The first ninety days are supposed to look like nothing

Ninety days into a new blog, the traffic graph is nearly flat. We say this in sales calls, before the contract, because the alternative is a customer who cancels in month four for hitting precisely the outcome we predicted. The first quarter does not buy traffic. It buys a pipeline that works and an archive with something in it, which is what traffic is made of later.

Month one is a plumbing month

The first thirty days go on things nobody puts in a case study. The engine reads your site and builds a style profile. Someone argues about whether the third post's central claim is true. A hero image renders badly on a phone and the template gets fixed. The approval loop meets a real week for the first time, including the week the approver is in Lapland.

None of that is visible from outside. From outside, month one is five or six posts on a domain that had none, which looks identical to a blog that will be dead by spring.

It matters anyway, because each of those small mechanical decisions is load-bearing for the eleven months after it. We have argued before that the brief is the product, and month one is when the brief stops being a document and becomes observable behaviour. A subject list that looked complete in a workshop meets its first real post and turns out to contain three subjects you cannot yet say anything true about.

Google's own timelines do not fit inside a quarter

This is the part most vendors leave out, and Google documents it in public.

Start with crawling, which is the cheapest step and still not instant. Google's guidance on asking for a recrawl says that crawling "can take anywhere from a few days to a few weeks", and that requesting one "does not guarantee that inclusion in search results will happen instantly or even at all". That is the front of the funnel, before ranking is even a question.

Then the slower clock. Google's guidance on core updates describes how long recognition takes once a site improves: some changes take effect in a few days, but "it could take several months for our systems to learn and confirm that the site as a whole is now producing helpful, reliable, people-first content in the long term". If nothing has moved by then, the same page's advice is to keep waiting — "if it's been a few months and you still haven't seen any effect, that could mean waiting until the next core update."

Read that phrase again. The site as a whole. Not the post. This is the same mechanism we described in Google doesn't penalize AI content: the assessment is of a body of work. On day ninety you do not have a body of work. You have a beginning.

Flat traffic is not the same as nothing happening

The difference is measurable, provided you look at the right four things.

Coverage comes first. At ninety days you should be able to hold up the list of questions your buyers actually ask and point to which ones now have a page. That is a fact about your archive, not a forecast. Indexed pages come second, and not as a vanity count — as a yes-or-no on whether the plumbing works. If forty posts are live and four are indexed, you have a technical fault you would otherwise have discovered in month nine.

Then the two that predict everything downstream. Are the posts accumulating links to each other, or is each one an island? And has anyone inside your company used one yet — sent it to a prospect, pasted it into a reply, cited it on a call? That last signal is soft, unglamorous and the best early indicator we know, because it means the archive has started being useful to the people closest to the buyer.

Our four reportable numbers are the right steady-state dashboard. In the first quarter most of them are zero, and reporting zeroes every month is how a sound programme gets killed early.

If you need pipeline this quarter, do not buy a blog

Here is where we argue against our own product.

If the board wants qualified pipeline by December and it is September, a blog is the wrong instrument and we would rather say so now. Paid search puts you in front of the same query next week. A webinar produces named humans faster. Content published today runs on a clock that is indifferent to your quarter, and volume does not compress it. Forty posts in month one are not assessed forty times faster.

The right time to start is when you can afford a quiet first quarter. That is usually sooner than people think, because it is a fixed, one-time cost. You pay it once. You never pay it again in year two — and a company that keeps deferring the start because this quarter is busy pays it forever, which is why in-house blogging stalls in the first place.

Judge the machine, not the graph

At ninety days, ask whether the machine works. Are posts going out on schedule without anyone chasing them? Is the voice right, or at least wrong in shrinking increments? Does the archive cover the questions you genuinely get asked? Has anyone internally reached for a post without being told to?

If those are yes, the ninety days did their job and the traffic conversation belongs in month nine, once publishing rhythm has had time to compound. If they are no, you have learned that for the price of one quarter, which is the cheapest that lesson is ever offered.