How to brief Fanzy AI so your first storefront comes back usable

You typed one sentence — something like "I sell digital products for creators" — and half a minute later a whole storefront appeared. Theme, layout, font, a headline, descriptions under each card. It looked professional. It also could have belonged to any of forty thousand other people, and you knew that before you finished reading it. That gap between looks finished and is mine is almost always a briefing problem, and a briefing problem is a cheap one to fix.

Why a capable assistant returns a bland page

An assistant working from the word "coach" has to write something that would not be wrong for any coach alive. The safest sentence that fits everyone is the sentence nobody remembers. This is not a flaw in the tool so much as arithmetic: vague input has to be answered with an average, and averages are forgettable by design.

The second reason is that you have stopped noticing the things that make your work particular. You know that most of your buyers arrive on a Sunday evening, three days before a deadline, already slightly panicked. You know they always ask the same two questions first. None of that reaches the draft unless you say it out loud, because it feels too obvious to mention.

And here is the part that is genuinely hard, rather than just fiddly: some people cannot brief well because they have not yet decided what they sell. If you are still choosing between three directions, no assistant can pick one for you, and the page will wobble in exactly the way your thinking does. Deciding first is not a shortcut you can skip. The Fanzy page describes what the assistant does; it does not do the deciding.

The usual way of getting a storefront up, and the briefed way

The difference is not speed. It is the order of the decisions. Most people choose a look before they have written a word, then bend the words to fit the look. Briefing properly reverses that: the sentences exist first, and the design is judged against real language rather than placeholder text. Everything the assistant sets stays an ordinary control in the storefront designer, so none of it is a one-way door.

The six lines

Write these in a notes app before you open anything. Five minutes here saves the evening you would otherwise spend nudging a font. Keep each line short and factual; you are giving the assistant material, not writing marketing copy.

1. 1What you make, at the level of the object. Not "resources for teachers" but "printable phonics worksheets for reception classes, twelve to a pack". If a stranger could picture the file, the line is specific enough.
2. 2Who buys it, in one clause. Newly qualified teachers in their first year, not "educators". Job title plus one qualifier beats any demographic bracket.
3. 3The moment they arrive. This is the line people skip and the line that does the most work. "Sunday night, planning tomorrow, does not want to design anything" tells the assistant what tone to hold and what to put first.
4. 4The tone, in plain adjectives. Calm, blunt, warm, technical, dry. Avoid brand language — "authentic yet aspirational" produces mush, while "plain and slightly dry" produces something you can edit.
5. 5What you actually sell right now, with prices. Two downloads and a call is a real answer. Prices matter because a page holding one item at 8 and one at 300 needs a different shape from a page holding three at 19.
6. 6Anything fixed. A colour you already use everywhere, a name your audience knows you by, a phrase you refuse to drop, a photograph you want at the top. Say it now rather than undoing it later.

Phrases that make the draft worse

• "Digital products" — true of roughly everything, so it tells the assistant nothing.
• "Modern and clean" — the default already is; you have spent a line and bought nothing.
• "Passionate about helping people" — this reliably comes back to you as filler in the headline.
• "Multiple income streams" — describes your ambition, not the page a buyer will read.
• "Coming soon" items — briefing products you have not built pushes the layout towards a catalogue you cannot fill.

A worked example: the same person, briefed twice

Say you edit podcasts. Brief one reads "audio editor, professional, sells services and templates". What comes back is competent and interchangeable: a neutral typeface, an even grid, a headline about quality sound, three descriptions that could be swapped between any three audio people without anyone noticing.

Brief two reads: "I edit interview podcasts for solo hosts who record on one microphone in a spare room. They come to me after episode six, when they have realised editing is the reason they have stopped publishing. I sell a 19 noise-cleanup preset pack, a 45 one-episode edit, and a 120 monthly slot for four episodes. Tone: plain, slightly dry, no studio jargon." The draft that returns has somewhere to put its feet — it leads with the monthly slot because that is the offer the described buyer needs, it holds the three prices in an order that makes the middle one look sensible, and the descriptions mention a spare room rather than a studio. You will still rewrite half of it. Rewriting half is a much smaller job than writing all of it.

The mistake most people make

The common mistake is briefing the business you intend to have rather than the one you have today. People describe the membership they will launch in the autumn, the course they are half-recording, the community they hope to open. The assistant designs for all of it, and you end up with a page built to hold nine things while two exist — which reads, to a first-time visitor, as a shop that has been abandoned. Brief today's catalogue, launch it, and rebrief in the autumn when the membership is real. Adding products later is a five-minute job in the dashboard, and the features list will still be there when you need the rest of it.

A brief is not paperwork. It is the only moment in the process where being specific is free, and everything downstream — the theme, the order of the cards, the sentence under your name — inherits whatever precision you put into it. Six lines, five minutes, and the page that comes back is one you can argue with instead of one you have to invent.