Marketing

Product text, posts and translations are always last in the queue

Drafts in your own voice, built from what you already know about the product, in both languages.

Who this is for Online shops, makers and service businesses that have to write regularly, and in two languages.

The problem

Descriptions and posts there is no time for

New products sit without text, because writing takes an hour and there is no hour. The social accounts go quiet for weeks, then three posts land in a day. The Lithuanian and English versions drift apart, because the translation happens whenever somebody remembers. So some products do not sell because nothing was written about them, not because they are bad.

You will know it by

Products go up with one line of text, or none.
The second language lags behind the first.
Posts come in bursts rather than steadily.
The text in different places reads like different companies wrote it.

Possible solutions

Not every road leads to me

Writing it yourself to a pattern
One clear skeleton for a description that you fill in within minutes.
Where it stands Often enough, and it produces the best text there is. The limit is that across fifty products it is still a day's work.
Hiring a copywriter
A person who writes your text regularly.
Where it stands The best choice for the text that matters most: the front page, a campaign. For every single product it rarely pays.
Machine translation
Running finished text through a translation tool.
Where it stands Fast and often good enough, provided somebody reads the result. An unread translation on a public page is a risk rather than a saving.
Drafts from your own data and voice
From what you already know about the product, a description, a post and the second language version.
Where it stands This is the one I build. Worth it when there is a lot to write, regularly, and the voice has to stay the same.

What I set up

From what you already have, the attributes, the size, the material, what it is for, a draft description is written in your voice and your structure, in both languages at once. The same source gives a shorter version for social and a sentence for an email. The voice is described once from your own best existing text, so new text sounds like the old text rather than like a tool. You read and approve, and your people publish.

How it goes in

We pick a few pieces of text you actually like and describe the voice from them.
We agree the structure: what a description always says and what it never claims.
Drafts are produced in batches, and you approve them from a list.
The second language version is read by a person before it goes out.

The limit Product text may claim only what is true. Specifications, materials, origin and any health or environmental claim come from your data rather than from a model, because it is the seller who answers for a false claim.

The work The first part, describing the voice, happens once. After that it becomes routine work, sized by your volumes.

The prices on this site are final.

What it brings

Products go live with text

The description stops being the reason a product waits a week.

Both languages at once

The second language stops being a debt you come back to later.

One voice everywhere

The site, the shop and the social accounts sound like the same company.

How you check it yourself

Count how many products have no real description today, and how long a new product takes to reach a page with text on it. The second number is the one worth watching.

Is this your job?

Tell me how this runs at your place now. On a 30 minute call I will say whether it is worth automating, where I would start and what it needs.

Book a 30 minute call

The call is free, with no obligation and nothing to sign.

Not sure which job is first for you? The free assessment says where the hours go. Start the assessment

Common questions

Do search engines penalise AI text?

Google states publicly that it judges whether content is useful rather than how it was produced, and that it acts against mass low-value pages. That is why this text is built from your own product data rather than generic phrasing, and why a person reads it before it is published.

Do I have to disclose that AI wrote it?

For a product description there is no such requirement, provided the text is accurate and you stand behind it. Disclosure is required elsewhere, for instance where a generated image or a chatbot could be mistaken for the real thing. That is how I read the rules rather than legal advice; if a decision turns on it, check with someone who gives that.