Anailís Eochairfhocail

Clóscríobh frása cuardaigh amháin. Féach cad a fhreagraíonn cúig inneall.

Ritheann an Anailís fiosrúchán amháin i gcoinne ChatGPT, Google, Perplexity, Gemini agus Google AI Mode ag an am céanna, ansin taispeánann sé duit an freagra a thug gach ceann acu, na foinsí a luadh, agus an raibh do shuíomh ann áit ar bith.

Léargas saor in aisce ar na hinnill AI. Níl cárta ag teastáil.

Cad a thagann ar ais

Do shuíomh, in aghaidh an innill

Faigheann gach inneall a chárta féin le fíorasc simplí ag an mbarr: tá do shuíomh le feiceáil ag suíomh N, nó níor aimsíodh sna torthaí seo é.

An freagra féin

Maidir leis na hinnill AI taispeánann muid an freagra ginte go hiomlán, ní achoimre air — an mhír iarbhír a léifeadh custaiméir.

Na foinsí a luadh

An liosta rangaithe de leathanaigh a raibh gach inneall ag brath orthu, le teideal, URL agus blúire. Tá do leathanaigh féin marcáilte sa liosta.

Inneall taobh le taobh

Marcáil an oiread inneall agus is mian leat agus ritheann siad go comhthreomhar, ionas gur féidir leat an easaontas eatarthu a léamh ar scáileán amháin.

Conas a oibríonn sé

1

Roghnaigh an staidéar a bhfuil tú ag obair air

Is éard atá i staidéar ná branda móide a phríomhshuíomh. Teastaíonn ceann amháin ón Anailís toisc gurb é an príomhshuíomh an fearann a mheaitseálann muid torthaí leis, agus cinneann tír agus teanga an staidéir cén margadh ina ritheann an fiosrúchán.

2

Clóscríobh frása amháin

Ní liosta eochairfhocail — téarma cuardaigh amháin, nó an cineál ceiste a chlóscríobhfadh ceannaitheoir i ndáiríre. Coinníonn frása amháin in aghaidh an reatha na freagraí inléite.

3

Roghnaigh do chuid inneall

ChatGPT, Google, Perplexity, Gemini agus Google AI Mode. Seoltar gach inneall marcáilte ag an am céanna agus líonann sé a chárta féin de réir mar a thagann a fhreagra.

4

Léigh trasna an tsraith

Suas le deich dtorthaí in aghaidh an innill, do shuíomh marcáilte, na luanna liostaithe. Stóráiltear gach rith i gcoinne an staidéir, ionas go dtógann na frásaí a thástálann tú isteach i dtaifead seachas imeacht nuair a dhúnann tú an cluaisín.

Cén fáth a sheiceáil frása amháin ag an am

Tá an chuid is mó d'uirlisí infheictheachta tógtha le haghaidh scála: na céadta eochairfhocal, scór, líne treochta. Is é sin an cruth ceart le haghaidh monatóireachta, agus an cruth mícheart don cheist a bhíonn agat de ghnáth ar dtús — cad a deir inneall AI i ndáiríre nuair a iarrann duine faoi seo?

Is é an Anailís an taiscéalaí ad-hoc. Níl aon teimpléad le socrú agus níl aon rud le sceidealú. Clóscríobhann tú an frása, faigheann tú cúig fhreagra ar ais, agus léann tú iad. Is é an t-aonad úsáideach is lú den phost é d'aon ghnó.

Tá sé tábhachtach toisc go n-easaontaíonn na hinnill níos mó ná mar a mbeifeá ag súil leis. Is féidir le frása ina bhfuil tú go compordach i dtrí cinn is fearr Google freagra ChatGPT a tháirgeadh a ainmníonn trí iomaitheoir agus nach luann tú riamh. Níl comharthaí rangaithe clasaiceacha agus comharthaí luanna mar an gcéanna, agus is é an t-aon bhealach iontaofa chun a fháil amach cé acu de do chuid frásaí a iompraíonn ar an mbealach sin ná ceist a chur.

Is é an chúis eile ná teanga. Is é an freagra ginte an táirge anois. Níl sé go leor a fhios a bheith agat nár luadh thú; má léann tú an mhír a taispeánadh ina ionad, agus má fheiceann tú cé na leathanaigh as ar tógadh é, insíonn sé duit cad a chreideann an t-inneall faoi do chatagóir agus cé a bhfuil muinín aige as faoi láthair chun é a mhíniú. Is leathanach é sin is féidir leat a léamh, agus go minic bearna is féidir leat a líonadh.

Nuair a bhíonn frása tábhachtach, bog é isteach i dteimpléad sábháilte sa Visibility Scanner, a ritheann na seiceálacha céanna thar seacht n-inneall ar bhonn in-athdhéanta agus a chuireann na ritheanna i líne ionas gur féidir leat gluaiseacht a fheiceáil.

Frásaí is fiú tosú leo

A blank query box is a hard place to start, so here is the order we use. It costs four runs and usually settles the question of whether you have a visibility problem at all.

Your own name. Ask the engines about your brand directly. This is the cheapest sanity check there is: if an engine gets your own business wrong, describes a competitor's product as yours, or cites a directory listing instead of your site, you have found something worth fixing before you look at anything else.

The category, without your name in it. "Best [thing you sell] for [who you sell to]". This is the phrase your prospective customers actually type, and it is the one where absence costs you money. Whoever the engine names here is your real competitive set in this channel, which is often not the set you thought.

The comparison. "[You] vs [a rival]". Engines answer these confidently and at length, and they will happily assert things about your pricing, features or reputation that came from a three-year-old review. Reading that paragraph is frequently the most uncomfortable minute of the exercise and the most useful.

The problem, not the product. "How do I [solve the thing your product solves]". These are the questions where an engine recommends an approach and then a supplier, and they reward explanatory content rather than product pages — so they usually show a different set of winners again.

Run those four, read what comes back, and you will know whether the gap is that you are absent, that you are present but described badly, or that you are fine and the worry was unfounded. All three are useful answers. Whichever it is, the phrases that turned out to matter belong in a tracked template afterwards so you can tell whether your fixes moved anything.

What a position number does and does not tell you

The position is a fact about one answer, from one engine, at one moment. It is not a rank in the way a Google position is a rank. Answer engines are not deterministic: the same query can return a slightly different answer and a slightly different citation list an hour later, because the model samples and because its retrieval layer sees a moving index underneath it.

So read a single run as evidence, not as a score. "Not found" on one engine on one day is a prompt to look; "not found" across four engines and two weeks is a finding. This is exactly why the tracked version exists — one measurement cannot separate a real absence from the noise, and a series can.

The citation list is the more durable signal in any single run. Which domains an engine reaches for is far more stable than the exact wording of its answer, and it tells you what kind of source it trusts for your category: your competitors' own sites, a review aggregator, a marketplace, a news outlet, a forum thread. That pattern usually points straight at the work.

What you will see

One card per engine. At the top of each, whether your domain appeared and where. Below it, the AI answer in full for the engines that generate one, then the ranked sources behind it — title, URL and snippet, with your own pages marked. Where you are absent, the card shows you exactly who was cited in your place.

Frequently asked questions

Which engines can I check here?
Five: ChatGPT, Google organic search, Perplexity, Gemini and Google AI Mode. Tick as many as you want in one run and they are dispatched in parallel. The wider seven-engine sweep, which adds Claude and DeepSeek, runs from the Visibility Scanner.
How do you work out my position?
We take your study's primary site, reduce it to a bare domain (protocol and www. stripped) and walk the engine's result list from the top. The first result on that domain is your position. On the AI engines the generated answer itself is skipped in that count, so position 1 means the first cited source rather than the answer. No match anywhere in the list means "not found in these results" — which is a real finding, not an error.
Do I need an account?
Yes, and a study. The study holds the primary site we match results against and the market the query is run in, so there is nothing sensible to compare a result to without one. Creating both takes under a minute.
Which country and language do the results come from?
Your study's. Google is queried with the study's location and language codes, and the AI engines are asked to answer in that market's language and to prefer in-market sources where they exist. Change the study's market and the same phrase can return a genuinely different picture — which is usually the point.
Does a run cost anything?
Each engine you tick draws credits from your plan's allowance, charged per engine per run — so a five-engine run costs five times a one-engine run. If you do not have enough left, the page says so instead of running. Pricing has the allowances.
Why does one engine come back slower than the others?
Some engines return a stored result set immediately; others generate an answer on demand. Cards fill in as their answers land, and the page keeps checking in the background for the stragglers for about half a minute before it stops.
How is this different from the Visibility Scanner?
Analysis is one phrase, right now, no setup. Visibility Scanner is the tracked version: a saved template of queries run across seven engines, stored run after run so you can watch positions move. Most people use Analysis to find the phrases worth tracking, then put those into a template.
It says I am not cited. Will this tell me why?
Not by itself — it tells you who was cited instead, which is where the investigation starts. For the causes on your own side, Site Pulse grades whether your pages are readable to an engine at all, and Schema Audit finds the structured-data gaps that stop them being understood.
Are my runs kept?
Yes. Every query is saved against the study it was run under, so the phrases you have tested stay available alongside the rest of that study's data on your dashboard.

Ask an engine what it says about you

Run a free scan
Féach ar do Rangú

Cuir isteach eochairfhocail agus féach cá bhfuil tú rangaithe ar innill chuardaigh AI.

George
Online
0%