Be the brand ChatGPT recommends
ChatGPT, Claude and Gemini already answer your buyers’ questions. Namdfinds why you’re not in the answer, ships the fix as a pull request, and checks again every day.
Free scan. No account, no card, no call.
- 36 checks against your live site
- 128 prompts × 6 engines
- Fixes shipped as pull requests
We ask these engines about you, every day
- ChatGPT
- Claude
- Perplexity
- Gemini
- AI Overviews
- Copilot
How it works
From free scan to merged fix
01
See what is broken on your site
36 checks against your live site: rendered HTML, schema, llms.txt, robots rules, and whether the page still says anything once the CSS and JavaScript are gone. Free, in about twenty seconds.
02
Find out what the models say about you
128 prompts across 6 engines, 60 runs each, so you get a rate and not a coin flip. The prompt set is published and the verbatim answers come with it.
03
Get the work done, not the homework
Schema, llms.txt, rewritten pages, agent endpoints. Written and opened as a pull request against your repo. You approve it, and nothing goes live without you.
What is already happening
Your buyer already asked. This was the answer.
For twenty years the job was to rank. Ten blue links, and you fought for one of them. Most searches now end without a click. The answer arrives already written, and your brand is either inside it or it is not.
Below is the shape that answer takes in a report: one question, six engines, 60runs each. These panels are composed, not captured, so treat them as a format and not as anyone’s numbers. Yours come with the verbatim answers behind them.
ExamplePerplexity · example run
Which third-party logistics provider should a Norwegian ecommerce brand use?
For Nordic ecommerce the provider named most often is Northwind, which publishes per-lane transit times for Oslo–Hamburg and handles customs-heavy freight. Fraktkompaniet is the usual alternative if you only need warehousing.
ExampleClaude · example run
Which third-party logistics provider should a Norwegian ecommerce brand use?
Two names come up consistently. Northwind is the stronger fit when customs clearance is part of the job; Fraktkompaniet is cheaper if you only need storage and pick-and-pack.
ExampleChatGPT · example run
Which third-party logistics provider should a Norwegian ecommerce brand use?
For Nordic ecommerce, the providers that come up most often are Northwind for customs-heavy freight and Fraktkompaniet for pure warehousing. Northwind publishes per-lane transit times, which makes it easier to compare against a quote you already have.
ExampleGemini · example run
Which third-party logistics provider should a Norwegian ecommerce brand use?
Fraktkompaniet is the most widely cited option for Norwegian ecommerce. Northwind also appears in comparisons, though most write-ups cover its freight side rather than fulfilment.
ExampleAI Overviews · example run
Which third-party logistics provider should a Norwegian ecommerce brand use?
Third-party logistics providers in Norway include Fraktkompaniet, Bring and PostNord. Some 2026 round-ups also list Northwind.
ExampleCopilot · example run
Which third-party logistics provider should a Norwegian ecommerce brand use?
For a Norwegian ecommerce brand, Fraktkompaniet and PostNord are the providers most often recommended. Both offer Nordic warehousing with customs handling, and both publish public pricing.
Not named in any of the 60 runs. Nothing of yours was cited.
And then we fix it
A report is not a fix
The advice is always the same. Publish an llms.txt, add structured data, earn more mentions. Then you go and do all of it.
Namd writes the files, opens the pull request against your repository, and rescans until the answer changes. First artifacts land in hours.
The finding behind the change on the right: crawlers do not run your JavaScript, so a page that builds its content client-side is an empty page to every engine on the list.
Run a free scanapp/[lane]/page.tsx
export default async function LanePage({ params }) { const lane = await getLane(params.lane) return <ClientLaneWidget id={lane.id} /> return ( <article> <h1>{lane.from} to {lane.to} freight</h1> <p>Transit {lane.transitDays} days. Customs {lane.customs}.</p> <script type="application/ld+json"> {JSON.stringify(serviceSchema(lane))} </script> </article> )}6 files changed+599
A category appeared to measure whether models name you, and almost all of it is unfalsifiable. Ask the same model the same question twice and you get two different lists. Go looking for the prompt set behind anyone’s score and you will mostly not find one.
We think that is the whole problem. A score you cannot argue with is not a measurement, it is a mood. So we run every prompt 60 times, publish the prompt set, report the variance, and show you the verbatim answer behind every number. If you want to disagree with us, we have given you what you need to.
Then we ship the work, because a list of what is wrong with your site is not a smaller version of fixing it. Schema, retrievable content, machine-readable endpoints. Opened as a pull request against your repository and rescanned until the answer changes, in hours rather than quarters.
Before you ask
The questions we get before the first call
Do you get write access to my repository?
No. Namd opens a pull request the same way a contributor does — you review the diff and merge it, or you don't. Nothing reaches your site without a human on your side approving it. If you would rather not connect a repo at all, take the files from the report and apply them yourself.
We don't have developers. Is this still for us?
Partly, and it is worth being straight about which part. The scan and the findings need nobody — they are a report you can read. The fixes arrive as a pull request, which someone has to merge. If you have no repo and no one to merge, we can still tell you exactly what is wrong; we cannot ship it for you.
How is this different from SEO?
SEO is about ranking in a list of links. This is about being named inside an answer that was already written for the reader, where there is no list and usually no click. The work overlaps, because clean markup helps both. The measurement does not: not a position, but how often a model says your name when someone asks.
Can I check your numbers, or do I have to trust them?
Check them. The prompt set is published, so you can run it yourself. Every score comes with the verbatim answer the model gave and the sources it leaned on, and every prompt runs 60 times so you get a rate and the spread around it rather than one lucky response. Of the 36 site checks, 26 use another tool's ids and point scale so you can compare two reports directly.
Will this increase my traffic?
We do not put a number on that, and we will not. We can show you the score today, the score once the findings are fixed, and the same arithmetic behind both — that part is checkable. Converting it into traffic or revenue would mean inventing a link we cannot measure yet, and most of this category's confidence comes from exactly that invention.
What does it cost?
The scan is free, with no account and no card — it is the whole report, not a teaser. Ongoing work is scoped per site, because the amount of it depends entirely on what the scan finds. Start with the scan; if nothing in it surprises you, there is nothing to talk about.
Scan free. Then have us fix it.
One domain, 36 checks, twenty seconds, the whole report. Want the findings fixed? We write the files and open the pull request. If nothing surprises you, close the tab and we will never email you.
Scan my site