Hashtags aren’t dead, they’re just ‘on life support’ as AI reshapes discovery

Hashtags aren’t dead. But they’re no longer doing the job they were built for. Once the backbone of content discovery, hashtags have been quietly pushed to the margins as platforms evolve into systems that understand content without needing to be told what it is.

What’s replacing them isn’t another tactic, it’s a move away from how content is interpreted, distributed and discovered.

“Hashtags aren't dead, but they're on life support,” Marcus Willis, CEO of Kill Boring Dead, says. “Most marketers are still treating them like it's 2018. The platforms have moved on. The algorithms have moved on. The audience has moved on.”

SEE MORE: TikTok Shop is booming globally, just not in Australia

Across TikTok and Instagram in particular, discovery is no longer driven by what a post is labelled as. It’s driven by what it is, how people engage with it and whether it holds attention. That’s a very different game.

The end of manual labelling

For years, hashtags were a crude but effective way of telling platforms what content was about. They were a form of manual categorisation - publisher-supplied metadata designed to help algorithms match content with audiences. That layer is no longer required.

“Historically, they were the primary way a publisher told an algorithm how to categorise a piece of content,” Luke Gosha, head of search and AI strategy at Edge Marketing, says.

“Today, the algorithms have advanced in their ability to understand content, its relevance to audiences and engagement.”

The shift is from explicit signals to inferred understanding. Platforms now analyse everything: what’s said in a video, what appears on screen, how long people watch, whether they rewatch, whether they share, save or comment. Content is no longer indexed - it’s interpreted.

“AI can now read your content the way a human does,” Willis said. “It understands what the video is about, what tone it's taking, who's likely to care.”

That has effectively removed the need for hashtags to act as a discovery mechanism. “You don’t need to tag it, you need to make it good,” he added.

Context, not content, is now king

If the last decade of digital marketing was built on the idea that content is king, the current one is tilting toward something more nuanced. Context.

“A piece of content that taps into something people actually care about right now will outperform a perfectly hashtagged post every single time,” Willis said.

Bryce Coombe, managing director at Hypetap, goes further. “Context and even subcontext are king. For a brand to succeed today, it’s less about having the right campaign tag and more about delivering content the algorithm identifies as high engagement.”

That shift is subtle but important. Content still matters, but it’s no longer enough on its own. Timing, cultural relevance, tone and audience fit now carry more weight than structured tagging ever could.

“The real shift is from structured tagging to behavioural and contextual signals,” says Miki Sim, platforms and culture director at VaynerMedia APAC.

“The algorithm no longer requires creators to manually label the content. It already knows what the content is about.”

Performance isn’t tied to hashtags anymore

One of the more telling aspects of the shift is how little impact hashtags now have on performance.

“If hashtags disappeared tomorrow, would performance actually change? For most brands, barely,” Willis said.

That’s a confronting reality for teams that still spend time debating which hashtags to include on every post.

“The content that performs does so because of what it is, the hook, the context, the emotional signal, the watch time, not because someone appended a hashtag at the end,” he said.

Coombe agrees the impact is minimal. In creator-led environments, he said, performance is now driven almost entirely by how the platform interprets the content itself.

“Content is now assessed automatically via image recognition, speech to text, caption search and a range of other signals,” he said.

Hashtags, in that context, have become a “low-level” signal - present, but not particularly influential.

Habit over strategy

Despite that, hashtags haven’t disappeared from marketing playbooks. They’ve lingered. Partly because they still serve a purpose in specific scenarios be that campaign aggregation, branded communities or live events, but largely because teams haven’t fully adjusted to how discovery now works.

“Most businesses still use them out of habit,” Coombe said. “But it’s no longer the engine that’s going to get you onto the ‘For You’ page.”

That distinction matters. Hashtags haven’t vanished, but their role has shifted from discovery tool to organisational layer.

“They’re more like community anchors than growth drivers,” is how one agency framed it.

It’s a subtle demotion, but a meaningful one.

B2B matters, sort of...

If hashtags have been sidelined on consumer platforms, LinkedIn remains a partial exception.

“LinkedIn is probably the last bastion for the hashtag,” Coombe said, pointing to the platform’s reliance on professional terminology and industry-specific categorisation. Even there, though, the direction is changing.

“Achieving organic reach on LinkedIn has become much harder,” Gosha says. “The signals that matter most are dwell time and the quality of the conversation in the comments.”

In other words, even in B2B environments, performance is increasingly tied to substance rather than structure.

“You can’t hack that with a string of B2B hashtags,” he said.

From SEO to AI-led discovery

The shift away from hashtags is part of a broader change in how discovery itself works.

Search is no longer confined to search engines. Discovery is happening across feeds, platforms and increasingly through AI interfaces that don’t rely on keywords in the traditional sense.

“We’re living in a world where AI can infer the meaning, tone and intent of a piece of content just by watching it,” Willis adds. That has implications beyond social.

“The next frontier takes this even further - brand discovery is increasingly happening directly on conversational AI,” Sim said.

When users ask AI tools for recommendations, hashtags don’t enter the equation at all. Content is surfaced based on how well it can be understood, contextualised and matched to intent.

“No hashtag helps you there,” Sim said.

That introduces a new layer to content strategy, one that moves beyond platform optimisation and into how brands are represented, interpreted and surfaced by AI systems. So what is replacing the hashtag? 

If hashtags are no longer the organising principle of discovery, what takes their place is less tangible but more powerful. Increasingly that is behaviour, context and relevance. These are not not added at the end of a post, they are embedded within it.

Gosha points to personalised activity, engagement patterns and dwell time as the dominant forces shaping visibility. Coombe highlights visual cues, natural language and what he describes as “the vibe” of content. And Willis strips it back further: “Make remarkable content, know your audience and trust the machine to connect the dots.”

For something that once sat at the centre of digital strategy, the decline of hashtags has been relatively quiet.

There hasn’t been a clear moment where they stopped working. No formal announcement. No switch flipped. They’ve simply been overtaken. Overtaken by systems that are better at understanding content than the people publishing it. Overtaken by algorithms designed to prioritise attention, not labels. Overtaken by AI that doesn’t need a prompt to know what something means.

Hashtags aren’t gone. But they’re no longer the shortcut they once were.