Almost Everyone Uses AI to Write Blog Posts in 2026. Results Just Hit a Twelve-Year Low.

In 2026 almost everyone writing blog posts uses AI. Only 13.9% of marketers say their blog delivers strong results, the lowest in twelve years of this survey (Orbit Media, September 8, 2026). AI is not the villain. The survey points at everything that quietly got dropped around the writing.

Everyone Switched, and the Results Went Down

Orbit Media has asked content marketers the same questions every year for twelve years. The 2026 round surveyed 1,042 of them (Orbit Media, September 8, 2026).

Two numbers carry the story. In the 2026 survey, 92.4% use AI for blogging, and 13.9% say their blog delivers strong results (Orbit Media, September 8, 2026). That second figure is an all-time low for the study, six points under the previous twelve-year low (Orbit Media, September 8, 2026). Search Engine Land put the comparison plainly: 14% in 2026, down from 26% in 2022 (Search Engine Land, September 10, 2026).

Writing got faster along the way. The average post took 3 hours and 20 minutes in 2026, down from a peak of 4 hours and 10 minutes in 2022 (Orbit Media, September 8, 2026). Search Engine Land works that out to roughly 50 hours saved per marketer per year (Search Engine Land, September 10, 2026). Posts got shorter too, averaging 1,312 words in 2026, down from a peak of 1,427 in 2023 (Orbit Media, September 8, 2026).

So: faster, cheaper, shorter, and working less than at any point since the survey started.

The Survey Does Not Actually Blame AI

Here is the part the headlines skip. Orbit found that AI use correlates with nothing (Orbit Media, September 8, 2026). Search Engine Land reported the same read: AI adoption showed no link to better results (Search Engine Land, September 10, 2026).

Read that again, because it cuts both ways. Using AI is not what dragged results down. It is also not helping. It is a speed tool, and speed was never the thing that made blogs work.

What changed is what people stopped doing with the time they saved.

Six Things That Quietly Got Dropped

The survey tracks practices that go with strong results, and marketers report doing fewer of them every year: keyword research, original research, working with experts or influencers, paid promotion, formal editing, and consistent use of analytics (Search Engine Land, September 10, 2026).

Some of the drops are steep. Collaborating with an influencer or outside expert fell from 25% of marketers in 2017 to 7% in 2026, and that practice beats the survey benchmark by 2.6 times (Orbit Media, September 8, 2026). Publishing original research improves the likelihood of performance by 50%, and fewer marketers do it (Orbit Media, September 8, 2026). Keyword research is converging at about 23% for both "always" and "never" in 2026, which means as many people ignore keywords entirely as always check them (Orbit Media, September 8, 2026).

Human editing shows up on the list too. Marketers who use a formal editor beat the benchmark by nearly two times (Orbit Media, September 8, 2026).

The stacking effect is the real finding. Marketers using six or more of these practices are nearly three times as likely to report strong results (Orbit Media, September 8, 2026). Not one respondent in the 2026 survey was doing all eight (Search Engine Land, September 10, 2026). Ann Handley of MarketingProfs called the result "a wake-up call" (Search Engine Land, September 10, 2026).

SEO Experts Draw the Same Line

A separate survey from the same week backs this up from the search side. Zyppy asked 131 SEO professionals to rate 103 possible Google ranking factors, collecting 13,665 ratings (Search Engine Land, September 9, 2026).

Their view of AI content was not good or bad. It was conditional. They rated AI content favorably when it got substantial human review and added something unique, and negatively when it was published at scale with little added value (Search Engine Land, September 9, 2026). Original research and first-party data pulled some of the highest ratings in the content quality group (Search Engine Land, September 9, 2026).

Both surveys landed in the same week with the same answer. The machine part is fine. The dropped part is the problem.

How This Blog Is Built, Since We Are Preaching

Fair question: does this blog follow its own advice? Every post starts from one question a real customer types into a search box or an assistant. A person edits every post before it goes live, and the AI part is disclosed at the bottom. One post a week goes out on LinkedIn, and the numbers get checked on a dashboard every Monday. That covers four of the six practices the survey ties to results.

The other two are being added one at a time, because the survey says practices stack. First up is a number only we could have collected, and here is one for the record: the agent draft behind this post took under four minutes on September 16, 2026. The human editing took about 30 minutes, on purpose.

So Is Blogging Still Worth Doing?

Yes, if you do it the way the survey describes. Marketers using six or more of those practices are nearly three times as likely to report strong results (Orbit Media, September 8, 2026), so the gap is a practices gap, not a dead channel.

There is also a newer reason. When someone asks a search engine or an AI assistant who to hire, the answer comes from what is written on your site. Two earlier posts here covered that: assistants recommend local businesses from the website, not the reviews, and the customers who arrive that way tend to show up further along in their decision. A blog post that answers a real question is the page an assistant can quote. That is how a post becomes a lead in 2026, and it only happens if the post is worth quoting.

What to Actually Do This Week

• Count your practices, not your posts. Six or more is where the results change (Orbit Media, September 8, 2026). Write down how many you are actually doing. Most people find the honest count is lower than they assumed.
• Put a human name on the editing. Marketers with a formal editor beat the benchmark by nearly two times (Orbit Media, September 8, 2026). Someone has to read the draft against reality before it publishes, and know your business well enough to catch what is wrong.
• Add one thing only you have. A number from your own jobs. A photo of real work. The answer you give customers every week. That is original research at small-business scale, and original research improves the likelihood of performance by 50% (Orbit Media, September 8, 2026).
• Pick the keyword before you write, not after. About 23% of marketers never consider keywords in 2026 (Orbit Media, September 8, 2026). That is an easy place to get ahead of people.
• Plan the promotion in the same hour you plan the post. Publishing is not distribution, and paid promotion sits on the list of practices marketers dropped (Search Engine Land, September 10, 2026).
• Check your numbers on a schedule. Consistent use of analytics is on that dropped list too (Search Engine Land, September 10, 2026). Monthly is enough. Once a year is not. For Google's AI results, the free Search Console report that shows whether your pages appear in them is an easy place to start (added October 7, 2026).

One Thing to Skip

Skip the volume plan. More posts, faster, with nothing added is exactly the pattern the 2026 survey describes, and it is the pattern SEO professionals rated negatively (Search Engine Land, September 9, 2026). A couple of posts a month you would put your name on beats a pile you would not.

One honest caveat on all of this. These are marketers reporting on their own results, not measured traffic, so read every figure as "marketers say." The direction is still hard to argue with after twelve years of the same question.

Since this blog practices what it preaches: this post was drafted by Ninth Degree's agent pipeline and edited by a human before publishing. The service we sell is the whole list above. AI writes the first draft fast. People do the research, the editing, and the promotion.