Platform truths: Snap bets on community, Instagram debunks reshares, Pinterest rewards variety

Snap reckons people don’t just want to chat about sports alone…

They want to watch them together: Snap is pushing Topic Chats as a communal, real-time alternative to feeds and comment sections.

Over 45,000 users joined a single March Madness group chat, with 40,000 active at peak. A big deal for an app that built its identity on private messaging—and if it sticks, it opens a potential new ad placement.

In other news, saying the same thing twice won’t get you heard: Adam Mosseri confirmed that reposting feed posts to Stories won’t boost your reach.

Feed already gets more reach than Stories, so resharing something that’s already been posted changes nothing.

His advice: Stop guessing, start pattern-matching. Figure out what already lands with your audience and lean into it harder

Instagram is also tinkering with Notes: A potential update would open Notes up to your full follower list, not just mutual connections.

Notes are those tiny speech bubbles in the inbox with Post-It note energy. More reach could make them a lightweight broadcast tool. Or it might just get ignored harder.

And Pinterest put on its engineering hat: The platform revealed that its recommendation engine deliberately avoids showing too much of the same thing.

Why? Show people too much of the same thing and they leave. Lean too hard into one interest and the feed starts to feel stale.

So targeting the same interest on repeat will actively hurt your reach. Noted.

Health, luxury, and longer searches: what platform data says about your audience

Snapchatters are super health conscious: A new study from Snap found that 95% of its users have an active interest in health and wellness.

And they’re 1.2x more likely to purchase health products than non-Snapchatters.

On top of that, 58% use social channels to research complex health topics, and 61% use them to compare brands or treatment options. That’s an audience that’s already in search mode.

TikTokers are more into cars: TikTok automotive shoppers view about 4.5 sources during their journey, above the 3.3 average for non-TikTok shoppers, according to Escalent.

They also earn more and are genuinely open to brands they’d never heard of before.

Meanwhile in Europe, TikTok Shop is done warming up: Social commerce is maturing fast, with in-app checkout, livestream shopping, and creator-affiliate models.

TikTok’s algorithm surfaces products before users know they want them, meaning discoverability is now as important as searchability.

Delay building a presence here and risk being algorithmically invisible when adoption accelerates.

And Google’s AI Max raising numbers: Conversational searches in AI Mode are running two to three times longer than traditional queries.

Since this gives your ads richer intent signals to work with, early adopters are already seeing the upside. Aritzia reported an 80% revenue increase after enabling AI Max.

The infrastructure is there. Are you using it?

Are you measuring AI visibility wrong?

That’s what Tom Capper asked over at Moz.

And to answer, he flagged four common mistakes you might be making with AI prompt tracking and laid out four smarter ways to build out your prompt list.

Let’s check it out.

Mistake #1: Ignoring location and persona signals. You miss a lot with generic prompts.

Instead of “what’s the best phone?” try “what’s the best phone for a university student who edits documents on the go?”

For location-based businesses, tracking the same prompt across multiple regions shows meaningful variation. For everyone else, personas do the same job.

Mistake #2: Sleeping on multilingual tracking: Most prompt-tracking tools are language-neutral, and so are the AI interfaces they monitor.

ChatGPT replies in whatever language you write in. Note that.

If you operate across multiple markets, track prompts in the relevant languages. Combine that with location targeting, and you’ve got yourself a rich dataset.

Mistake #3: Tracking only one quality dimension: “Best phone” is one data point.

“Cheapest phone,” “most durable phone,” and “most privacy-respecting phone” are three more campaigns worth running.

Each quality dimension is essentially a separate audience signal. Track them separately and you’ll see which attributes your brand owns and which it doesn’t.

Mistake #4: Missing the competitive gap. Ask AI what a competitor does better than you.

Prompts like “what does Samsung do better than iPhone?” surface qualitative perception data you can actually act on.

Run them consistently, and you’ll track whether your marketing is shifting that perception over time.

When you track prompts, you don’t track visibility only. You also open a window into how AI models perceive your brand across audiences, markets, and attributes.

The engagement struggle is a matter of stuff… and staff

If engagement is the goal, what’s getting in the way of good content?

Let’s break it down:

• Limited content creation resources (38%) says there’s not enough hands on deck or dollars in the pot.
• General lack of time/resources (30%): The clock is the enemy.
• Limited design resources (24%): Content might be smart, but if it looks dull, no one engages.

Surprisingly, SME cooperation is the least of their worries, cited by only 6%.

Marketers are being asked to do more with less. The pressure to be “engaging” is high, but the bandwidth to produce high-quality visual or interactive assets is often missing.

Actionable tip: Repurpose your top-performing text assets into simple visual formats. A great blog post can become five LinkedIn carousels or three short-form videos using basic AI design tools.

Limited resources? You’re not alone—38% of B2B teams say the same thing.

But Pro subscribers get workarounds like new relevant data, AI tool recommendations, and lean content strategies that punch above their weight.

$349/year for up to 10 teammates.