Subscribe
Share
Share
Embed
.
The Social Trend Report. Weekly Tuesday deep-dive on content, social media strategy, and AI for operators: platform algorithm shifts, creator economy, format research, posting cadence, and what's actually working this week. Hosted by giovanni gallucci with Rachel Donovan. Three operator stories per episode, under ten minutes.
.
.
.
| ai-assisted content
[HOST] Good morning. This is giovanni gallucci, with Rachel Donovan, and this is the Social Trend Report.
[GUEST:excited] Morning, gio. Coffee's hot, tank's full, keys already in my hand, so let's not sit still.
[HOST] Three this week. Why X just handed you the receipt for every time your reach fell off a cliff. Why the answer box is eating your click, and what to do instead of panicking about it. And why five small parts will beat one big post every single time. Let's get into it.
[STINGER]
[SECTION 1: Pull The File Before You Blame The Room]
[GUEST] Clark put this one at the top of the stack. On September twenty-first, X expanded a page called Under the Hood. It shipped back in August, and it lets you download a report of every safety label sitting on your account. The new piece names when a country asked X to restrict one of your posts, and which posts drew it.
[HOST] And for a small operator, that file is a bigger deal than it sounds.
[GUEST] Bigger how.
[HOST] Because it kills the guessing. Here's the pattern every business account knows. Reach is fine for eight months, then one week it drops by two thirds and just stays there. So you start guessing. You change your posting times. You cut the links because somebody on a podcast swore links get buried. You add them back. You post more, then you post less.
[GUEST:thoughtful] Six weeks of that and you still don't know what happened.
[HOST] Six weeks gone, real hours burned, and you have no idea whether you did something, a competitor mass-reported you, or nothing happened at all. Under the Hood replaces every one of those guesses with a document. You either have a label holding your posts down in the feed or you do not.
[GUEST] So what does a food brand actually do with it.
[HOST] First thing, go get it. You need an account older than a year and at least ten posts in the last month, and then it's just a download. Read every label on the account and on the individual posts.
[GUEST] And then.
[HOST] Then the part people skip. You line the label dates up against your own analytics. If there's a label and it lands the same week your reach fell, congratulations, you found your cause. But if the report comes back clean and your reach still dropped, that is also an answer, and it's the more important one.
[GUEST:curious] Why is the clean report the more important answer.
[HOST] Because it stops you from burning down a strategy that was working. A clean file means the cause sits upstream of any penalty. Maybe it was a slow season. Maybe the format just ran its course. If you rip up your whole plan chasing a punishment that never happened, you did more damage than the algorithm ever did. I've watched brands torch a format that was perfectly fine because their category just goes quiet in July, and this file would have told them that in about thirty seconds.
[GUEST] Where does this go sideways.
[HOST:annoyed] It goes sideways when you treat the file as the whole truth. Transparency published by the same company doing the penalizing is still a document that company controls. It shows the labels X decides to show you. It does not show the ranking weights, it does not show the model that applied the label, and it does not hand you an appeal that resolves on a schedule you could plan a launch around.
[GUEST] So it's one input, not the answer key.
[HOST] Right. Read it that way and it saves you weeks. Read it as gospel and you'll just build a new superstition on top of the old one.
[GUEST] Three to run with.
[HOST] One. Pull your Under the Hood file this week. Settings, download the report, read every label on the account and on your posts. That's your starting point instead of a guess.
[HOST] Two. Line the label dates up against the week your reach fell. If they match, you found it. If the file is clean, stop blaming the platform and go look at your calendar and your formats.
[HOST] Three. Fix what the file names, then get back to posting. Do not sit there refreshing it like a weather report. It's a receipt, not a strategy.
[STINGER]
[SECTION 2: The Answer Box Ate Your Click]
[GUEST] Travis ran the numbers on this one, and they are ugly. Researchers at Penn and Northeastern finally ran a real experiment on Google's AI answers. Not a survey, a randomized test, about a thousand people. When an AI Overview showed up at the top of the page, the clicks out to actual websites fell almost forty percent.
[HOST] And here's the part that makes it worse. The clicks on the paid results didn't budge.
[GUEST:surprised] So the free traffic evaporates and the ad slots stay put.
[HOST] Right. The organic traffic dries up and the paid slots don't move an inch. If you built your whole discovery plan on ranking a page and earning the click, that plan just lost about half its fuel, and nobody sent you a memo. Those same users also came away trusting the results less and searching less, when everybody going in assumed they'd search more.
[GUEST] Okay, so if the click is drying up, where does a brand even show up.
[HOST] That's the right question, and Kendall mapped the answer. A firm called Trendos went through a hundred and seven million AI answers and pulled the top sources those answers actually cite. For consumer goods it sorted into three buckets. Community stuff, so Reddit and YouTube and forums. Retail and brand pages, so Amazon and your own product detail. And independent write-ups.
[GUEST:curious] So the job stops being about ranking.
[HOST] The job becomes being the thing the model quotes. And that changes what's worth making. A vague brand paragraph never gets cited. The model reaches for the specific, checkable thing. The exact number, the real ingredient, the straight answer to how you're different from the can sitting next to you. And those three buckets tell you exactly where to spend. If community posts and product pages are what the engines quote for consumer goods, then a plain answer you leave in a forum and a product page with a real spec on it are worth more than another glossy post nobody ever cites.
[GUEST] And the places it pulls from.
[HOST] Look at that list again. YouTube. Community answers. Your own product pages. That is not some new discipline you have to go learn. That's the social content you were already supposed to be making, pointed at a new job.
[GUEST] Give me the version I can act on Monday.
[HOST] Monday morning, run the check on your own category. Open Perplexity, open Google's AI Overviews, whichever answer engines your buyers actually use, and ask them the three questions a real buyer asks before they hand you money. Write down every source those answers cite.
[GUEST:thoughtful] And that list is the target.
[HOST] That list is exactly where you need to show up. Then you go make the straight, specific answer live on the platforms doing the citing. For most food brands that's YouTube, your own pages, and the corners of the internet where your category argues with itself.
[GUEST] And the trap you keep warning me about.
[HOST:sarcastic] The panic take. Somebody reads that forty percent and declares search dead, throw it all out. That's lazy. The click was never the prize. The trust was. If the answer quotes your number and the buyer never clicks, you still won, because you're the brand the machine vouched for.
[GUEST:mischievous] One caveat on the Trendos map before we move.
[HOST] Good catch. Those figures come from the vendor's own tool, and nobody audited them. So don't take their list as your map. Run the check on your own category and build the real one.
[GUEST] Three to close.
[HOST] One. Run the citation check yourself this week. Ask the three engines the questions your buyer asks, and log every source they quote. That's your placement list, built in twenty minutes.
[HOST] Two. Put one specific, checkable fact where the models pull from. A number, a date, an ingredient, sitting on YouTube or your product page in plain words. Specific gets quoted. Brand copy gets skipped.
[HOST] Three. Stop scoring yourself on the click alone. Start asking whether you're the cited source, because that's the question that decides whether you exist in an answer nobody clicks out of.
[STINGER]
[SECTION 3: Five Parts Beat One Post]
[GUEST] Jayme flagged this last one, and I did not expect to care about it as much as I do. Snapchat and Ipsos surveyed three thousand people across three countries, and forty-two percent of Snapchat users are watching microdramas every single day. Little serialized vertical episodes. And the part that got me is the sharing. People don't watch them alone. They send clips to friends and argue about what happens next.
[HOST] Hold onto that last detail, because it's the whole reason this matters. And set the genre aside, because the genre is a trap.
[GUEST] The genre is a trap how.
[HOST] The lesson underneath the soap opera is that a series beats a one-off, and a small brand can run a five-part story when it could never afford a campaign. One post is a coin flip. You ship it, it catches or it dies in an afternoon, and the next day you start over cold.
[GUEST:curious] But a series is more work. Why is that worth it for a shop with one person posting.
[HOST] It's actually less work, and that surprises people. A series takes one good idea and spends it across five posts. It doesn't demand five separate ideas.
[GUEST] And coming up with five ideas is the thing that breaks you.
[HOST] Exactly, that's what breaks a one-person shop. Coming up with one idea and telling it in parts, that you can do on a Tuesday. And every part after the first shows up to an audience that already met you, so you're not paying the cold-start tax five times over.
[GUEST] So what does that look like for somebody who makes, say, cheese.
[HOST] You've got a batch that failed on a Wednesday and you fixed by Friday. There's your series. Part one, here's what went wrong and the tray I had to throw out. Part two, the theory on why. Part three, the change I made. Part four, the fix in the tank. Part five, the taste test.
[GUEST:thoughtful] And you number them.
[HOST] Put a number in the first frame of each, so the feed reads it as a run and not five loose posts. You didn't need actors. You needed a story that comes in parts, and you already had one sitting in the shop.
[GUEST] Now the sharing piece.
[HOST] The sharing piece is your distribution, and it's free. Those microdramas travel because people argue about them. So you build one thing worth arguing about into every part. A cliffhanger, an open question, a call somebody will push back on. And that daily number is the real tell. Forty-two percent watching every day is a habit, not a one-week spike, and a habit is the one thing a small brand can build toward without a budget behind it.
[GUEST] And when they send it to somebody.
[HOST] When a viewer sends part three to a friend to settle a bet, that friend is a new viewer you didn't pay for. The format does the reaching so you don't have to rent it.
[GUEST] Where do brands blow this.
[HOST:annoyed] They chase the costume instead of stealing the mechanic. They see the word microdrama, decide they need a script and a cast, figure that's not them, and walk away from the actual lesson. The mechanic is serialization plus a reason to come back. That is it. You are not making a television show.
[GUEST] Just one true story with parts.
[HOST] One story, told in parts, with a hook. And a wheel of cheese aging in a cave has more real drama than most of what's on there anyway.
[GUEST] Three to send us out.
[HOST] One. Take your single best one-off idea and cut it into a five-part run this month. Put the episode number in the first frame so the feed sees a series, not a scatter.
[HOST] Two. Build one hook into every part. A cliffhanger, an open question, something worth disagreeing with, so the audience sends it around and does your reach for you.
[HOST] Three. Steal the mechanic, skip the genre. No actors, no script. One real story that comes in parts. The batch that broke Wednesday and got fixed Friday is a series you already own.
[STINGER]
[OUTRO]
[HOST] That's the report. If you're a food or beverage brand and your reach fell off a cliff and you honestly can't tell whether it's a penalty, a slow season, or a plan that just quit working, that's the problem I solve at gallucciNET. Organic social media strategy. Finding the real audience, building the community, and growing your reach and your loyalty without buying a single ad. Fifteen-plus years running organic social for food and beverage brands. Find me at gallucci dot net, or DM at gallucciNET on Instagram or LinkedIn.
[GUEST] Talk next week.