Yes, but the job description changed. A blog used to exist to win clicks from Google. It now exists to do two things: be the source AI systems read when they answer your category's questions, and capture the smaller number of higher-intent clicks that survive.

Be straight about the click decline first. Google referrals to publishers fell 34 percent year over year, zero-click searches grew from 56 to 69 percent after AI Overviews rolled out, and organic CTR on queries with an AI Overview still runs well below queries without one. If your blog strategy is "publish generic posts, collect traffic", that strategy is over.

Now the other side of the ledger, which is where the case for blogging gets stronger, not weaker.

The clicks that remain are worth more. Semrush measured AI-referred visitors converting at 4.4x standard organic, and Ahrefs found AI search delivering 0.5 percent of its traffic but 12.1 percent of its signups. On Shopify specifically, AI-referred sessions convert about 50 percent above organic with 14 percent higher order values. Fewer readers, better buyers.

Being cited pays even inside Google. Seer Interactive's 2026 study found brands cited in an AI Overview earn 120 percent more organic clicks per impression than uncited brands on the same query.

And the door is more open than the old SERP was. Only 17 percent of AI Overview citations now come from pages ranking in the organic top 10, down from 76 percent in mid-2024.

AI systems also favor fresh material: about 65 percent of citations come from content under a year old. A store publishing current, specific answers can get quoted without ever outranking a big publisher. That was never true in classic SEO.

There is also a reader you cannot see in analytics: the model itself. Citation research across 2.4 million domains shows LLMs build their understanding of a brand from your owned content plus everything written about you.

Your blog is the one part of that record you fully control. It is how a model learns what you sell, who it is for, and which questions you are an authority on. No blog means the model learns your category from competitors and Reddit.

So the honest answer is conditional on what kind of blog. Worth it: posts that answer real buying questions in the first lines, structured so an engine can lift the answer, connected to products, and organized as topic clusters so the coverage compounds into authority.

That is the logic behind topical maps and content briefs, and behind matching post intent to product embeds so posts sell instead of just ranking.

Not worth it: thin listicles chasing volume keywords, because the AI answers those directly and keeps the click.

Two practical shifts follow. Measure the blog on revenue per visitor and on citations, not sessions; the conversion math of AI-era traffic makes traffic alone a misleading KPI. And publish consistently, because the freshness bias means a dormant blog decays out of AI answers faster than it ever decayed out of rankings. Consistency is an operations problem, which is exactly what an AI blog agent exists to solve.

The blog did not die. The lazy version of it did.