For agencies · under your brand or none
When your client asks if they appear in ChatGPT
Open saved answers, sources and next actions. Share a report and prepare the conversation using data about their brand.
Free, no card needed: four questions to ChatGPT and Gemini. See the plans
One complete task, start to finish
- Task
- Bring what AI says about the brand to the monthly meeting
- Deliverable
- A report with the answers and one agreed action
- Done when
- The action is logged so you can see what changes afterwards
The meeting with your client
One report, one question, one agreed action
An example with demonstration data.
- The report
- Visibility over the last 30 days, with the answers and sources for each question.
- The client’s question
- “Why does ChatGPT recommend Quantly and not us?”
- What it shows
- The answers that name Quantly cite a comparison and a guide; none names your client.
- The agreed action
- Prepare their comparison page with facts the client confirms; their team reviews and publishes it. It goes into “What you’ve done” so you can see what changes afterwards.
Data for your next meeting
Real data for the next client meeting
When your client asks whether they show up in ChatGPT, open the saved answers instead of improvising.
The problem
The question has already reached your meetings
Your clients read that AI is replacing search and ask you what to do. Improvising an answer is a risk. Measuring it with method is a service you can offer tomorrow.
Answers
What AI answered, with its date
Every question of your client's is measured daily and every answer is saved in full, with its sources. You can show it as it is.
- Full answer
- Linked sources
- Daily measurement
Share of voice
How much your client is talked about against competitors
The part of the mentions each brand takes, by platform and by period, over the same answers.
- By platform
- By period
- Against each competitor
Assistant
Ask and get the figure
The assistant answers with the brand's data and says what it read, so you walk into the meeting with figures you can check.
- Answers with the workspace's figures
- The data it read, in view
- A link to each screen
Next step
One concrete priority for each client
The dashboard shows the questions where your client is missing and the sources that cite its competitors; each action links to the answers behind it and says how many there are, not the effect it will have.
- Questions where it is missing
- Sources that cite its competitors
- Each action, with its evidence
Frequently asked questions
What agencies ask us most.
Can't find your question? Write to us
What if I would rather measure many clients myself?
The platform’s Scale plan is built for that: up to 3 brands and 50 questions, unlimited team members, reports under your agency’s identity, API and MCP, for $349/month + applicable taxes. Measurements, listening, priorities and included drafts share one workspace. Compare the two plans on the pricing page.
How much does AnswerGravity cost?
Two plans: Grow ($109/month + applicable taxes) and Scale ($349/month + applicable taxes). They are paid monthly, with no minimum term and no yearly billing. The only free thing, with no card, is the visibility checker. When a plan offers the 7-day trial, it asks for a card.
What does AnswerGravity do automatically?
It measures your questions every day, searches for conversations around your keywords and prioritises next steps with their evidence. Turn on automatic preparation for a brand and drafts arrive in your workspace even when the website is closed. You review them, decide what to use and publish from your own account.
What changes between Grow and Scale?
Both share the same workflow. Grow tracks 10 questions at the same time and measures them every day on ChatGPT, Gemini, Google AI Overviews and Perplexity, with 1 brand and 2 team members. Scale tracks 50 questions at the same time and measures them every day on ChatGPT, Gemini, Google AI Mode, Google AI Overviews, Microsoft Copilot (Spain only) and Perplexity, with up to 3 brands and unlimited team members. Claude and Grok are measured every week on both, for all your questions. Listening on Reddit and Quora: Grow, 20 keywords once a day; Scale, 75 keywords once a day. Drafts a month: Grow, 4 new articles and 4 updates of pages you already have; Scale, 12 new articles and 12 updates of pages you already have. Capacity is shared across workspace brands.
What is the difference between questions and keywords?
Questions are what we ask AI platforms to measure their answers. They are tracking capacity: how many you have active at once, not a monthly allowance. Keywords drive listening for conversations on Reddit and Quora, a separate function. The monthly allowances are the drafts.
What is an update?
We use AI to rewrite a page your website already has so it answers your customers’ questions better. You review it and publish it. A new article, by contrast, is a page you do not have yet.
Is a draft prepared every day?
Not necessarily. Preparation follows the available evidence and spreads your monthly allowance. It avoids repeated opportunities and keeps at most three automatic drafts awaiting review. You can pause it or prepare a draft manually using the same allowance.
When do included drafts start?
With the paid plan, after any free trial. The allowance renews every month, with no rollover. Each draft includes two additional versions within seven days. Included drafts use no credits; once the allowance is used, each extra draft is paid with credits at your plan’s rate, as are thread replies, new threads and documents. They do not include publication.
Which AI platforms are measured?
ChatGPT, Gemini, Google AI Overviews, Google AI Mode, Microsoft Copilot (Spain only), and Perplexity. Each question is measured from its brand’s country, in that country’s language: Spain, the United States, or Germany. Grow includes ChatGPT, Gemini, Google AI Overviews and Perplexity. Scale includes ChatGPT, Gemini, Google AI Mode, Google AI Overviews, Microsoft Copilot (Spain only) and Perplexity. Perplexity is measured through its API with web search, and the others in their web versions. Claude and Grok are measured every week for all your questions, on both plans and outside the daily AI platforms, through their APIs with web search: not their apps or a personalised account. Weekly balances cost and follow-up; their answers can vary between queries. Copilot is measured when the measurement market is Spain, wherever your company is based. Each answer keeps its platform, date and sources; a failed measurement does not count as your brand being absent.
Are media placements included?
Media selection is part of the product. Sponsored placements cost extra and require your approval. Inventory will be added as supplier agreements and connections are verified. A publication needs a confirmed offer before any charge, and a paid placement never guarantees an AI mention.
How do I start and pay?
Choose a plan, create your workspace and pay by card through Stripe. Prices are in euros plus applicable VAT or, for customers outside the European Union, in US dollars plus applicable taxes; you see the total before confirming. If your plan offers the free trial, its button says so before you pay: it lasts 7 days, needs a card and measures up to 20 questions (10 with Grow) on 2 platforms a day and 3 listening keywords, with no drafts and no Claude and Grok, which are measured every week; drafts start with the first payment. We email you two days before it ends; unless you cancel, the chosen plan is charged when the trial ends. Otherwise the first charge is made when you buy.
How does cancellation work?
Plans are paid monthly, with no yearly billing. Plans lets you manage the subscription, review changes and cancel the next renewal. Access continues until the end of the paid period. The purchase terms explain the applicable rights and time limits.
What if I already have a plan?
If you already have a Starter, Growth, Enterprise or Local plan, you keep it with its limits and its price. Grow and Scale are the plans for new subscriptions.
Can I use it for clients?
Yes. Scale supports up to 3 brands, with unlimited team members, and its reports can carry your logo and colours. Invite your team, control access and share reports. The plan capacity is shared across all brands in the workspace. Working for several brands under your own identity? See the white-label platform.
Does it guarantee appearing in ChatGPT or another AI?
No. AnswerGravity preserves answers and sources, identifies opportunities and prepares work for review. Providers decide what to answer; a later increase alone does not prove that an action caused it.
Which other tools measure visibility in ChatGPT?
Several, such as Peec AI, Otterly.AI, Profound, Semrush, Ahrefs Brand Radar or MentionScore. We compare them with ours, with sources, dates and when not to choose us, in tools to measure your brand’s visibility in ChatGPT.
Your next step starts here
Measurements in ChatGPT, Gemini, Copilot, Perplexity and Google (plus Claude), evidence-backed priorities and included drafts to review. Choose your capacity and get started.
See plans