Write down the ten questions your sales conversations keep repeating, answer each one properly on its own page, and stop there until those pages are earning something. That is the whole method. Everything else in content marketing is scheduling, and scheduling is what people reach for when they do not want to decide what to say.
Start with the question, not the calendar
Most content programmes fail at the first step, which is choosing the subject. A publishing calendar fills itself with topics that are easy to write rather than topics somebody needs. Six months later you have forty posts, decent traffic on none of them, and no way to tell which ones influenced a sale.
The fix is unglamorous. Sit in on three sales calls or read the last thirty support tickets, and write down the questions in the words the customer used. Those questions are your brief. They come pre-loaded with search demand, because the person asking you has almost certainly typed something similar into a search box first. We use the same list to shape the technical work described in SEO Fundamentals Every Business Should Know.
- The question, in the customer’s words, not your product vocabulary.
- Who asks it, and what they are about to decide.
- The honest answer, including the cases where your product is the wrong choice.
- One number, price or constraint that makes the answer concrete.
- What you want the reader to do next, and the single link that lets them.
What the research says about effectiveness
It helps to know how everyone else is doing, because the honest answer is: not brilliantly. In CMI and MarketingProfs’ 2026 B2B research, fielded between 24 June and 14 August 2025 with 1,015 B2B marketers, respondents rated their own marketing effectiveness over the previous twelve months. Twelve per cent said highly effective, meaning they exceeded goals. Forty-seven per cent said somewhat effective. Thirty-one per cent were neutral, and 10% said they fell short.
Read that as permission to lower the drama. If nearly a third of professional marketing teams describe their own year as mixed, your blog not transforming the business in a quarter is normal rather than a failure of nerve.
Strategy beats tooling, and the data is blunt about it
The same study asked how content strategy effectiveness had changed. Ninety-seven per cent of marketers said they have a content strategy, and among them 13% said it significantly improved, 48% somewhat improved, 30% stayed stable and 9% declined. So 61% reported improvement, which is the cheerful headline.
The interesting part is why. Seventy-four per cent of those reporting improvement pointed to strategy refinement. Fifty-one per cent credited new technology, including AI and automation, and 40% cited team restructuring and resource changes. Tools came second to thinking. That ordering matters when a vendor tells you a platform will fix your content.
On AI specifically, the research is neither hype nor dismissal. Among marketers using AI for content creation, 87% said productivity improved, 6% saw no change and 3% said it got worse. Productivity is not the same as effectiveness, and the report’s own framing is that the teams winning are building fundamentals rather than playing with prompts. Our view after a year of watching this: draft faster if it helps, but keep a human deciding what is true and what is worth saying.
Paraphrasing the CMI 2026 B2B researchTools do not create strategy. They make the absence of one obvious faster.
Which formats and channels actually land
Nearly every B2B marketer in the study (96%) says their organisation creates thought leadership content, but few manage it well: 17% described their programme as exploratory and 36% as developing, against 11% at the advanced or leading end. Employee participation is the quiet bottleneck, with 37% saying minimal participation, meaning under 5% of the people with specialist knowledge actually contribute.
| Where thought leadership is published | Rated most effective | What it demands of you |
|---|---|---|
| 76% | A named human posting consistently, not a company page broadcasting | |
| Email newsletters | 54% | A list you own and a reason to open it; median open rate across all industries was 43.46% in 2025 |
| Speaking events and webinars | 52% | One good talk, reused: slides, a recording, a written version, three clips |
Notice what is missing from that list: posting volume on every network. The organic social channels still matter, but as distribution for something that exists rather than as the thing itself. We argued the cadence case in Social Media Marketing Without the Burnout.
How do marketers judge whether thought leadership worked? Eighty per cent look at audience engagement such as views, downloads and shares. Sixty-three per cent look at business impact including leads and pipeline influence. Forty per cent gather audience feedback and 38% track brand authority signals like speaking invitations and citations. If you only have capacity for one, use the 63% answer.
Writing for people first, and for AI search second
Google’s guidance on this has barely changed in years, and it is still the clearest brief available. Its documentation on creating helpful, reliable, people-first content asks whether the page would be useful to someone who already trusts you, whether it demonstrates first-hand experience, and whether it leaves the reader feeling they have learned enough to achieve their goal. Those are editorial tests, not keyword tests.
The AI search panic has not changed the brief either. Google’s own page on AI features is explicit that the best practices for search remain relevant, that there are no additional requirements to appear in AI Overviews or AI Mode, and that you do not need to create new machine-readable files or special schema.org markup for them. Structured data still helps for the features it was designed for, and that is all.
- The 2025 Web Almanac found meta descriptions on only 67.7% of desktop pages, against title tags on 98.6%. The easiest on-page win is also the most commonly skipped.
- Median alt attribute coverage was 60% of images per page. Write alt text while you write the post, not in a cleanup sprint six months later.
- Median title length was 77 characters on desktop, so long titles get truncated in results.
- Only 2.13% of desktop sites served a valid llms.txt file. Treat it as an experiment, not a requirement; Google has said it does not use it.
A monthly cadence a small team can survive
Here is the cadence we actually recommend to a small team with no dedicated writer. One substantial page a month, 1,200 to 2,000 words, answering one of the ten questions properly. One shorter update that improves a page you already published. One email to your list. That is it, and it is more than most competitors sustain past month four.
- Week one: pick the question, interview whoever on your team knows the answer, and collect the real numbers you will quote.
- Week two: draft it. Lead with the answer in the first paragraph; the explanation comes after.
- Week three: check every claim against a source you can link to, add the table or chart that saves the reader a calculation, then publish.
- Week four: send it to your list, post it once on the channel where your buyers actually are, and update one older page that now deserves a link to it.
Updating old pages is the most underrated line in that list. A page published two years ago with current figures and a clearer answer usually outperforms a new post on the same topic, because it already has history. Refreshing beats republishing.
Measuring it without lying to yourself
Pageviews are a vanity metric in isolation and everyone knows it, which has not stopped them appearing at the top of most monthly reports. The 2026 CMI research shows where better signals come from: 77% of B2B marketers collect first-party data through direct customer engagement such as subscriptions and communities, 68% through gated assets and webinars, 63% through CRM and sales interactions, and 52% through behavioural site data.
| What you measure | Why it is worth the effort | Reported by |
|---|---|---|
| Leads and pipeline influence | Connects a page to revenue, however roughly | 63% use business impact to judge thought leadership |
| Audience engagement | Fast feedback on whether the subject was right | 80% track views, downloads and shares |
| Improved targeting from first-party data | The most commonly reported payoff of collecting your own data | 52% of marketers |
| Customer insight | Changes what you write next, not just how you report it | 44% of marketers |
| Conversion rate or ROI improvement | The claim everyone wants and fewer can show | 26% of marketers |
One caution from the same study: 52% of marketers rate their data governance as established or better, which leaves 48% still exploring or developing it. Collecting more data without agreed standards for quality, access and lawful basis is not a measurement programme, it is a liability. If you are weighing content against buying traffic while you build all this, SEO vs Paid Ads: Where Should You Invest? has the numbers for the comparison, and our digital marketing service runs the whole cycle remotely for clients who would rather not staff it.
Frequently asked questions
How long before content marketing produces enquiries?
Plan on two to three quarters before organic search sends meaningful traffic to new pages, and longer in competitive categories. Email and direct sharing pay out faster because you are not waiting to be indexed and ranked. The honest framing is that content is a compounding asset with a slow start, not a campaign with a launch date.
Should we use AI to write the posts?
Use it to draft, outline and tidy, and keep a person responsible for accuracy and point of view. In the 2026 CMI research, 87% of marketers using AI for content creation reported better productivity, but the report’s conclusion is that fundamentals decide outcomes. A fast draft of a page nobody needed is still a page nobody needed.
How many words should a post be?
Long enough to answer the question and no longer. We aim for 1,200 to 2,000 words on decision-stage topics because those readers want the trade-offs and the numbers, and 400 to 600 for a narrow how-to. Length is an output of the brief, not a target.
Is blogging dead now that AI answers questions directly?
No, but shallow blogging is. Google’s documentation says there are no special optimisations for AI Overviews and that normal SEO practice still applies, which means the pages worth citing are the ones with original detail, real figures and first-hand experience. Generic summaries were always the weakest format and now they are also the easiest to replace.
What should we do with old posts that get no traffic?
Audit them once a year. Merge the thin ones into a better page and redirect, update the ones that are nearly right with current figures, and delete the handful that were never about your business. A smaller library of strong pages is easier to maintain and easier for search engines to understand.
If you would like a content plan built from your own sales questions rather than a keyword tool, send us the ten questions you hear most and we will draft the brief. Get in touch with Eudora Technology to talk about your project.
.jpg){kind=link}
.jpg){kind=link}
