Content and brand
Writing for two readers at once
A person needs the text to convince. A model needs to lift a whole paragraph without losing the meaning. Both can be satisfied, and there is a concrete way to do it.
- Paragraphs that stand alone, so quoting one does not misrepresent you
- Your brand voice loaded as a constraint, not as a suggestion
- Several versions per document, comparable against each other
The problem
- Text a model generates is noticeable, and your buyer spots it before you do.
- You optimise to be cited by a model and end up with a page nobody wants to read.
- Everyone on the team uses a different tone and the site sounds like four companies.
If the text does not convince a person, being cited by a model fixes nothing.
How it works
It is not about stuffing keywords: it is about each important claim fitting whole into one paragraph.
- 1
You load your voice
Examples of your own text that works, words you never use and the register that suits your buyer. Without this, what comes out is the average of the internet.
- 2
It writes in self-contained paragraphs
A model that cites lifts a fragment and drops it into another context. If your paragraph starts with this also applies to, quoting it loses what this is and your claim comes out deformed. Writing self-contained is self-defence.
- 3
Versions get compared
Several versions of the same text, stored, so you choose with judgement instead of keeping the first. The previous version is never lost: going back is one click.
What you get
Text you can publish, not a draft that must be redone entirely.
- The text at the length you asked for, respecting the brief's structure.
- The stored versions, with date and who requested them.
- Marked gaps where a figure of yours is needed that the model cannot invent.
Measure before you pay anything
The scanner runs 10 real responses in about 60 seconds. No signup, no card. If the number it returns tells you nothing, do not buy.
Measure my brandWho it is for
- Small teams that publish a lot and have no in-house writer.
- Anyone with a clear structure who needs the first draft fast.
- Shops with many product pages where tone gets lost across hundreds of items.
It is not for you if...
It is not for you if you want to publish without reading. Text generated and published with no human review ends up asserting things you cannot support, and in a measurement product that is expensive.
An honest comparison
| By hand | With Citenza | |
|---|---|---|
| What it costs you | Pasting the brief into any chatbot and copying what comes out. | The brief and the text live in the same place. |
| The method | No brand voice: out comes the average of the internet, and it shows. | Brand voice is a constraint applied every time. |
| Repeating it next month | The previous version was lost when the tab closed. | Every version stored, with one-click rollback. |
Availability by plan
Available from Starter, with the number of documents and versions growing by plan. Cost depends on the length generated, and the estimate is shown before launching as everywhere else in the product.
Honest questions
- Is it obvious a machine wrote it?
- Less so if you load your voice, and it still needs reviewing. What gives generated text away is not the vocabulary: it is that it asserts nothing concrete and takes no position. The marked gaps are precisely so you insert the figures that make the text say something.
- Can it invent figures?
- It is forbidden from doing so and marks the gap instead. When a sentence calls for a figure, a flagged gap appears rather than a believable number. It is more awkward to fill in and it is the only way the text does not assert things you cannot support.
- Does Google penalise generated content?
- Google says it penalises worthless content, generated or not. In practice, what sinks a page is adding nothing that is not already in ten others. Generating the draft is not the problem; publishing without adding anything of yours is.
- Does it write in all five languages?
- Yes, and it writes in the target language rather than translating from Spanish. The difference shows: a translated text keeps the source's sentence structure and sounds imported, which is exactly the opposite of what you want in a new market.
Write to be read and to be cited
And check whether it works: measure before and after with the same question set.