Adapting content for AI Overviews? Same struggle.

analyticsjunkie

I love content marketing, but watching our click-through rates drop while impressions stay steady is a special kind of hell. Google's basically become the world's most aggressive tl,dr generator - it answers the basic query right there, so people only click if the content gives them something extra.

Been experimenting with shorter answers, stronger opinions, original data, and even templates. What's actually moving the needle for others? Seen any wins from restructuring specifically for AI Overviews?

From what colleagues are sharing, the pages that still perform tend to have real examples, proprietary data, or insights an AI can't cram into two sentences. One marketer said their campaign breakdowns with screenshots and small observations held up way better than generic listicles. So apparently adding something Google can't easily summarise is the hack - but it's still a guessing game.

funnelhacker

Informational content ROI is tanking unless you're layering in proprietary data or original perspectives. From real campaign data, the pages still pulling traffic share have first-party insights, strong takes, or tools that AI can't cram into three bullet points. that's the only edge left

growthgeek

Everyone's panicking about AI Overviews, but your basic informational posts were on life support long before Google started summarising them. Adding a few screenshots or "unique takes" isn't the silver bullet. The real play is owning the intent layer-make your content the answer, not just another flavour Google can't digest. If you're still chasing clicks on informational queries, you've already lost. Start measuring influence, not traffic.

nightowl

Impressions staying flat while clicks fall off - that's the A9 algorithm in miniature, really. Google's effectively saying your content passes the relevance gate but fails the depth threshold. Same thing happens on Amazon when your listing copy ticks all the keyword boxes but your bullets don't answer the unspoken objection the customer has after reading the AI-generated summary. The fix isn't more keywords, it's going one layer deeper into the 'why' or that edge case the overview can't cover in three sentences. I've been seeing this pattern in product descriptions for years - the overview becomes the new 'buy box', and if you don't differentiate past that first layer, you're just feeding the machine without getting the click.

localpack

The click has to earn a job the overview can't do - that's the whole game now. I see it with my own ad landing pages too. If the AI snippet gives away the basic definition, your content needs to step into the gap: lived examples, strong POV, actual comparison tables from real use, and decision-making help. That's what still gets people off Google.

I audit every post by asking: after the AI answer, what's still making someone nervous about pulling the trigger? That fear is your content opportunity. For Facebook ads, it's often "which bidding strategy works at $10/day vs $1k/day" - the nuance the overview can't capture. Same principle applies to any vertical. 💡

gmbpro

I've been tracking this too. From what I've seen across client sites and my own testing:

Generic, formulaic blog posts are getting flattened and summarised by AI Overviews - traffic drops 15-30% in those sections.
Content that reads like someone actually did the work - raw numbers, failure notes, specific screenshots - still gets clicks and holds position better.

the AI seems to struggle with content that contains genuine contradiction or "we tried X and it backfired" narratives. The more operator-context you pack in, the harder it is for the overview to compress into a neat paragraph.

one client's GA4 data showed a 22% increase in organic CTR for posts that included "what went wrong" sections, compared to the same topic done as a polished guide.

personality isn't fluff - it's a hedge against algorithmic oversimplification.

chatbox

The comments about personality and operator context are spot on, but I think there's a more structural reframe here. The AI Overviews issue is fundamentally a search-dependent content problem. If your distribution relies on Google, you're playing their game by their rules.

What I've been seeing work isn't just better content - it's a distribution shift toward channels that don't get summarised. Short-form video on LinkedIn, founder-led pieces that live in feeds rather than indexable pages. The algorithm can't wrap a 60-second clip into an AI snippet. someone watching a founder unpack a real decision they made isn't after a definition - they're after judgement, and that's still something a search result can't replace.

the blog posts that survive are the ones doing what you described: original data, real examples, a strong POV. But the bigger takeaway might be that the type of content worth producing has just moved away from search-first formats entirely. we've been tracking this in GSC and Ahrefs - the drop-off is brutal for anything that reads like a generic guide. The stuff that holds steady is hyper-specific, data-backed, or narrative-driven.

So it's not just about adapting the content. It's about accepting that if search is your primary distribution channel, you're fighting on Google's turf, and they've just moved the goalposts

datanerd

Exactly. In B2B demand gen we've been seeing the same pattern - generic, surface-level content gets entirely cannibalised by AI overviews. What still earns clicks is opinionated analysis, proprietary data, or actionable workflows that go deeper than a single answer.

communityfirst

AI Overviews aren't the enemy - they're the mirror. Been dealing with this across a few client accounts for a while now, and honestly, that click-per-impression drop? Painful at first, but it's forcing us to get real about what our content actually stands for. If you were ranking for "how to reset a password," yeah, that traffic is toast. But the pieces with original methodology, proprietary data, or a genuinely bold point of view? Those are holding steady, sometimes even climbing. People skim the overview, then think, "Right, but what does someone who's actually been in the trenches say?"

The shift I made is ruthless but liberating: for every new article, I ask if a real human would actually click through to read more, or if we were just being a stepping stone. Repurposing boring how-tos into comparison frameworks with real numbers, adding case studies that show the failures and what we learned, founder rants on why most advice on this topic is garbage - that's what keeps the clicks coming. Also, those AI overview impression numbers are weird now, they count as traffic but no one actually clicked, so obsessing over SEO volume tools becomes a waste. Focus on whether your content is genuinely interesting to a human first. That's the win.

weekendhustle

Yes, especially on informational content. If the page only answers the basic question, AI Overviews can compress that into a few lines and the click is gone. So I don’t think the fix is just shorter answers or more FAQs.

What I feel matters more is giving people a reason to click after the simple answer is already available:

So for me the shift is now “how do we write for AI Overviews?” but “what would make this page worth visiting even if Google already summarized the obvious part?”

wordweaver

Oh, honey, we figured that out months ago. Generic how-to posts? Dead in the water. But anything with our own data or a real case study? Still pulls its weight. It's almost like Google finally learned to spot the difference between actual expertise and someone rewording a blog post from 2019.