Your B2B blog's most loyal readers are not buying anything

The people who read your company blog every week are not going to buy anything. They are your competitors, the candidates you are about to interview, and your own sales team. The buyer you built the blog for shows up once, late, reads perhaps three pages and leaves. That asymmetry, rather than your traffic number, is what should decide what the archive contains.

Your analytics cannot tell you who anyone is

Every reporting tool you have shows sessions, sources, and time on page. None will tell you that Tuesday morning's long read of your pricing explainer was a product manager at your closest competitor, or a candidate deciding whether to reply to your recruiter. The numbers are real. The identities are inferred, and mostly inferred wrongly.

What fills that gap is a funnel diagram. Strangers arrive at the top, get nurtured through the middle, convert at the bottom. It is a tidy model that describes some businesses well. For a company with a few hundred realistic customers in its whole market, it misdescribes almost everyone actually reading.

What we can say is that the reader most archives are built for — the curious stranger at the top of a funnel — is the one least likely to be there.

The buyer arrives late and reads a handful of pages

By the time a serious buyer reads your blog, they know they have the problem, know roughly what a solution looks like, and have a shortlist with your name on it. They are not being educated. They are checking whether you are credible, and more precisely whether you have already thought about the thing worrying them.

That reader does not work through your archive. They read the post that matches their worry, glance at one or two others to see whether the first was a fluke, and judge the company on that very small sample. Three pages is a realistic budget. Depth on the subject they came for beats breadth across the category.

This is why an archive of competent introductory explainers underperforms so reliably. Introductions are written for a reader at the beginning. Your buyer is near the end, and can tell the difference in a paragraph.

The regulars are competitors, candidates and your own team

Competitors read you closely and consistently, and that is not a problem to be solved. A blog that gives away nothing to a competitor gives away nothing to a buyer either, because both read the same paragraphs for different reasons. The only real constraint is not publishing your roadmap.

Candidates read you before an interview. For a small company the blog is the most substantial public artefact of how the people there think, and a technical candidate who reads three posts learns more about your engineering culture than the careers page ever will.

Your own sales team is the audience with the fastest measurable return, which is the case we made in why the first year is sales enablement. A salesperson who sends a post after a call has answered an objection without writing the email, and told you which objection keeps coming up.

Google's first question assumes a reader who did not search

Google's guidance on creating helpful, reliable, people-first content opens its self-assessment by asking whether you have an audience that would find the content useful "if they came directly to you".

That is an odd question to lead with if search traffic were the whole point, and it is the most useful test on the page. It asks whether the post would still be worth reading to someone who arrived without a query: a customer following a salesperson's link, a candidate reading the night before an interview, a reader who already knows you.

A post that only makes sense as an answer to a search query fails that test. A post written for someone who came directly tends to pass both. It is the argument from our cornerstone piece on why Google doesn't penalize AI content: the durable question is whether anybody needed the page, not how it was produced.

What this argument does not license

It does not mean search traffic is worthless. Non-brand impressions remain the cheapest evidence that your archive covers subjects people care about, and a post nobody can find can only ever be sent by hand.

It does not mean writing nothing but bottom-of-funnel comparison pages. An archive of pure sales collateral reads as exactly that, and the competitor and candidate audiences will stop reading.

And it does not mean we know your ratio. The readership of a small B2B blog is largely unobservable, the funnel model is a guess, and ours is a better-fitting guess rather than a measurement. Nobody can see your mix from outside, including us. The test available to you is cheap. Ask your salespeople which posts they send, and your last three hires which ones they read.

What changes if you accept this

You stop judging a post by whether it could attract a stranger, and start asking whether a salesperson would send it, a candidate would be impressed by it, and a competitor would find it annoyingly good. Those three tests are harder to pass and much harder to fake.

The brief changes before the archive does. Subjects stop coming from keyword volume and start coming from the objections your sales calls surface and the positions you are willing to defend. The archive gets narrower and deeper, the opposite of what a traffic target encourages.

And the traffic number stops being the headline. It becomes one input among several, none of which can tell you who anyone was. That was true the whole time.