Limit
Three models, not all of them
GLIMPSE by Weevo queries Gemini, ChatGPT, and Claude. Perplexity and Copilot bring visits but are not queried.
Brand presence in AI answers
How often AI models mention your brand. With the number, and with the margin.
GLIMPSE by Weevo measures how often AI models mention a brand: it asks Gemini, ChatGPT, or Claude the same question dozens or hundreds of times and states the margin. It counts how many answers mention each brand. It is not a survey: it measures how the models behave.
To try it right away: three free trials, no credit card required.
3
models queried: Gemini, ChatGPT, and Claude
20 to 2,500
answers per question, chosen by you before you pay
±2.0 points
the margin with 2,500 answers on a 50% share
What has changed
Anyone who asks an AI for advice reads a single answer.
GLIMPSE by Weevo starts from a change: until recently, a search returned ten links and the reader made the choice. Today a question returns one paragraph with three or four names in it, and a brand that is not in that paragraph is not rejected: it is never considered at all.
The point is not to be first. It is to appear, and to know in how many answers out of a hundred you actually appear.
Four steps, with three free trials
From your website to a number with its margin.
01
Your website
It starts from your domain. GLIMPSE by Weevo reads the homepage, works out what you do and where you sell, and does not keep it. If the website cannot be read, two lines describing the business are enough.
02
Five questions, five intents
The questions a customer who does not know you yet would ask, one per category: five of the same kind would measure only one thing. Your brand name appears only in the Brand category question, and the questions can be edited before they are measured.
03
The free taste
One question, ten answers from Gemini and ten from ChatGPT, side by side, with no credits used. A brand that one model mentions and the other ignores has a specific problem, and it shows only when the columns sit next to each other.
04
How many answers you need
From the free taste, GLIMPSE by Weevo calculates how many answers you need for a 5-point margin on your share, before you spend anything. Then the full analyses start, one for each chosen question, on Gemini, ChatGPT, or Claude.
Three free trials, no credit card required
Try GLIMPSE by Weevo on your brand before you spend a single euro.
You create your GLIMPSE by Weevo account, enter your website, and within a few minutes you have the five questions and the free taste: ten answers from Gemini and ten from ChatGPT, side by side. Then you decide whether to measure, and how much.
No subscription: after the trials, €0.40 per answer.
The metric, stated
“AI visibility” is not one thing. GLIMPSE by Weevo always states which one it is measuring.
Mention
For GLIMPSE by Weevo, a mention is the percentage of valid answers in which the brand is mentioned at least once. An answer that mentions it five times counts only once.
Citation
The percentage of valid answers in which one of the domains you declared appears among the sources the model consulted. Same denominator, different event.
Support
How many sources support you, and how many of those are your own pages versus other sites' pages. Being cited only by your own website is a fragile position, and a different one from not being cited at all.
A brand with a 90% mention share and a 0% citation share has a specific problem: the model mentions it, but when it looks for a source it goes to someone else. This is why two tools measuring the same brand give different numbers that are both correct: they measure different things under the same name. GLIMPSE by Weevo states which one, inside every result and inside every exported file.
Gemini, ChatGPT, and Claude
Three models, measured one at a time.
Given the same question, two models can mention completely different brands. GLIMPSE by Weevo queries them separately, at the same cost in credits, so the difference becomes data instead of a surprise.
Gemini
Answers by searching with Google, and GLIMPSE by Weevo counts both the pages it consulted and the phrases it typed to find them.
Sources consultedyesSearch phrases typedyesCost1 credit per answer
ChatGPT
Shows its sources in only some of its answers and does not say what it searched for. GLIMPSE by Weevo records the sources that are there and does not invent the others.
Sources consultedonly partlySearch phrases typednoCost1 credit per answer
Claude
Returns the cited pages and the phrases it searched for, like Gemini, and GLIMPSE by Weevo counts them the same way.
Sources consultedyesSearch phrases typedyesCost1 credit per answer
Perplexity and Copilot are not queried. If traffic to your website shows visits coming from them, GLIMPSE by Weevo tells you, but it cannot explain those visits with a share.
±107%
Nobody writes it down, and it decides everything
Two hundred answers are enough for one market and not for another.
With 200 answers, the margin GLIMPSE by Weevo states is about ±6.3 points on a 30% share and ±2.1 points on a 2% share. In points, the second looks more precisely measured; relative to the value, they are ±21% and ±107%, and in the second case the interval is wider than the number itself.
In a crowded market your share is structurally small, which is why the same depth buys much less precision. GLIMPSE by Weevo tells you before you pay, not after.
95% confidence interval, Wilson method. It appears next to every percentage, on every page, and in every export.
After the number
A number alone does not tell you what to do. The Brand center does.
For every website you declare, GLIMPSE by Weevo brings together what the answers say and what your website knows.
Who cites you, and who cites your competitors
For each question, the sources that back you and the sources that back the others. A passage is attributed to a brand only when it mentions a single brand: if it mentions three, there is no way to know which one that source belongs to, and GLIMPSE by Weevo does not guess.
Sentiment and trends over time
Two brands with the same share can be described in opposite ways. And change from one month to the next.
Being mentioned is not enough: the words matter. A brand often described as “expensive” or “outdated” has a different problem from a brand that is not mentioned, and no share shows it. That is why GLIMPSE by Weevo also measures sentiment, brand by brand, on the same answers it uses to calculate the shares, and tracks it over time.
The terms, with their sign
In each answer, GLIMPSE by Weevo collects up to five terms the model uses to describe each brand — “creamy,” “expensive,” “good for kids” — and counts them across all answers. Each term has a score from −1 to +1, and the brand's sentiment is the average of its terms.
Share and sentiment together
The chart that crosses them separates four situations: mentioned and liked, mentioned but described poorly, liked but rarely mentioned, absent. Each one calls for different work.
Sentiment over time
Searches repeat automatically every week, month, quarter, or six months. Sentiment and shares each have their own trend, and for the main brands you can see how the most used terms change from one measurement to the next.
The series that doesn't add up
If the question, provider, model, country, web search setting, or metric changes between two runs, the time series is flagged as not comparable and the log shows where and when it changed. Increasing the depth, on the other hand, breaks nothing.
Many “drops” in visibility data are actually changes of instrument: a step in the chart between two different measurements is not news about the market. And sentiment does not affect the shares: if classification fails, the search stays valid and sentiment is reported as missing, not neutral.
Phil and client reports
Phil tells you where to start. You deliver the report with a link.
Phil is the assistant inside GLIMPSE by Weevo reports: an AI that analyzes the answers of AI models, using the data from the search in front of you. The report is made for people who do not use the tool: the client, management, a colleague in another department.
A comment for every brand
Phil reads the answers, analyzes sentiment and key terms, compares the brands, and for each one writes strengths, areas for improvement, and strategic advice.
A chat about your data
When a number doesn't add up, you ask Phil: why a brand stops at 50%, which terms hurt it, what a competitor does better. Phil answers based on that search.
A link for the client
A switch makes the report public, and the link is ready to copy. The client does not sign up and does not enter your account: they see only that search, in the browser and read-only. If you turn the switch off, the link stops working.
The words are yours
Before publishing, you can change the title and rewrite every text in the analysis: the client receives a reading in your voice, not automated output.
Need to work on it together? A search can also be shared with another GLIMPSE by Weevo account. The report downloads as CSV and searches as JSON, each file with the metric measured. The interface and reports are available in eight languages.
No subscription
One credit per answer. €0.40 per credit.
Trial
Free
Three complete free tastes of GLIMPSE by Weevo, each on Gemini and ChatGPT with ten answers per model. No credit card required.
Credits
€0.40
One credit costs €0.40 with any of the three models. Packs of 50, 100, 200, 500, 1,000, and 10,000 credits: a search with 100 answers costs 100 credits.
No surprises
€0 a month
Credits do not expire and there is no volume discount. If a batch is interrupted, the credits for the answers not produced are returned.
| Depth | Answers | Margin on a 50% share | Cost |
|---|---|---|---|
| Exploratory | 20 | ±20.1 points | €8.00 |
| Basic | 100 | ±9.6 points | €40.00 |
| Extended | 500 | ±4.4 points | €200.00 |
| Deep | 2,500 | ±2.0 points | €1,000.00 |
The price is the same for every pack. No plan costs more for the same answers.
What GLIMPSE by Weevo cannot do
We write them here because you would find them anyway, and it is better to know before.
Limit
Answers change
The same model gives a different answer to the same question from one month to the next, and a provider can change the model behind the same name. That is why a time series is needed, not a single measurement.
Limit
ChatGPT cites less
ChatGPT shows sources in the text in about one answer out of three in the tests GLIMPSE by Weevo ran in September 2026, Gemini almost always: the citation share is compared over time on the same model, not across different models.
Limit
Merging makes mistakes
The automatic merging of name spellings gets some cases backwards, and they have to be corrected by hand.
Limit
Analytics undercounts
Google Analytics sees only visitors who accepted cookies and have JavaScript enabled, and nobody knows how many that leaves out. That is why GLIMPSE by Weevo never adds those numbers to other traffic measurements.
A tool that measures how AI models describe brands has to withstand the question “what are your limits?” If the answer is not written down anywhere, it means nobody looked for them.
Ask a question about your market and see who shows up in your place.
Trying GLIMPSE by Weevo takes about a minute and no credit card is required. At the end you will have two columns side by side, Gemini and ChatGPT on the same question, and you will know whether your name is in them.
Published on updated on
StrengthsDescribed as creamy and healthy, almost never negatively.
Areas for improvementAbsent when the question is about protein.
Strategic recommendationsA page on protein content, comparable with Greek yogurts.