Search whether blogging is dead and nearly every answer is written for someone who earns money from ads or affiliate links. Their question is whether traffic still pays. A business owner is asking something else: whether publishing still brings in customers. The pages ranking for this search rarely get to that question.
The short version. The blog as a company diary is dead, and for most small businesses it never worked. The page that answers a question a buyer types before they buy is not dead, though AI answers now take some of its clicks. Publish fewer, narrower pages built around those questions and judge them by enquiries rather than visits. Skip the blog entirely if nobody searches for what you sell.
- The blog as a company diary is dead, while the page that answers a buyer's question before they buy is not.
- AI summaries cut clicks: Pew found users clicked a result in 8% of visits with a summary against 15% without, and question-style searches were the most likely to show one.
- Publish one question per page, starting with prices, comparisons, timelines, diagnostics and who should not buy.
- Judge content by non-branded clicks and enquiries rather than pageviews, and give a page three months before deciding.
- Skip the blog if nobody searches for what you sell, the site cannot convert, nobody will keep it going, or you need customers this month.
Which kind of blog died
Two different things share the word blog. The first is the news feed: the team day out, the new hire, the award shortlist, a post because it was Tuesday. It was written for the business, and nobody outside the business was looking for it. Search engines had no reason to show it and customers had no reason to read it.
The second is the page that answers a question a buyer has before they buy. What does this cost? How long does it take? Is it worth it for a business our size? Those pages were never a diary. They are the part of a sales conversation that happens before anyone calls you.
The first kind is the one that died. Google's own guidance on people-first content warns against producing lots of content on many topics in the hope that some of it performs. That is a fair description of most company blogs. The same page asks whether you are writing to a word count Google prefers, and answers itself: "No, we don't."
What AI answers changed, and what they did not
This is the part of the question that has changed in the last two years, and the evidence is better than most of what is written about it.
Pew Research Center tracked the browsing of 900 US adults in March 2025. When a Google results page carried an AI summary, users clicked a traditional result in 8% of visits. Without one, they clicked in 15%. Searches that began with a question word such as who, what, when or why produced an AI summary 60% of the time.
SparkToro's analysis of Similarweb panel data, published June 2026, found 68% of US Google searches ended without a click in the first four months of 2026. Separately, Ahrefs compared Search Console data for 300,000 keywords and found the top result's click-through rate was 58% lower in December 2025 when an AI Overview appeared.
So a page whose only job is to define a term or answer a simple factual question is losing clicks, and that is documented. What we infer from it, and it is inference, is that questions closer to a purchase lose less. Someone choosing an accountant for an import business wants to compare options and contact one, not read a summary. How to check whether it is happening on your own site is set out in AI Overviews and your traffic.
Google's documentation on AI features says there are no additional requirements to appear in AI Overviews, and that existing SEO fundamentals still apply. A page that answers a buyer's question plainly is also the page an AI system can quote, which is why GEO, AEO and AIO are one job.
What to publish instead
Start from the questions buyers ask before they buy, in their own words. Your inbox and your sales calls hold a better list than any tool, and the method is in write for the question, not the keyword. Five kinds of page earn their place.
- The price page. What it costs, what it covers and what moves the number. A page carrying a real price can be quoted. A page saying "contact us for a quote" cannot.
- The comparison. Your option against the other ways a buyer could solve the problem, including doing it themselves.
- The timeline. How long it takes, in stages, and what makes it slower.
- The when-not-to. Who should not buy from you. Most businesses refuse to write it, and it is the page that makes the rest believable.
- The diagnostic. Something has gone wrong; here is how to find out what, in order.
Each is one question on one page. None needs a date to stay useful, which is the practical difference from a diary. Update them when the facts change and they keep working.
How often, and what it costs
Less often than you have been told. Four good pieces a month builds depth in one topic area within a quarter, and eight thin ones lose to four good ones. Plenty of businesses need fewer. A service business with ten buyer questions needs ten excellent pages, then updates.
Writing them yourself is a real option. Budget three to five hours for a researched, edited piece that answers one question properly. If you know the subject better than any agency will and you have the hours, write them yourself.
From us, content marketing is $950 a month for four pieces as a standalone service. That covers the quarterly plan, buyer-question research, distribution and revenue attribution. Inside the Grow retainer it is the $1,100 SEO and content line, alongside the technical work.
What moves the number is scope. Ten pieces a month is the $1,200 content line in Scale. A second language or specialist subject matter is scoped in writing before anything is signed. How urgently you want it does not move it. The full breakdown is in what content marketing costs.
How to tell whether it is working
Ignore pageviews as the headline. A piece read by two thousand people who will never buy is worth less than one read by forty people holding a quote.
- Non-branded clicks in Search Console: people who found you without already knowing your name.
- Enquiries that landed on, or mention, a specific page.
- What AI assistants say when you ask them about your prices and services, and whether it is correct.
Give a question-led page three months before judging it. If a page has neither clicks nor enquiries after six, rewrite it around a sharper question or remove it. For a UAE law firm, non-branded clicks rose 3.4 times in six months. That is one client over one period, not an average, and the case study sets out what was done.
Count the questions your buyers asked last month that your website does not answer. That number is your content plan, and it is usually longer than a year of diary posts. It is also the plan our content marketing starts from, at $950 a month for four pieces on the pricing page.
If the count is zero, read the next section before spending anything.The $950 standalone content marketing rate, the $1,100 SEO and content line inside the Grow retainer and the $1,200 ten-piece content line in Scale are our published prices, listed on the pricing page. The 8% and 15% click rates and the 60% AI summary rate for question-word searches are from Pew Research Center. Pew tracked the browsing of 900 US adults in March 2025 and published on 22 July 2025. The 68% zero-click share for US Google searches in January to April 2026 is from SparkToro. Its analysis of Similarweb panel data was published 9 June 2026. Google's statements on people-first content and AI features are from its Search Central documentation, both pages last updated 10 December 2025. The 3.4x non-branded clicks figure is from one client engagement and is set out on the case study page. The diary versus answer-page distinction, the five page types and the four-pieces cadence are editorial judgement. So are the three-to-five-hour writing estimate, the three- and six-month review points, and the four situations where a business should not blog. The view that questions close to a purchase lose fewer clicks is our inference, not a documented finding. The 58% lower top-result click-through rate when an AI Overview appears is from Ahrefs' study of 300,000 keywords, comparing December 2023 with December 2025, published 4 February 2026.
When a business should not have a blog
Some businesses should not, and the honest answer to the headline depends on which of these describes yours.
Nobody searches for what you sell. If every customer arrives by referral or through a sales team, articles are an expensive way to feel busy. Spend the effort on the pages a referred buyer checks: who you are, what you do, what it costs.
The website cannot convert. More readers landing on a site that does not say what you do means more people leaving. Fix the site first.
Nobody will keep it going. A feed whose last post is two years old tells a visitor the business may have stopped too. Five answer pages with no dated feed beat a feed that stopped.
You need customers this month. Question-led pages take months to earn clicks. If the business needs enquiries in weeks, paid search is the honest answer.
If none of those apply, write down the ten questions buyers asked you last month. That list is the blog worth having, and it is not dead.