just published the viral-carousels trendwatching skill you guys asked for it returns the carousel accounts by niche ranked by saves and like-rate instead of views, with the slides of the top posts. $11 on apify, everything's in the repo 👇

map your whole category's facebook ads before you make a single one pull every competitor's facebook ad library with the apify api have gemini describe every image and video. angles, promises, outcomes now you have a map of how everyone in the category talks about themselves have claude find the gaps nobody is claiming validate the gaps by scraping reddit with exa ai make ads for the gaps with nano banana and seed dance upload with the facebook ads api next day, read the results from your warehouse, kill losers, move winners to their own ad sets once it works, deploy it to a server so it runs itself

How would you steal customers off a competitor with outbound? Easiest campaign in B2B to run, because every account on the list has already proved they'll pay for what you sell. Here's how we'd build it with Clay, Apify, Trigify & EmailBison, after 10M+ cold emails: Step 1: Find every company paying them Nobody hands you this list, but every customer leaves a trail. > BuiltWith or Apollo's tech filter for their tool > G2 & Capterra reviewers, pulled with an Apify scraper > the logos on their case study page > job posts listing their tool as a required skill > people engaging with their LinkedIn posts, via Trigify Run the case study logos through AI Ark or Ocean as seed domains. You get every company that looks like their best customers. Then qualify every account with a Claygent column that reads the site & confirms the fit before you pay to enrich anyone. Step 2: Score them on how unhappy they are A happy customer isn't a lead. You want the ones already looking for the door. > a 1 to 3 star review in the last 12 months > hiring for a role their tool should've replaced > a new VP in the seat, auditing the whole stack > engaging with posts about switching Two of those on one account puts it at the top of the list. Step 3: Group them by complaint Pull every review & sort by what they're angry about. Slow support is one group. Surprise price rises are another. Each complaint becomes its own campaign with its own message, sent to every account that raised it. That's relevance without a single personalised line, because the list did the work. Step 4: Make switching cost them nothing The reason they haven't left is the work of moving, so the offer removes the work. > a free migration of their data > 30 days running both side by side > a comparison against their current setup, built for them Skip the audit & skip calling your product better. They already know what's broken, they told G2. Step 5: Write one email per complaint Keep it under 45 words & open on the complaint in their own words, lifted from the reviews. "Hey Sarah, saw a few {competitor} users mention reporting takes a day to refresh. Ours updates live & we'll move your data over for free. Worth a look?" Subject line stays at 2 words, lowercase: "{competitor} reporting" Spintax every sentence & launch 15 variants across the complaint groups. Step 6: Treat "we already have a vendor" as the best reply you'll get It confirms they buy the category, they have budget for it, and there's a contract somewhere with an end date. > reply once & ask when it renews > log the date & the incumbent against the account > re-contact 90 days before, with the comparison > warm call 30 days out, when they're deciding Step 7: Call anyone email can't reach Enterprise accounts sit behind Barracuda & Proofpoint. Pull the ESP before sending & route those straight to the phone. Call every positive reply the same day, while they're still annoyed enough to move. Step 8: Read it by complaint group One complaint will carry almost all the meetings. Move the volume there & retire the rest. Step 9: Keep the whole list on a 90 day rotation Contracts renew whenever they renew, so the list gets contacted every 90 days until they do. A marketing automation client of ours sells into a market where most buyers are locked into 36 month contracts with a competitor. Some of their prospects are 32 months from renewal. Their fastest deal closed in 4 to 6 weeks, because the email landed the month a competitor let that prospect down. Start with whichever competitor your prospects complain about most on G2.

AI agents are now buying and using products on behalf of humans, and the natural evolution from PLG is ALG (Agent-Led Growth) There are only a handful of operators who have iterated on and designed agent-first growth funnels, and I'm so excited to bring some of them on stage at our very first ALG event hosted by @2027dev 👋 @JaySahnan, growth engineer @browserbase @lavanyaai, prev. VP growth @wandb @0xLukasB, product marketing @apify Come join us! October 7th, 6pm If you're working on ALG and have a demo/insights to share, pls comment here + we'll try to include you too!

Jev is incredibly good! If you haven't heard, Jev is a new "System One" model optimized for decision-making. For example, you can give it a set of possible choices, and Jev will classify the input text based on those choices, including a confidence score for each. Best of all: Jev is really fast and cheap, so you can use it in all sorts of applications. The first thing I built with it is a classifier to pick the hotels I want to visit. I use @apify to collect thousands of reviews, then classify them with Jev based on specific criteria I care about. You just couldn't do this reliably, but now you can. Link to the repository: github.com/svpino/hotel-revi…