A business blog is the section of a website where a company publishes content that answers its customers' questions before they become customers. It is not a diary, it is not the news section and it is not the place to announce the trade fair you went to.
The current state of things needs saying up front, because it changes everything else. In the 2026 edition of the annual survey by Orbit Media, only 13.9% of 1,042 content marketers say their blog delivers strong results. It is the lowest figure in twelve editions of the same survey. Another 18% say they do not know whether the blog is producing anything.
This does not mean starting a business blog is a mistake. It means the reasons almost everyone starts one no longer hold, and the few that work are different from the ones you usually read about.
The business blogging statistics nobody checks
Guides on business blogging keep repeating the same handful of numbers. We traced them back to the source one by one. None of them leads to a study you can actually read.
| Statistic in circulation | Who it is attributed to | What you find when you trace it |
|---|---|---|
| "Companies that blog generate 67% more leads" | DemandMetric, cited by HubSpot | No year, no published study. Other sources attribute it to someone else |
| "A blog brings 55% more visitors" | HubSpot, citing itself | No year, no dataset, no method |
| "434% more indexed pages" | BKA Content, or OptinMonster | Goes back to a post from over ten years ago, with no raw data |
| "13 times more likely to see positive ROI" | HubSpot | No year, no study |
| "80% of bloggers report strong results" | Orbit Media, no year given | The 2026 edition of the same survey says 13.9% |
| "2,450 words is the ideal length" | No attribution | The average length measured by Orbit Media in 2026 is 1,312 words |
The last row deserves attention. The same organisation that ten years ago measured 80% of blogs with strong results now measures 13.9%, with the same methodology. The old statistic keeps circulating because it suits people who sell content. If someone puts it in a proposal, you now know where it comes from.
What has really changed between 2022 and today
Three things, all measured.
The first is about clicks. The Pew Research Center tracked the browsing of 900 US adults across 68,879 Google searches in March 2025. When an AI-generated summary appears on the page, users click a traditional result in 8% of visits; without a summary, in 15%. 26% end the session straight after reading it, against 16% on pages with only classic results. The sample is American and should be read as such, but the same feature runs in Google UK, and we have already covered the phenomenon in our piece on zero-click searches.
The second is about whoever ranks first. Ahrefs compared 300,000 keywords using Search Console data, March 2024 against March 2025, on desktop: where an AI Overview appears, the CTR of the first organic result drops by 34.5%. Ranking first is worth less than it was two years ago.
The third is about questions. Semrush analysed over a billion rows of clickstream data from a US panel between October 2024 and February 2026: between 65% and 85% of prompts typed into ChatGPT do not resemble any traditional search keyword. They are long sentences with context built in. A blog built only on short keywords captures a shrinking share of what people actually ask.
How much traffic a business blog really gets
This needs an order of magnitude, because it is the part missing from every guide in circulation.
"Business blog" gets around 210 searches a month in the UK, according to Google Ads data (DataForSEO, Google UK, September 2026). "How to start a business blog" gets 30, "company blog" 30. These numbers say one thing: the UK market is a fraction of the US one, a B2B niche is a fraction of that again, and any traffic forecast copied from an American article is out by an order of magnitude.
A concrete example from our own projects. For an Italian industrial manufacturer we work with, we track 135 product keywords. Eighty of them currently sit in Google's top ten. In the data window synced from Google Search Console, those same 135 keywords produced 7,156 impressions and 70 clicks: an overall click-through rate of 0.98%. Seventy of the eighty keywords on page one did not bring a single click.
| Indicator | Measured value |
|---|---|
| Product keywords tracked | 135 |
| Keywords in the top ten positions | 80 |
| Impressions recorded by Search Console | 7,156 |
| Clicks recorded by Search Console | 70 |
| Overall CTR | 0.98% |
| Page-one keywords with zero clicks | 70 of 80 |
| Monthly volume reported by the database for the 135 keywords | 32,880 searches |
The gap between the 32,880 monthly searches the database assigns to those keywords and the 7,156 actual impressions is not a tracker error. It comes from how volumes are built: the Google Ads documentation says that average monthly searches aggregate the keyword and its close variants, and that the figure is rounded. In our set, fourteen volume values are shared by several keywords: the value 50 appears on fourteen different queries, 30 on fifteen. Anyone adding up the volumes of variants to estimate a blog's traffic is counting the same question several times. This is why keyword research based on volume alone produces inflated content plans.
When a business blog is worth it, and when to leave it
It is worth it when the product has a long buying cycle and the buyer arrives with technical questions before asking for a quote. In that case the blog is not there to get visits, it is there to be found by the four or five people a month who are deciding who to give a supply contract to.
The case we describe in full in SEO for manufacturing works like this. Macropix, a Milan-based maker of custom LED screens, went from position 88 to second place on Google Italy for "monitor pubblicitario" (advertising display) and now holds 25.55% share of voice on the ledwall cluster, ahead of Amazon.it at 18.08%. The nine keywords that bring it the most qualified leads do not contain the company name: they are descriptions of products and applications.
It is not worth it in three recurring situations. When nobody in the company can spend time reviewing what goes out, because editing decides the outcome: according to Orbit Media, those with a formal human editor beat the results benchmark, while those who leave everything to AI fall below it. When the stated goal is "raising brand awareness" with no number attached. And when market demand simply does not exist, which you can check in half a day by looking at the search intent of the sector's queries rather than their volume.
There is also a downside risk that is rarely mentioned. Ahrefs analysed 1,600 SaaS companies and their worst traffic drops: in one case 99.7% of an entire domain's loss was concentrated in the /blog folder; in another, a site that had published 1,800 mass-produced articles lost 99.52% of its traffic after a manual action by Google. A badly run business blog is not neutral, it weighs on the rest of the site.
Decisions to make before the first article
Where to put it. A subfolder on the same domain, such as company.co.uk/blog/. Not a subdomain, not a separate domain: the authority signals the blog builds up need to benefit the commercial pages, and that only happens if they sit on the same host and link to each other. We have written a guide on how to connect them with internal links.
Who writes. The person in the company who answers customer emails, not an outside writer who has never seen the product. An outsider can rewrite and optimise, but the raw material has to come from inside. The best questions are already in lost quotes and after-sales phone calls.
How often. Orbit Media finds an average of 3 hours 20 minutes per article, down from 4 hours 10 minutes in 2022, and an average length of 1,312 words. Worked out for a part-time internal resource, two articles a month is already a real commitment. Better two done well than eight fillers: the question of the ideal blog post length matters far less than how often you can keep up the standard.
What to publish first. Not the introductory article on the industry. The pages that answer the question before the purchase: how much it costs, what the difference is between two solutions, how it is installed, which certifications are required. The method for choosing the order is the one we describe in the content marketing plan.
What to measure (almost certainly not traffic)
In Orbit Media's 2026 survey, 39% of marketers say they track blog traffic. That group reports strong results in 11% of cases, below the overall average of 13.9%. Those who track qualified leads and deals in the CRM are above it. Measuring visits is not just insufficient, it correlates with doing worse.
The three numbers we recommend keeping for a B2B business blog are contact requests attributed to blog pages, cost per request compared with the paid channel, and the share of the sector's commercial queries on which the site appears. The other SEO KPIs help you work out what to fix, not decide whether to carry on. On how to turn all this into a figure, we have written a guide to SEO ROI.
The thing almost nobody does
Publishing your own data. Orbit Media finds that those who publish original research are about 50% more likely to report strong results, and that the practice has been declining since 2018. The Content Marketing Institute, in its 2026 B2B research with MarketingProfs on 1,015 marketers, measures the same imbalance from another angle: 95% use AI applications, 87% report higher productivity, but only 39% see content performance improve.
A manufacturer has data no competitor can copy: average delivery times, failure rates, energy use measured in the field, five-year maintenance costs. That is material people cite, and citing means links. It is worth more than twenty articles explaining what the product is.
The timing helps more than it seems. According to the Office for National Statistics, the share of UK businesses with 10 or more employees using at least one AI technology rose from around 12% in September 2023 to around 35% in June 2026. Generic content is getting cheaper to produce for everyone; figures that only you hold are not. Whoever brings real numbers starts ahead for another couple of years.
Frequently asked questions
The main cost is time. Orbit Media finds an average of 3 hours 20 minutes per article in 2026, plus promotion. Two articles a month written in-house and reviewed by an outside editor come to roughly one working day a month. If the sums do not add up against the value of an average contract, the blog is not the right channel.
On the industrial projects we run, the first qualified leads from the blog arrive between the sixth and the twelfth month, and consolidating positions takes longer still. The Macropix case, from position 88 to second place, spans five years.
It does, provided the goal changes. According to the Pew Research Center, in the United States an AI summary on the page cuts clicks on organic results from 15% to 8% of visits. The blog is less and less about collecting visits and more and more about being the source that search engines and assistants cite when they answer.
A subfolder on the same domain. It keeps the site together and lets internal links pass authority from informational pages to commercial ones. A subdomain is only worth considering when technical constraints prevent installing the blog on the same host.
For drafting, yes; for substance, no. Orbit Media finds that 92.4% of content marketers use AI and that those who use it do not get better results than average, while those who do not use it at all tend to get worse ones. The difference is the human editor: those with a formal one sit above the benchmark.
One thing the industry rarely says: a business blog is becoming more useful as an internal archive than as an acquisition channel. Answers written well once end up in quotes, sales emails, presentations, and now also in the systems that answer in place of search engines. Measuring it on visits alone means throwing away the part that pays. If you want to know whether it makes sense for your company, look at the projects we have worked on or talk to us about SEO.
