Write posts people can actually find

There are two ways to write a post nobody finds. One is to write something nobody wants. The other is to write something good and never decide what question it answers.

The second is more common and much easier to fix.

Decide the question before the topic

A topic is "onboarding". A question is "how long should onboarding take for a new marketing hire". People type questions.

I want to write about how we onboard new client accounts.

Give me eight questions a potential client might actually type that this
post could answer. Rank them by how close the person asking is to needing
us.

For each, say in one line what a post answering it would have to contain to
be worth reading.

That last column is what stops you picking a question you cannot answer well.

Write the metadata first, then the post

Doing it the other way round is why so many posts have a title that describes the topic and a meta description written in a hurry afterwards.

For the question we picked, give me:
  a title under 60 characters that contains the phrase someone would type
  a meta description under 155 characters that says what they will learn,
    not what the post is about
  the three subheadings the post needs to actually answer it

Do not write the post yet.

"What they will learn, not what the post is about" is the line that earns its place. "In this post we explore our approach to onboarding" describes a document. "How long onboarding takes, what the client has to provide, and what happens if they are slow" tells someone whether to click. The second is harder to write and it is the whole job of a meta description.

Then write it, against the outline

Write the post against those subheadings, in our brand voice.

Answer the question in the first paragraph. Do not build up to it.

Use specifics from our own work rather than general advice. If you do not
have a specific, say so and leave a note rather than inventing an example
or a statistic.

What this playbook does not claim

It does not tell you what people search for in what volume. That needs a keyword tool, and nothing here has one connected. What you are getting is a post that clearly answers one specific question, with metadata that matches it.

That is most of what makes a post findable, and it is the part usually skipped. If you have search volume data, bring it to the first step and pick from evidence rather than judgement.

What good looks like

Every post has one question it answers, a title containing the words someone would type, and a description that would make you click.

The measure is not traffic in week one. It is whether, six months later, you can say what each post is for.