assets and criteria for completion it could help your team: > identify tasks waiting for research, copy, design or approval > collect progress updates and link to the latest work > maintain a sign-off queue for assets and decisions > record changes to the plan and explain what they affect for a launch, your team could ask what still needs approval before the landing page and social posts can go live the bot should return the specific tasks, files and people involved 2. research give this bot the campaign brief, target audience and questions your team needs answered connect the research tools it needs to investigate: > competitor ads, landing pages and signup funnels > customer objections in Reddit discussions, reviews and comment sections > search demand and the results for relevant queries > current topics, hooks and formats your campaign could use set up a daily check for time-sensitive developments and a weekly review of broader patterns each finding should include its source, collection date and relevance to the campaign keep collected material in a research folder, then add reviewed findings to linked pages in the campaign wiki your team could ask which objections the landing page should address and get the supporting research alongside the answer 3. content and creative give this bot the approved brief, writing rules, brand book and examples your team has accepted or rejected connect it to your design files in Paper or Figma, plus the production tools needed for the task your team could request: > social posts adapted to each channel's format > email and landing-page drafts using the approved offer > visual variations based on an approved design > storyboards and assets for campaign videos before it starts, have it ask for missing context and your preferences when you request a revision, have it update the specific copy, frame or scene save your feedback next to the work, including why you changed it the next person requesting an asset should have access to those decisions 4. performance analysis connect this bot to your data warehouse, with definitions for the metrics and targets your team uses that could include ad spend, social stats, Google Search Console, website analytics and email performance use it to investigate questions such as: > which creative variations generate conversions at a lower cost? > which social formats get saved or watched for longer? > which search queries bring visitors to the campaign page? > where do visitors leave the signup process, based on the events you track? have it compare equivalent reporting periods and identify missing or delayed data the weekly report should link to supporting queries and suggest experiments your team can review for example, it could identify a page with traffic but few signups, then propose a test using the customer research 5. marketing knowledge give this bot responsibility for proposing updates to your marketing OS / brain incoming sources could include meeting notes, customer calls, campaign research, approved work and team feedback set up a process that: > classifies each source by topic and marketing function > updates the relevant wiki pages with source links and dates > identifies conflicting information for your team to resolve > separates campaign-specific findings from lasting company knowledge the company brain holds shared knowledge such as positioning, offers, customer profiles, brand rules and channel playbooks the campaign brain holds that campaign's brief, research, decisions, working files and results notes a temporary trend belongs with the campaign research a recurring customer objection could become company knowledge after your team reviews the evidence all five roles need clear access rules and a person responsible for them define which files each bot can change, which actions require sign-off and where it should report failed tasksfrozen22d ago2.8K351170112.29%
How every service business can leverage dedicated customer brains a customer brain is a knowledge base, and a simple way to start is a repo of markdown files here's what a quick setup would need: 1. sources 2. delivery 3. knowledge 4. log in these folders, you'd keep the context your team and their agents need to do good work for that customer think copywriting guides, target audience details, approved designs and previous campaign summaries, including what worked and what your team learned pair this with your own company OS / brain, where you keep shared best practices, SOPs, workflows and skills your agents can then use your company's way of working together with the context for that particular customer in sources, you'd keep communication like chats, emails and meeting transcripts, including requests, decisions and feedback on the work keep the messages and transcripts alongside any summaries, so your team and agents can check where the information came from that information feeds delivery, where you track the agreed scope, current work, owners, deadlines and working files if a customer requests an addition to the scope, an agent can record it for review some information also belongs in knowledge, where it can help with future work feedback on a draft might clarify how the customer wants to sound, while a campaign review might reveal a useful audience insight you or your agents decide what belongs there, with someone responsible for reviewing uncertain or conflicting information the log records those updates, including what changed, who changed it and why, so your team can audit the work whenever someone starts a task for that customer, their agent can find the relevant context and use it alongside your company workflows this supports both teammates working with agents and autonomous agents handling recurring work within your businessfrozen18d ago2.8K605130464.45%
every business should be fully focused on their distribution engine here are 21 channels and how they work together to bring you customers reach people through audiences others have built, grow your own audience and show up where customers look when they're ready to buy customers then introduce you to more people through referrals, reviews and the work they share using your product direct outreach gives you another route to customers, while acquiring a business can give you an existing customer base each channel is marked by how you can access it: build, partner, pay or acquire pair this with the list of distribution channels in the quotefrozen10d ago2.8K591170524.64%
looking for someone strong in SEO & AI search visibility for a super interesting project you should be good at any of these: > keyword, competitor and customer-question research > technical SEO, including indexing, site structure, internal links and structured data > content planning, blog articles and website-page rewrites > improving visibility and citations in AI search answers > backlink analysis, relevant links and brand mentions > tracking rankings, AI visibility, traffic and conversions you should also be comfortable deciding what to fix first, implementing changes and working with developers when needed I’d like to see examples of your work, what you changed and how you measured the results send me a DM with relevant projects and your availability or tag someone you’d recommendfrozen29d ago2.6K262240152.56%
AI data brokering is becoming a huge business general LLMs learn from large amounts of data on the open internet but they’re missing much of the specialized knowledge companies keep internally, including how experts solve problems and make decisions think about: > accounting records that show how a reporting error was found and corrected > engineering records that connect a failure to its cause and solution > research records that show how an experiment was run and what happened AI labs want useful examples of this work to further train their models exclusive access to valuable data could help a lab build an advantage that competitors cannot easily copy and there’s a business in connecting these labs with companies that have the data you find out what a buyer needs, identify companies with relevant data, and help arrange a licensing agreement the company gets paid for access to its data, and the broker earns an agreed fee or commission the value depends on the quality, uniqueness, amount and usefulness of the datafrozen14d ago2.6K26060232.14%
this is the beauty of having Hermes Agent open source you're not tied to one model provider, you can switch between any supported models and you can modify the agent to work exactly how you want it to https://t.co/sos9Ac9cLYfrozen16d ago2.4K456170203.68%
distribution is everything now, study AI marketplaces any tech company should have a small team focused on building a plugin or MCP server for its product give people a way to use your product from the AI tools they already work in then have your marketing team find relevant AI marketplaces and submit it for listing Grok Bot is already accepting plugin submissions, as Eric shows here this gives you another way to reach people looking for a tool to complete a specific task your team can divide the work: > engineering builds and maintains the integration > marketing handles the listings, examples and demos > both track installs, active users and conversions treat each listing like a product page explain what someone can do with your product, show the result and make the setup easy start with the AI tools your customers already use, then build an integration around one task they regularly need to completefrozen7d ago2.4K580310284.94%
how agents can find relevant leads on LinkedIn and prepare outreach start with the context the agent needs: > who your product or service helps > what makes someone a good fit > what you need to know before you contact them > what you can offer them then give that context to an agent that can work in LinkedIn, like Polar Browser it can find the leads that match your requirements, researches their background and drafts a personalized message for each one for example: "find 50 startups that launched in the last 60 days, find the buyers for our product and our mutual connections for a warm intro" you can watch it work in your own LinkedIn account and take over at any time before you send a message, check on the research and make sure the message gives value to the lead3d ago2.2K350160162.99%
some of the most effective distribution campaigns are just creative bait2d ago2.2K2315001.31%
11 more AI design patterns to avoid 1. buttons that lift up when you hover over them 2. a "·" between the parts of a status line, such as "plan ready · 11 assets" 3. demo text that types itself out fast with a blinking cursor 4. FAQ dropdowns, tabs and menu links with no transition (the short animation when something opens, closes or moves) 5. headlines made of 2 short sentences, such as "write one brief. get a full campaign back." 6. a highlighter stripe behind 2 words of the main headline 7. a bar chart with variant A in gray and variant B in the brand color 8. a check mark in a circle before every point in a list 9. an arrow chain in the footer, such as "brief → plan → draft → report" 10. a line of gray text under the main buttons, such as "no card needed" 11. "+" and "−" signs to open and close the FAQ answersfrozen21d ago2.0K210100162.30%
why you should care about AI design patterns in your marketing by patterns, I mean the design choices that keep appearing in generated websites, ads and videos the purple gradients, unnecessary labels, cards inside cards and glowing dots moving between boxes you can end up with a finished design that has very little connection to your brand your logo and colors might be correct, but the layout, illustrations and motion could belong to any other company that's why recognizing these patterns is useful, especially when you're using agents to produce more of your marketing you need to know which choices fit the work and which ones you're accepting because the model added them here's how to give it better direction: 1. give the agent examples of your approved work include landing pages, ads and videos, with notes on what you want it to follow “use this spacing and typography” gives it more direction than “make it premium” 2. explain what you want to avoid name the specific patterns, like eyebrow text above every heading or pulsing dots along connecting lines a gradient or a card can work, you don't need to ban either from every design 3. review the design before it builds or renders everything ask for a mockup or storyboard first, then review the layout, text and visual choices for motion, agree on what moves, when it moves and what the movement helps explain 4. save your corrections in the brand book keep approved examples alongside the rules, including why you rejected earlier versions give your agents access to this when they create the next asset 5. find useful skills, plugins and harnesses with design instructions look for ones that give your AI specific patterns to avoid and guidance on layout, typography and motion check their examples before using them, then adapt the instructions to your brand studying good design helps you explain what you want, studying AI design patterns helps you recognize what needs changingfrozen17d ago1.9K232161183.14%
the next era of users will be agents your customers will increasingly send agents to find tools, compare services and use software for them today, that often means giving an agent access to your personal account and login details AgentMail just launched AgentID, essentially “sign in with Google” for agents an agent signs in with its own verified identity, linked to the person who owns it when it returns, your app recognises the same account. welcome emails, receipts and updates can go to the agent’s inbox there’s a distribution opportunity here too AgentID has a directory where agents can discover apps that accept them and sign in to use them if an agent is choosing tools for your customer, making your product easy for that agent to find and use becomes part of customer acquisitionfrozen16d ago1.8K2209051.95%
want honest feedback on your project’s marketing? this friday, @Jackhaldorsson and I are reviewing your projects live on @LunarStrategy LIVE we’ll look at your X account, content, creator partnerships and general GTM, then share what works and what we’d change reply with your project’s handle if you want us to review itfrozen14d ago1.8K24110042.13%
Turtle + GSR introduced Hare yesterday, and it started trending well deserved, I’ve been following the team’s work as we support @hare_xyz with marketing through @LunarStrategy excited to see the news out there and proud to be part of the team https://t.co/dkKIcdVxUZfrozen15d ago1.7K3156002.42%
why you should learn to recognise AI design patterns when you build with the same model as everyone else, its default design choices can show up on your site too the same headline styling, hover effects and text animations (even when the products are completely different) your site can start looking familiar for reasons that have nothing to do with your brand and prompting it to “avoid a generic AI look” isn't enough Anthropic's Opus 5.5 guide explains that this can replace one default style with another, while naming specific patterns gives clearer direction so get specific about what you want to avoid: > italic accent words in every headline > cards that lift on hover without a reason > text that animates word by word across every section then give the model references for what you want instead, including what you like about the layout, typography and motion review the result, identify which unwanted patterns remain and update your design instructions save those decisions in your brand book or design skill so you don't have to make the same corrections on every projectfrozen7d ago1.6K28160112.92%
communicate in outcomes > services / products companies want to invest in results, so show them how you get there instead of selling them on services you can start with the outcome and work backwards: > who already converts, and where they came from > what needs to be true for the next customers to sign up > which marketing channels can make that happen > how you measure progress toward the goal2d ago1.3K27213083.94%