Is blogging still worth it in 2027?

Why does blogging feel so much harder right now?

Blogging is harder today because publishing has become much easier while getting attention from Google has become less reliable.

Three changes happened almost together. AI made decent articles cheap to produce. Google started answering far more informational questions directly inside search. And conventional content marketing began producing weaker results for a larger share of publishers.

Orbit Media’s latest annual blogging survey captures the shift unusually well. It questioned 1,042 content marketers and found that only 13.9% said their content generated “strong results.” During most of the previous 12 years, that figure stayed between roughly 20% and 30%. Another 18.9% of respondents could not even say whether their blog was producing results.

At the same time, 92.4% of the marketers surveyed now use AI somewhere in their blogging process. Ahrefs reached a similar conclusion from a completely different angle when it analyzed 900,000 recently discovered English-language webpages: 74.2% showed signs of at least some AI-generated content.

Producing another competent article has therefore become much easier. Winning attention with that article has not.

Google has also changed what happens after someone searches. AI Overviews now answer many questions before the user visits a website, while AI Mode pushes search further toward a conversational interface.

That combination explains why blogging suddenly feels more uncertain. Publishers can create more content than ever, yet the old route from article to ranking to click has become less dependable.

Did Google AI Overviews kill blog traffic?

Google AI Overviews have already cut a meaningful amount of traffic from informational blogs, although Google Search remains far too large to dismiss.

One of the strongest recent pieces of evidence comes from an experiment involving 1,065 U.S. desktop users. Researchers randomly showed users Google results either with AI Overviews or with the Overview removed.

When an AI Overview appeared, users made 39.8% fewer outbound organic clicks. Zero-click searches increased 34.5%.

The exposure was also heavily concentrated where bloggers care most. AI Overviews appeared on 53% of informational searches in the experiment, compared with only 15% of transactional searches and 6% of navigational searches.

A site answering questions such as “how long does cooked chicken last?” or “what does magnesium do?” sits directly in the part of search most exposed to AI answers. A website targeting someone looking for a specific software provider, lawyer or expensive product faces a different level of risk.

The source links inside AI Overviews did not compensate for the lost clicks either. Only 7.1% of outbound clicks on pages containing an Overview came through links inside the AI answer itself.

Pew Research had already found the same pattern from observed browsing behavior: users clicked traditional search results on 8% of visits when an AI summary appeared, versus 15% when it did not.

Google says overall organic click volume has remained relatively stable and argues that the visits generated by AI-enhanced search can be more valuable. The latest independent experiment found no measurable improvement in time spent on destination websites, bounce behavior or user satisfaction, so that part of Google’s argument remains disputed.

For informational publishers, the direction is hard to ignore. Google can answer more questions without handing the reader to the website that supplied the information.

Is SEO still worth doing for a blog in 2027?

SEO is still worth doing in 2027 because Google remains the dominant search engine, but building an entire blog business around SEO alone has become much riskier.

Statcounter’s latest worldwide measurement still puts Google at roughly 91% of the search-engine market. Google itself says people perform more than five trillion searches per year.

Even after allowing for more zero-click searches, that leaves an enormous amount of traffic available.

Google has also made clear that AI Overviews and AI Mode still use its underlying search and ranking systems to retrieve sources. The same basics that help a page rank in ordinary search—crawlability, relevance, authority, useful content and strong sourcing—also affect whether Google’s AI systems can surface it.

What has changed is which searches are worth chasing.

An article targeting “how many grams are in an ounce?” has little room to create value once Google can answer the question instantly.

A detailed comparison of payroll software for a French company employing U.S. contractors is harder to compress into one sentence. Original testing, pricing differences, tax implications and first-hand experience give the reader reasons to open the source.

SEO still makes sense. The easy informational end of SEO simply carries much worse economics than it used to.

Can a completely new blog still rank on Google?

A completely new blog can still rank on Google, but anyone expecting large organic traffic within a few months is starting with the wrong assumptions.

Ahrefs followed one million newly published URLs for a year and found that only 1.74% reached Google’s top ten for at least one keyword. When Ahrefs narrowed the sample to two million substantive English-language pages, the success rate improved to 6.11%.

Neither number suggests that ranking has become impossible. They do show how unusual first-page success is across the full population of new content.

Older pages also occupy more of Google’s valuable positions than they did several years ago. Ahrefs found that 72.9% of URLs ranking in the top ten were more than three years old. In its comparable 2017 research, the figure was 59%.

The average number-one result in the newer study was around five years old.

Age itself is unlikely to be the magic factor. Older pages have had time to earn backlinks, accumulate mentions, demonstrate consistent traffic, get updated and establish a ranking history.

For a new blogger, the practical result is the same: you are often competing with assets that have been compounding for years.

That pushes blogging closer to a long-term publishing business and further away from a six-month SEO experiment.

Has AI made blogging too competitive?

AI has made generic blogging extremely competitive because almost everyone now has access to the same production shortcut.

Orbit Media found that 92.4% of surveyed content marketers currently use AI for blogging. Ahrefs’ large sample of new webpages found some AI involvement on almost three quarters of them.

The interesting part is what happened to productivity.

Orbit estimates that the average article now takes around 3 hours and 20 minutes to create, down from a peak of 4 hours and 10 minutes in 2022. Across the average annual publishing volume in its survey, that saves roughly 50 working hours.

Results did not improve alongside that efficiency. Strong reported outcomes fell from about 26% in 2022 to 13.9% in the latest survey.

Orbit also found no meaningful difference in the probability of reporting strong results between AI users and people who do not use AI.

If one blogger can generate a competent first draft in 20 minutes, thousands of competitors can do the same.

AI helps with production costs. It gives very little protection from competition.

What kind of blog content can AI not easily replace?

The safest blog content today contains something the AI could not have known without somebody first doing the work.

That includes original datasets, experiments, first-hand product tests, interviews, proprietary calculations, expert experience, useful tools and reporting that produces genuinely new information.

Google’s own content guidance explicitly asks publishers whether their pages contain original information, reporting, research or analysis and whether the author demonstrates first-hand expertise.

Orbit Media’s 2026 data points in the same direction. Marketers using original research, expert collaboration, stronger keyword research, deeper articles and formal editing processes were more likely to report good results.

Some of the least common tactics produced the largest gaps. Only about 7% of respondents said they regularly collaborated with influencers or experts, yet that group reported strong results at roughly 2.6 times the overall rate.

Orbit’s own annual blogging research shows how this works in practice. Its recurring survey has attracted links from more than 1,600 websites over its lifetime because other writers need the underlying dataset.

An AI system can summarize that research. A competing blogger cannot cheaply recreate more than a decade of survey history.

A useful test before publishing is to ask whether another writer could create essentially the same article by reading the same five Google results.

When the answer is yes, expect heavy competition.

When the article depends on something we measured, experienced, tested, calculated or discovered ourselves, copying becomes much harder.

Can ChatGPT replace the Google traffic blogs are losing?

ChatGPT and other AI assistants are sending rapidly growing referral traffic, but that traffic is still far too small to replace Google for most blogs.

Similarweb measured about 1.13 billion referral visits from AI platforms during one comparison month in 2025. Google Search generated roughly 191 billion referrals over the same period.

That puts Google at around 169 times the referral volume.

Ahrefs found the same imbalance in its own datasets. In one analysis of 3,000 sites, 63% received at least some identifiable traffic from AI assistants, but AI accounted for only 0.17% of the average website’s visits.

Growth has been fast. In a later analysis of 81,947 websites, Ahrefs found average AI referral traffic had increased about 9.7 times year over year, while measured search traffic had fallen around 21%.

More recently, Ahrefs noted that AI referral growth had started flattening over several months, another reason to avoid simply extrapolating the earlier 10x growth rate indefinitely.

Small AI audiences can still be commercially interesting. Ahrefs found that on its own website, AI visitors represented only about 0.5% of traffic in one analysis but generated 12.1% of sign-ups. That worked out to a conversion rate around 23 times higher than conventional organic search.

One company’s conversion funnel is obviously not a universal benchmark. It does show why AI citations can be valuable even before AI sends large volumes.

For now, bloggers should treat ChatGPT, Gemini, Perplexity and other AI systems as additional discovery channels. Google remains several orders of magnitude larger.

Discovery channel	Useful current benchmark	What it tells us
Google Search	~191B referrals in Similarweb comparison	Still massively dominant
AI platforms	~1.13B referrals	Growing fast from a much smaller base
AI share in Ahrefs' 3,000-site sample	0.17% of average traffic	Usually nowhere near search volume
AI visitors on Ahrefs.com	0.5% of traffic but 12.1% of sign-ups	Some AI traffic can have unusually strong intent

Are people still willing to read and pay for written content?

People are still paying for written content at meaningful scale, which makes the idea that “nobody reads blogs anymore” hard to defend.

Substack currently says its network has more than five million paid subscriptions. More than half of new subscriptions now come through Substack’s own recommendation and discovery network.

Beehiiv shows a similar direction. Paid subscription revenue across its platform increased from $8 million in 2024 to $19 million in 2025, a 138% jump. The number of creators earning subscription revenue doubled.

Its wider 2026 newsletter report says publishers sent 28 billion emails during 2025 to more than 255 million unique readers. Average open rates remained above 41%.

Those numbers do not mean a traditional personal blog automatically becomes a viable subscription business. They tell us something more useful: written publishing still commands attention and money when the material gives people a reason to come back.

The format is also becoming harder to classify. A modern independent publisher may post the full article on a website, send it by email, distribute excerpts on LinkedIn, discuss the same research on YouTube and offer a paid version to subscribers.

For many successful writers, “blog” and “newsletter” are becoming parts of the same publishing business.

Can a blog still make real money from display ads?

Display ads can still produce a real blogging business, but they work much better once a publisher has substantial traffic and several ways to acquire it.

Premium ad networks are still actively serving independent publishers.

Raptive accepts eligible sites beginning around 25,000 monthly pageviews, subject to additional requirements. Mediavine’s Journey product starts much lower, while its main program now uses prior annual advertising revenue rather than the old traffic-only threshold.

Mediavine also has publisher tiers beginning at $100,000 in annual ad revenue and continuing through $250,000, $500,000 and $1 million.

Those thresholds do not tell us what the average blogger earns. They do confirm that large independent advertising businesses continue to exist.

The weaker part of the model is traffic acquisition.

A publisher receiving 500,000 visits per month from a mix of direct readers, email, Pinterest, search, social platforms and returning users owns a much more resilient business than a publisher receiving the same 500,000 visits almost entirely from informational Google searches.

Both can show the same advertising revenue today. Their risk is completely different.

Display ads still work. Building thousands of generic articles specifically to manufacture Google pageviews looks much less attractive these days.

Is affiliate blogging still worth it?

Affiliate blogging still works when the blog actually helps someone make a difficult purchase, while thin “best products” pages are losing much of their old advantage.

Brands continue spending heavily on affiliate partnerships. Impact.com surveyed 818 marketers across eight countries and found that 74% attributed between 11% and 30% of total company revenue to affiliate marketing.

The money available to publishers therefore has not disappeared.

The economics at the publisher level can still be tight. Amazon’s U.S. affiliate schedule currently pays around 3% in many large categories such as home, beauty, sports, baby products and tools. A $100 purchase at a 3% commission produces just $3 for the publisher.

That model needs either a large audience, expensive products, higher-paying programs or unusually strong purchase intent.

Search competition is also moving closer to commercial queries. Semrush’s large analysis of AI Overviews found that Google initially concentrated them on informational searches but progressively expanded them into commercial and transactional territory.

A reader comparing a $2,000 camera, enterprise software package or specialized tool can still benefit from a detailed article. The publisher has to contribute more than rewritten specifications.

Actually testing the camera, comparing image files, measuring battery life and explaining what happened after six months gives the article something durable.

Summarizing five other review pages does not.

Does the blogging niche matter more now?

The niche matters more now because 10,000 readers can be almost worthless in one market and extremely valuable in another.

Consider two blogs receiving 50,000 monthly visits.

One answers simple informational questions and monetizes primarily with ads. AI can answer many of those questions directly, while each remaining pageview is worth relatively little.

The other publishes specialist tax guidance for business owners. A single reader might buy software, book a consultation or become a client worth thousands of dollars.

Newsletter economics show the same gap.

Beehiiv’s paid-newsletter data found a median free-to-paid conversion rate of 0.62% across the platform, but results varied dramatically by category. Its data suggests that a 1,000-subscriber investing newsletter can generate more than ten times the annual subscription revenue of a similarly sized travel newsletter under median category economics.

Subscriber lifetime value also ranged from roughly $83 to $230 depending on the niche.

Traffic volume alone therefore tells us surprisingly little.

Before starting a blog, we should care much more about what each reader might eventually do: buy something, subscribe, hire someone, request a quote, recommend the publication or keep returning for years.

A small audience in the right niche can beat a huge audience built around low-value questions.

Should a blog still depend mostly on Google?

A new blog should avoid depending mostly on Google because direct audience channels now matter almost as much as rankings.

Email is the obvious starting point.

Once a Google visitor joins an email list, the next article can reach that person without winning another search result. Search becomes an acquisition channel rather than the only relationship with the reader.

The current growth of newsletter platforms shows how valuable publishers consider that direct connection. Beehiiv says its publishers sent 28 billion emails in 2025, while Substack says more than half of new subscriptions on its platform now come through its internal recommendation network.

Businesses are spreading content more widely too. HubSpot’s survey of more than 500 marketers found that companies with blogs commonly distributed articles through social media, email and paid promotion rather than waiting solely for organic search.

AI visibility may reward the same broader footprint. Ahrefs studied 75,000 brands and found a strong correlation between mentions across the web—particularly on YouTube—and visibility inside systems such as ChatGPT, Google AI Mode and AI Overviews.

Correlation cannot tell us that a YouTube mention directly causes an AI citation. It does suggest that brands visible in many places tend to appear more often in AI answers.

A modern blog therefore works better when one strong piece can travel: website article, newsletter, social posts, video discussion, expert outreach and eventually search or AI citations.

The website remains the permanent home of the work. Discovery can come from several places.

Does blogging still make sense if you already sell something?

Blogging currently makes more economic sense for many businesses and experts than it does for publishers whose only product is pageviews.

The difference comes down to what one visitor is worth.

A media blog displaying ads might need tens of thousands of visits to generate meaningful revenue. A consultant, lawyer, recruiter, agency or B2B software company can sometimes justify months of content work from a handful of customers.

HubSpot’s survey of more than 500 marketers found that companies maintaining blogs used them for several goals at once. Brand awareness was cited by 66%, customer engagement by 53%, lead generation by 49% and SEO by 34%.

Half of the respondents said blogging ROI had improved over the previous year, while 31% said it was roughly unchanged.

HubSpot sells marketing software, so we should treat its survey as industry evidence rather than independent proof that every company needs a blog.

The underlying economics are straightforward, though.

Imagine an article receives only 1,500 visitors in a year. That looks disappointing for an ad-funded publication. If the same article generates three customers worth $5,000 each, the business owner sees a completely different result.

That is why falling search click-through rates threaten mass-traffic publishers more severely than they threaten every company that publishes articles.

How much work does a successful blog take now?

A successful blog still takes a lot of work, and AI has mostly shortened the drafting stage rather than removing the difficult parts.

Orbit Media’s latest benchmark puts the average article at 1,312 words and around 3 hours 20 minutes of production time. Respondents publish an average of roughly 57 articles per year.

That alone works out to around 190 hours of writing.

Research, interviews, editing, original graphics, photography, data analysis, updating older posts, email distribution, SEO and promotion all come on top.

The harder tactics also appear to be some of the ones producing better outcomes. Orbit found stronger reported performance among marketers doing original research, expert collaboration, more substantial articles and formal editing.

Mass AI publishing offers the opposite proposition: very low effort per article and almost no scarcity.

A blog can eventually become easier to maintain once a large archive ranks, an email list grows and old posts continue generating leads or sales. Getting to that point still requires substantial upfront work.

Anyone looking for genuinely passive income should be cautious with blogging today. The compounding can become passive later; building the thing rarely is.

Who should still start a blog in 2027?

People with valuable expertise, a product to sell, first-hand experience or the ability to produce original information still have a strong reason to start a blog in 2027.

A consultant can turn articles into leads. A SaaS company can use technical guides and calculators to attract potential customers. A product specialist can combine real testing with affiliate revenue. A researcher can publish datasets that other websites cite. A niche publisher can build an email audience and layer advertising, sponsorships, subscriptions and products on top.

The weaker setup is easy to recognize too.

Someone chooses a high-volume niche, finds hundreds of low-competition keywords, generates articles from existing search results with AI and expects Google traffic plus display ads to create passive income.

That approach now runs into several problems at once: AI makes the content easy for competitors to reproduce, new pages face entrenched Google results, AI Overviews absorb more informational clicks, and low-value pageviews require enormous scale to monetize well.

The gap between those two blogging models has become wide enough that asking whether “blogging” works without specifying the model is almost meaningless.

Blogging model	How it looks heading into 2027	Main reason
Expert or consultant blog	Strong	A small number of readers can become valuable clients
B2B or SaaS blog	Strong	Articles can generate leads and support product acquisition
Original research publisher	Strong	Proprietary information is difficult to copy
First-hand review site	Viable	Real testing adds information AI cannot manufacture
Newsletter-led niche publication	Viable	Direct audience reduces dependence on search
Pure display-ad SEO site	Much harder	Needs large traffic while informational clicks are under pressure
Generic affiliate listicle site	Much harder	Weak differentiation and increasing AI competition
Mass AI-generated niche site	Weak	The production advantage is available to everyone

Is blogging still worth it in 2027?

Yes. Blogging is still worth it in 2027 when the blog builds an audience or supports a real business, while the old model of publishing interchangeable SEO articles and monetizing Google pageviews has become far less attractive.

The latest evidence is unusually consistent.

Google still controls roughly 91% of worldwide search, so ignoring SEO would make little sense. At the same time, recent experimental research found almost 40% fewer outbound organic clicks when an AI Overview appeared. New content faces a tough ranking environment, with nearly three quarters of current top-ten URLs in Ahrefs’ study already more than three years old.

AI has also removed much of the scarcity from basic article production. More than nine out of ten content marketers in Orbit Media’s latest survey now use AI, yet only 13.9% reported strong results. As seen above, that was the weakest result in the survey’s 13-year history.

The healthier parts of written publishing look quite different. More than five million paid subscriptions now exist across Substack. Beehiiv’s paid subscription revenue rose 138% in one year. Brands continue spending on affiliate partnerships. Large independent sites continue earning substantial display-ad revenue. Businesses still use blog content to generate leads and customers.

The common thread is reader value.

Original research can be cited. First-hand reviews can influence purchases. Expert articles can bring in clients. A useful newsletter can create recurring revenue. A proprietary tool can attract links for years.

Generic information has become much cheaper.

Starting a blog in 2027 still makes sense if we are building expertise, an audience, a brand or a business asset that can compound over several years. Starting one because we expect hundreds of AI-assisted articles to produce easy Google traffic is a much weaker bet today.

Blogging still works. The easy version of blogging mostly does not.