this is bigger than you think... Google's AI Mode and the Gemini app each have 1B+ monthly users now when someone asks for a product like yours, your MCP can show up embedded in the answer here's how it works ↓ if you have an MCP, or a SaaS tool, app or online service that could work through one, you can start on this today: 1. apply for the preview at developers .google .com/connected-apps 2. connect a remote MCP server with OAuth 2.1 login, so it can act in the user's account (or use AppFunctions for an Android app) 3. register your tools and test the app in the Google Connected Console 4. choose Search, the Gemini app or both, then submit the app for review 5. once it's live, track traffic, latency and errors in the console, with your verified logo on the app 50+ apps are already live, from Adobe, Peloton, Experian and more, and Google opens access in phases over the coming months MCPs and plugins are becoming one of the biggest distribution channels, so throw yourself at this, it's HUGE

how to reverse-engineer any company's growth playbook (for under $1) Jason just open-sourced /REA-growth, a Claude Code skill that builds a growth playbook all public marketing channels here's how it works ↓ 1. you name a company and a time window 2. the agent pulls the public data through api: traffic, search, ads, backlinks, posts, hiring, news and AI search answers... 3. it finds every launch & important date (posts, ads, launches, raises, hires) against their monthly traffic, and what kind of results it yielded 4. you then get a playbook: what got them results and when, that you then can get insights and apply to your own campaigns he tested it on Lovable with over 2 years of data, and saw that their website affiliate campaign badge was one of their most successful ones which company would you run it on?

LIVE This Friday on Upblock: the AI business race and Anthropic's IPO Join @TimHaldorsson, @shannholmberg, and @sirsamjenks for the week's headlines and a deep dive into who is winning the AI enterprise race and what the IPO means for the rest of the space Live on X at 11 AM GMT+1 on Friday

TLDR about the new X algo update > follow your reply guys back, a reply from a mutual counts 4x a normal reply > write posts people want to share, when someone copies and shares your post's link it counts 30x a like > post when your followers are online, if you're under 50k followers a new post can get boosted in its first 2 hours, and X picks the posts with the most likes per view > write posts people stop and read, time spent on a post counts 50% more since Oct 13 > post fewer, better posts, if several of your posts are in someone's feed, the 2nd only gets about 60% of its normal score and the 3rd less than half

how to fix AI design slop by pointing at it (@cursor_ai's new Annotate mode) you click on the page, and the agent gets the exact element, style or sketch you mean: > comment on any element, it knows the exact class > change colors, padding and radius in a panel > sketch a new section right on the page > send every note to the agent at once (batch revision comments) > check the page at any screen size

the future way of operating SaaS will be to sell outcomes instead of letting users do the work themselves, you package your product in two ways 1. self-serve the user runs your product themselves, in your app or with their own agent connected to your API they own the result, and they pay for access and usage 2. outcome the user tells you the result they want, and you deliver it. behind the platform, a marketing engineer owns the work for that client and orchestrates the agents that produce the outcome. they pay a higher price, because someone is accountable for the result. for an SEO and AI search tool, it could look like this: self-serve > the user signs up, tests the keyword research and gets an API key > they give the key to their own agent with a task, like "get us ranking for these 10 keywords" > the agent pulls the data through your API and writes the content itself > priced by usage outcome > the user asks for a result, like "[N] articles in the top [X] for [topic] in [Y] days" > a marketing engineer on your side runs your platform and agents to deliver it > priced per result, at a premium over the platform alone software gives the user the tools, and a marketing engineer with agents gives them the outcome (with the experienced from succesful outcomes prio)

how X ranks your posts (breakdown of the new algo update) the X algo predicts how each person will react to your post, and that then decides if and where it shows up in their feed here's how much each reaction counts, compared with a like: > someone copies the link: 30x > someone you follow back replies: 30x > a reply, a quote or a DM share: 7.6x > a follow from the post: 6x > a repost: 1.5x > a like: 1x > a bookmark: 0 the negative actions weigh far more, a report is -355x a like, a mute -89x, "not interested" -79x and a block -47x if X predicts someone is likely to mute or report your post, it ranks lower in their feed reports get a big weight because they are rare, a report is over 1,000x less likely than a like these rules decide how far a post can reach: > if you have 50,000 followers or fewer, X can boost a new original post up to about 16th place in someone's feed > the boost only works in the post's first 2 hours, and stops once the post has 200 views in the Home feed > X boosts one post each time a feed loads, and it prefers posts that got the most likes per view from their first viewers > after 48 hours, X stops putting your post in people's feeds > your replies and reposts only show up in the feeds of people who follow you > original posts can also reach people who don't follow you, X scores them 25% lower for those people > when several of your posts compete for one person's feed, the 2nd keeps 62.5% of its score and the 3rd keeps 43.75% > replies, quotes, DM shares and copied links only count when people find your post in their feed, so engagement from a group chat link adds nothing > if X labels your post low quality, it shows a notice on it and people can't like, reply, repost, bookmark or share it, misinformation labels can limit it too what I'd do with it (my read of the weights): > post when you can reply in the first 2 hours > write posts people want to quote, reply to or send in a DM > reply back to accounts that follow you > post originals when you want to reach new people

the future way of operating SaaS will be to sell outcomes instead of letting users do the work themselves, you package your product in two ways 1. self-serve the user runs your product themselves, in your app or with their own agent connected to your API they own the result, and they pay for access and usage 2. outcome the user tells you the result they want, and you deliver it. behind the platform, a marketing engineer owns the work for that client and orchestrates the agents that produce the outcome. they pay a higher price, because someone is accountable for the result. for an SEO and AI search tool, it could look like this: self-serve > the user signs up, tests the keyword research and gets an API key > they give the key to their own agent with a task, like "get us ranking for these 10 keywords" > the agent pulls the data through your API and writes the content itself > priced by usage outcome > the user asks for a result, like "[N] articles in the top [X] for [topic] in [Y] days" > a marketing engineer on your side runs your platform and agents to deliver it > priced per result, at a premium over the platform alone software gives the user the tools, and a marketing engineer with agents gives them the outcome (with the experienced from succesful outcomes prio)

with Hermes Agent + Claude SEO, you can build an agent that runs your SEO and AI search it researches keywords, collects context and prepares article drafts that can rank in Google and get cited by AI chats how it works: 1. find searches that matter to your business start with questions people ask when choosing a product: > best [category] for [use case] > [competitor] alternatives > [tool A] vs [tool B] add relevant Search Console queries where your pages get impressions but sit outside the top 10 use DataForSEO to compare search demand, competition and trends prioritize questions your product can help answer, then decide whether to improve an existing page or create a new one 2. build a research file for each target give the agent: > current search results and competing pages > sources cited in relevant AI answers > product documentation and approved claims > customer questions and sales objections > original examples, results or expert input DataForSEO’s LLM Mentions API can help find citations in its database of ChatGPT and AI Overview responses ask the agent to identify what the existing pages miss and what useful information you can add 3. define the workflow and writing rules Use Opus 5.5 in Hermes and install Claude SEO Plugin for the bot set a schedule for research, draft preparation and performance checks save the research, sources, drafts and previous results in project files that each run reads give the writing agent: > your voice DNA file > anti slop rules > approved examples > a source for each factual claim > instructions to flag missing information answer each section’s question directly, then add the evidence and detail needed to understand it update pages when the information changes or you have something useful to add 4. review before publishing have the agent prepare a review package with the draft, sources and unresolved questions check the claims, product details and voice use Claude SEO to check technical issues, internal links and structured data where relevant keep publishing behind your approval 5. find relevant places to earn mentions look at the third-party pages AI answers cite for your target questions find relevant comparisons, directories and resource lists that include your competitors ask the agent to prepare an outreach shortlist with: > the page and its audience > why your product belongs there > evidence that supports the suggestion > a draft message for your review give the editor a useful reason to consider your product 6. publish, measure and improve check that published pages can be crawled and indexed track: > Google impressions, clicks and rankings > citations in the AI answers you monitor > visits and conversions from those channels > pages that need better information use Bing Webmaster Tools’ AI Performance report to track citations across supported Microsoft AI experiences have Hermes send a weekly report with what changed, what needs attention and the proposed next actions the system needs to be able to choose useful targets, collect reliable context, check the work and learn from the results

we're past the stage where you look at vanity metrics and decide which creators to include in your campaign successful creator campaigns are a balanced mix of coordination, qualified creators and a shareable story Creator Wire now makes it easy to coordinate and find qualified creators through the new Insights feature we've been working on bringing better insights to help teams make better decisions and give creators better data on how their content is performing PS you get a sharable video / card when you sign up ↓

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 goal

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 lead

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 process