How to Plan a Website Instead of Writing Random Blog Posts (A 6-Step Map)

How Random Posting Usually Happens

Very few people set out to write random blog posts. It happens gradually. A customer asks a good question, so it becomes a post. A competitor publishes something, so you write your own version. A keyword tool suggests a phrase with decent volume, so that becomes a post too. Each decision is reasonable on its own.

A year later, the blog has a few dozen articles that are each fine, but they do not obviously belong together. Some cover nearly the same question. Some never link to anything. Your service pages, the pages that actually describe what you sell, sit off to the side with no posts pointing to them. Nothing is broken, and yet the site does not feel like it is building toward anything.

That pattern is usually not a writing problem. The posts were written before anyone decided where they belonged. The fix is not to write better posts. It is to decide where posts belong before writing them.

What a Website Plan Actually Looks Like

A website plan does not need to be elaborate. For most small sites, it is a spreadsheet with a handful of columns:

• Hub page: the main page for a topic, often a service page or a core guide.
• Post: a supporting article that sits under that hub.
• One job: the single thing that post does.
• Links: which pages it links to, and which pages link back.
• Status: idea, planned, written, published, or needs merging.

That is it. The value is not in the format. It is in forcing every post to answer "where do you live?" before it exists. The six steps below fill in that spreadsheet in order. If you want the background on why this kind of structure matters for AI search in the first place, why website architecture matters more than any single article covers that side. This guide focuses on the doing.

Step 1: Decide What the Site Needs to Be Known For

Start from the business, not from the idea list. Write down the few things you want someone, or an AI assistant, to associate with your site after reading it. A practical way to get there is to list:

• Each distinct thing you sell or do.
• The main problem customers come to you with.
• Any audience or situation that needs its own explanation.

For many small businesses, this lands somewhere between three and six core topics. To make it concrete, consider a hypothetical small bookkeeping firm. Its core topics might be monthly bookkeeping, payroll, tax preparation for small businesses, and "getting your books in order before you hire an accountant." That last one is a customer problem rather than a service, and it is often where the most useful posts come from.

If your list grows past what you could realistically keep current, it is probably too broad. A shorter list you can actually support usually works better than a long one that stays half finished.

Step 2: Give Each Topic One Hub Page

For each core topic, choose one page that will act as the main answer. This is the hub. Every post on that topic will point back to it.

Often the hub already exists: it is your service page for that offering. Sometimes it is a longer guide you have already written. If nothing suitable exists, the hub is the first thing to create, before any supporting posts. Writing five posts about payroll without a payroll page for them to point to is one of the most common ways a blog ends up feeling scattered.

Two practical rules help here. First, one hub per topic. If two pages could each claim to be the main payroll page, pick one and let the other support it or merge into it. Second, the hub should describe the topic plainly enough that a newcomer understands it without reading anything else. Supporting posts go deeper. The hub sets the frame.

Step 3: Sort Your Post Ideas Under the Hubs

Now take every post idea you have, from your notes, customer emails, keyword research, or the drafts folder, and rewrite each one as the question a reader is actually asking. "Payroll tips" becomes "What do I need before running payroll for my first employee?" Questions are easier to sort and easier to check for overlap.

Then place each question under one hub. Three things tend to happen:

1. Most ideas fit cleanly. These go into the plan.
2. Some ideas fit under two hubs. Pick the one where the reader most likely is when they ask it. A post can still link to the second hub, but it lives under one.
3. Some ideas fit nowhere. These get parked. They might be interesting, but if they do not connect to anything you want to be known for, publishing them mostly adds noise. If a parked idea keeps coming up, it may be a sign you are missing a core topic, which is worth revisiting in Step 1.

This step is where most of the random posting stops. Ideas still come from the same places. They just have to find a home before they get written.

Step 4: Give Every Planned Post One Job

Two posts can sit under the same hub and still compete with each other if they answer the same question. To avoid that, give each planned post one job. A short list of job types covers most posts:

• Diagnose: help the reader figure out what went wrong.
• Explain: show how something works.
• Walk through: give a step-by-step process.
• Check: offer a short list of things to verify.
• Interpret: help the reader read a number, report, or signal.

Then look across each hub. If two posts share the same question and the same job, they are really one post. Merge them in the plan now, before either is written. That is much easier than untangling two published pages later, which is its own process covered in how to fix content cannibalization properly.

Same topic with different jobs is fine. In the bookkeeping example, "Why does my payroll keep coming out wrong?" (diagnose) and "How to run payroll for the first time, step by step" (walk through) can both sit under the payroll hub without stepping on each other.

Step 5: Decide the Links Before You Write

Internal links are usually added at the end, if at all. In a planned site, they are decided first, because they are what turn separate posts into a connected section of the site. For each planned post, fill in three kinds of links:

• Up: a link to its hub page, using words that describe the hub's topic plainly.
• Sideways: one or two links to sibling posts under the same hub that a reader would reasonably want next.
• Next step: where the reader should go when they are ready to act. That might be a service page, a contact page, or a tool.

Then add the reverse: note which existing pages should link to the new post once it is live, and add those links on publish day rather than "later." A post that nothing links to is easy for both readers and search systems to miss. What AI search learns from your internal links goes deeper on why the direction and wording of these links can matter.

A useful side effect: deciding links first often sharpens the post itself. If you cannot name a natural next step for the reader, the post may not have a clear enough purpose yet.

Step 6: Bring Your Existing Posts Into the Plan

Most people reading this already have a blog full of posts written without a plan. That is normal, and it does not mean starting over. Go through existing posts and sort each into one of three groups:

1. Attach. The post fits under a hub. Add the up, sideways, and next-step links from Step 5, and add a link to it from the hub if it deserves one.
2. Merge. The post answers the same question with the same job as another page. Fold the useful parts into the stronger page and redirect the weaker one.
3. Retire. The post fits under no hub and brings little traffic or interest. Consider redirecting it to the closest relevant page or removing it.

This does not need to happen in one weekend. Doing a few posts at a time, starting with the ones that get the most visits, is usually more sustainable than a big cleanup project. If you are unsure how many pages the site should end up with once this is done, how many pages your website actually needs offers a simple way to think about that.

A Four-Question Check Before Any New Post

These four questions take a couple of minutes. Asked consistently, they are often enough to keep a blog from drifting back into random posting.

Where Tools Fit In

The planning decisions in this guide can be made with a spreadsheet and some honest judgment. Tools help mostly at two points.

Before Step 6, it helps to see how your existing pages currently relate to each other: which ones are connected, which ones sit isolated, and where pages overlap. Doing that by hand gets slow once a site has more than a few dozen pages. Website Architecture for AI Search looks at the whole site at once for exactly that picture, including how AI is likely to understand the business as a whole, so the cleanup starts from what is actually there.

After each post is written, the question shifts back to the single page: is this post clear, well structured, and easy for AI search to read? The Free AI SEO Checker handles that page-level check with step-by-step fixes. If you are not sure which tool fits the stage you are at, the WebTrek tools overview lays out what each one is for. For the broader picture of how strategy and structure come together for AI search, the AI SEO guide is a good place to start.

FAQ

How do I plan blog posts for a website instead of writing them randomly?

Start from what your business needs to be known for, not from a list of post ideas. Pick a small set of core topics, give each one a hub page that serves as the main answer, then sort every post idea under one of those hubs. Before writing any post, give it one job and decide which pages it will link to. Ideas that do not fit under any hub get parked instead of published.

How many core topics should a small business website have?

There is no fixed number. A useful starting point is one core topic per distinct thing you sell or do, plus one for the main problem your customers come to you with. For many small businesses that lands somewhere between three and six. If you list more than you could realistically keep up to date, the list is probably too broad.

What should I do with blog posts I already published without a plan?

Sort each existing post into one of three groups. Posts that clearly fit under a hub get attached by linking them to that hub and to related posts. Posts that cover the same question as another page get merged into the stronger one. Posts that fit nowhere and bring no visits or leads are candidates to retire or redirect. Most sites can do this gradually, a few posts at a time.

Does planning a website really matter for AI search?

It tends to help. AI assistants form a picture of a business by reading across several of its pages. When posts are connected to clear hub pages and do not overlap, that picture is usually easier to form. When posts are scattered and repeat each other, it can be harder for any single page to stand out as the clear answer.

Do I need special software to plan a website structure?

No. A simple spreadsheet with columns for hub page, post, its one job, planned links, and status is enough for most small sites. Tools can help you see how your existing pages currently connect, but the planning decisions themselves can be made with a spreadsheet and an honest look at what you already have.

Final Takeaway

Random blog posts are rarely bad posts. They are posts that were written before anyone decided where they belong. Planning the website first does not slow writing down much. It mostly moves a few decisions earlier: which topic, which hub, which job, which links.

If you only take one thing from this guide, make it the four-question check. Ask it before the next post, and the one after that. Over time, the blog stops being a pile of separate ideas and starts reading like one connected site.