How do you write a content brief when an AI writes your blog posts?

In short: A content brief is the two or three sentences that tell the model what you expect in every article: content type, scope, reader and the sections each article must have. The brief drives both topic selection and the shape of the article, so a vague brief means off-topic posts. In Wiozen Content Bot the brief is set per category; tone is a separate free-text field, length is chosen between 300 and 2,500 words, and an article that reaches less than 60 percent of the target isn't saved.

Which of the bot's decisions does the brief affect?

In the Content Bot the brief is the What kind of content should it be? field on the Categories tab, and it's used in four places. Candidates from RSS are matched against it to decide which fit the category. When the sources yield no suitable topic, the bot suggests one itself based on the brief. Find sources automatically searches for sites that match it. And the writing prompt includes it with the note that content type and scope must follow it.

So the brief is written once but read on every run. A category can be saved with a brief or with at least one RSS address; if both are empty, it isn't saved and you get "Add a brief or at least one RSS source; the category was not saved."

What goes into a good brief?

A weak brief looks like this: "Articles about loans." From that the model may produce news, rate commentary or bank promotion. A brief that works for the same category:

Practical questions about mortgages, car loans and personal loans: eligibility, fees, early repayment and restructuring. The reader is taking out a loan for the first time. Every article includes a worked example, a step-by-step application and a common mistakes section; no investment advice.

• Name the reader. "Someone taking their first loan" and "a loan professional" lead to very different articles.
• List the subtopics. When every topic the bot suggests already exists on the site, it asks again and expects the model to split the category into narrower subtopics. Subtopics written in the brief make that split more accurate.
• Spell out the sections you want every time: a worked example, a table, an ingredient list, a comparison. The model can use tables and numbered lists in the body.
• Say what you don't want as well. Limits such as no advice, no price predictions or no brand praise go into every article if they're in the brief.

Keep tone and length out of the brief; each has its own field and goes into the writing prompt as a separate line.

Can the bot write the brief for me?

Two ways. When you pick a category from the list that hasn't been set up yet, the bot fills in the brief itself: the model writes a 2–3 sentence brief from the category name, its WordPress description and the titles of the 10 newest published posts in it, then looks for matching sources. Filling in the category description in WordPress therefore makes that first brief noticeably better.

The second is the Write with AI button. It asks 5 questions specific to the selected category in 3 steps: Content (2 questions: content type, scope and subtopics), Reader and tone (2 questions) and Article structure (1 question: the sections every article should have). Each question has 4 ready answers plus a field for your own. Your answers are turned into a brief. The questions and the brief are two separate model requests billed to your balance, and the button doesn't work until an API key is connected.

The bot writes the brief in the language set on the Writing tab. Whichever way it's written, read it before saving and add what's specific to your site; the model doesn't know your readers as well as you do.

What should go in the tone field?

Tone is a free-text field per category. On an English site the default is "informative, plain and trustworthy". Two or three adjectives such as "friendly and practical" or "formal and measured" are enough; there's no need for long instructions.

Whatever the tone, the writing prompt has fixed rules: stock phrases like "in today's world" or "in conclusion" are banned, sentence length has to vary, sections shouldn't end on a summary line, and no invented personal experience or imaginary experts. Writing "write as if you tried it yourself" in the tone field doesn't override that last rule; real experience is something you add to the article yourself.

How should I choose the length?

Length is set between 300 and 2,500 words in steps of 50; the default is 900. The bot asks the model to use the focus keyword at least 5 times in the body, roughly once every 90 words on longer articles (10 times at 900 words).

Models tend to write under the target. In one measurement with the default model, a 900-word target produced articles of about 660 words on average. The bot sets a floor for this: if an article comes in under 60 percent of the target (540 words at 900) or has fewer than two H2 headings, it isn't saved. The log says "Article did not pass the quality check (… words, target …); topic skipped.", the topic is marked so it isn't retried, and the bot moves to a reserve topic for that category.

A higher target makes each article take longer and cost more. With a very long target the model's answer can hit its output limit; the log then shows "The response was cut off at the model's output limit; lower the article length or choose another model." Rank Math gives its full length score at 2,500 words, but for a recipe or a short how-to, 900–1,200 words is usually enough.

How do I test a brief?

1. Leave a single category active, set articles per run to 1 and publishing to draft.
2. Start a run with Run now at the top.
3. When it finishes, check the Log tab for which topic was picked and which were skipped and why, then read the draft.
4. If the topic drifted, narrow the scope in the brief; if the structure is missing something, name the required sections outright. A few runs are usually enough to settle it.

How topics are chosen from sources is in the RSS and Google Trends guide, and the full setup in the auto-publishing guide.

Frequently asked questions

Download the Content Bot, write a brief for one category and run the first round as drafts. The log shows on the first run whether the brief works.

Create your account

More guides: How can you use an AI API from Turkey without a credit card? · How do you connect an OpenAI-compatible API to a tool you already use? · How do you calculate the cost of an AI API? · What should I do if my AI API key has leaked? · How do you generate images with an AI API, and what does one image cost? · Why does my AI API return a 429 error, and how do I fix it? · Why do long AI API requests time out: streaming or non-streaming? · Anthropic Messages API or OpenAI Chat Completions: which one should I use? · What is a context window, and what is a 1 million token context good for? · Which AI model should I choose: deciding by price, context and quality · Why does a reasoning model return an empty answer, and why are thinking tokens charged? · Why is my API balance running out faster than expected? Tracking spend request by request · How do you convert a web page to Markdown to feed it to an AI model? · How do you get AI to write and publish articles in WordPress automatically? · How do you find blog topics from RSS feeds and Google Trends? · Does AI-written content show up on Google and in AI answers? · How do you get internal links into AI-written WordPress articles? · How do you get AI to generate a different featured image for every WordPress post? · What should you watch for when an AI writes recipes for your food blog? · How do you fix errors when connecting a WordPress plugin to your AI API account? · Why does my AI article bot stop halfway in WordPress? "Run is stuck" and LiteSpeed · Why are my scheduled WordPress posts late? WP-Cron and a real server cron · How do I get an AI article bot to write blog posts in English or another language? · What does "Content Bot cannot generate articles" mean? Reading the bot's log