kwerve
fr en

For freelance SEOs and small agencies

Be the answer AI gives.

Your clients are asking whether they show up in AI assistants. Here is your answer.

kwerve measures your brands' visibility every week on AI tools such as ChatGPT, Claude and Perplexity, and hands you a white-label report under your logo, ready to send to the client.

A report you re-bill to your client, every month.

15 places at €79/month for life. No card required.

A weekly report produced by kwerve. Fictional client, demonstration figures.

Page 1 of an AI visibility report produced by kwerve

The objection, up front

“Why not just ask ChatGPT yourself?”

Nothing stops you. But once proves nothing: AI answers change from one run to the next, and a single test has nothing to be compared against.

Knowing whether a brand is actually visible means asking the same question dozens of times, across several engines, week after week, and then counting. That is the difference between an anecdote and a measurement.

Nobody tells Semrush users to “just Google the keyword”.
Same logic, new search surface.

The arithmetic

What you will not be doing by hand

10
clients
15
prompts per client
3
engines, every week
1,950
manual checks per month

And that is before recording positions, spotting competitors and laying out the PDF: 30 to 60 hours a month. kwerve does all of it for €99 a month, however many clients you have.

The deliverable

The report is the product.

Every week, a PDF and a shareable link on your agency's letterhead. Forward it as it is, or re-bill it at €50 to €100 per client per month.

Fiduciaire Delacroix · week of 13 July · Rive Digitale Page 2

So what

The engines know the brand, but it is missing from the topics that matter and is almost never cited as a source. Three jobs, in order of return.

Findings & actions

  1. 01

    4 topics out of 12 with no mention at all

    Publish a page for each missing topic. These angles do not require taking a position, only existing.

  2. 02

    3 citations out of 87 lead to the client's site

    Target local.ch, bilan.ch and expert-suisse.ch. Being present there counts for more than any change to the site.

  3. 03

    BDO is cited 71 times, the brand 41

    Take back the topics BDO owns: the engines have made it the category's default answer.

AI visibility report Rive Digitale
The “so what” page: what the numbers imply, and where to start.
Download a sample report (PDF)

Full report, fictional client, demonstration figures.

Page 2 of an AI visibility report produced by kwerve
Share of voice and cited sources
Page 3 of an AI visibility report produced by kwerve
The detail, prompt by prompt

In practice

Three steps, then nothing left to do

  1. 01

    You describe the brand

    Name, domain, market, a few competitors. Two minutes per client, once.

  2. 02

    kwerve queries the engines

    Your prompts, asked several times on ChatGPT, Claude and Perplexity, every week.

  3. 03

    The report goes out under your logo

    PDF and shareable link, in French or English, ready to send to the client.

What we measure

Numbers a manual check cannot give you

42 / 100

AI visibility score

A weighted blend of the first three, 0 to 100. Published formula, never a black box.

Read the methodology

Visibility rate

How many answers name the brand. “4 out of 9”, not “it shows up sometimes”.

Average rank

Where the brand lands when it is named. The decision is made in the first three.

Share of voice

Your mentions against those of the competitors you declared, across every answer.

Cited sources

Which domains the engines lean on to answer, and whether yours is one of them.

▲ 5

Week-on-week movement

An arrow on every number from the second report on. That is what separates an audit from monitoring.

What we measure, and what we don't

Perplexity answers by searching the web: that is where the citations come from. ChatGPT and Claude answer as plain text, from what the model learned. We measure what the engines actually answer, not what your page contains. Google AI Overviews is not tracked: there is no official API, and we do not scrape.

The differences

Why not one of the other tools

Pricing
Per brand tracked: every client you add costs more.
One flat price, unlimited brands. Fair use of 400 checks per brand per month.
Language and market
Built for English and the US market.
French first, local markets (fr-CH, fr-FR, fr-BE), German next.
Deliverable
A dashboard you have to connect the client to.
A white-label report you send, or re-bill.

We name no one: the comparison is about models, not about specific competitors.

Pricing

One price, all your clients

15 founding places
€99 / month

€79 / month for life for the first 15

  • Unlimited brands: 400 checks per brand per month
  • ChatGPT, Claude and Perplexity
  • Weekly white-label report: PDF and shareable link
  • Findings and actions in every report
  • Week-on-week movement
  • Interface and reports in French and English

15 places at €79/month for life. No card required.

Payment isn't open yet. Leave your address: you keep the founding price and you'll hear from us when we open.

Questions

What people ask

Why not just ask ChatGPT myself?

You can, and once proves nothing. Answers vary on every run: what matters is how often a brand is named across dozens of answers, several engines and several weeks. That is a calculation, not a question.

Which engines are tracked?

ChatGPT (OpenAI), Claude (Anthropic) and Perplexity, through their official APIs. Google AI Overviews is not tracked: there is no official API, and we do not scrape results pages.

Does the report carry my logo?

Yes. Your logo and agency name on the letterhead, a neutral layout, no kwerve colour anywhere in the document. It is your deliverable, not ours.

Can I re-bill it to my clients?

That is what it is for. Agencies bill this kind of monitoring at €50 to €100 per client per month; kwerve's price is set so that margin exists from the second client on.

How many brands can I track?

As many as you like, within fair use of 400 checks per brand per month (one check = one prompt sent to one engine). Twenty prompts across three engines every week stays well below that.

Does kwerve guarantee better visibility?

No, and be wary of anyone who promises it. AI answers are not deterministic; nobody controls what ChatGPT will recommend next week. kwerve measures, shows where to act, and proves what moved.

What happens to the measured data?

It is kept. History is what turns an audit into monitoring: week-on-week movement comes from it, and nothing is erased.

Which languages?

French and English at launch, for the interface and the reports alike. German follows.