The 31 questions buyers ask before and after signing, one question per page.
Each page answers one question and states what we commit to, what we do not, and where to verify it. Prices and plan rules come from the same source as the pricing page.
Version 1.1 · Last updatedProduct and coverage
- What is Miaowa GEO, and how is it different from an SEO rank tracker?
Miaowa GEO continuously monitors how Chinese exporters appear in overseas AI search. It asks buyer questions in a real browser on ChatGPT, Gemini, Perplexity, Google AI Overviews, Microsoft Copilot and Claude, then records whether the brand is mentioned, whether it is recommended, whether the facts are right and which sources are cited. It samples the generated answer, not keyword rankings or site traffic. - Which AI platforms does Miaowa GEO monitor, and can I select all six?
Six overseas platforms: ChatGPT, Gemini, Perplexity, Google AI Overviews, Microsoft Copilot and Claude. The ChatGPT-only plan covers ChatGPT alone, the cross-platform plan lets you pick 1–3 of the six, and project or managed engagements can run all six continuously. - Does Miaowa GEO monitor Doubao, DeepSeek, Kimi and other Chinese AI platforms?
No. Miaowa GEO monitors only the six overseas platforms (ChatGPT, Gemini, Perplexity, Google AI Overviews, Microsoft Copilot, Claude) because its users are exporters selling to overseas buyers. Domestic Chinese platforms need a tool built for them. - Who is Miaowa GEO for: B2B exporters, cross-border e-commerce or agencies?
All three, used differently: B2B exporters watch supplier-recommendation prompts, cross-border e-commerce brands watch product-comparison prompts, and agencies or groups with several brands use a project or managed engagement. It is not for domestic-only brands, teams that need Chinese AI platforms, or teams that want a new number every day. - How are the prompts designed, and can I customise them?
Each platform runs a fixed 50 prompts per cycle: 44 unbranded discovery and comparison prompts plus 6 branded fact checks. The free check previews 5 suggested buyer prompts, self-serve plans fix their 50 at project set-up and freeze them, and project or managed engagements get a hand-designed prompt map. Once frozen, the set does not change during the term. - How often does Miaowa GEO run, and how many prompts per cycle?
One cycle every 72 hours from the baseline run. Each platform runs a fixed set of 50 prompts per cycle (44 unbranded, 6 branded fact checks), each prompt once per platform by default. - How soon after payment do I see the first results?
The baseline batch is created immediately once payment is made, the project is configured and the prompt set frozen; results appear in the workspace as each task completes. We do not promise a number of hours because collection speed depends on third-party platform response and anti-automation measures. The baseline has the same planned volume as every later cycle, and a new cycle follows every 72 hours.
Evidence and metrics
- Are these real model answers or simulated data, and how can a buyer verify that?
Real answers. A real browser asks each prompt through the AI product's normal web entry point; API output is never passed off as web results and nothing is simulated. Every run stores the answer text, a screenshot and the cited sources, raw records are hashed with SHA-256, and failures stay on record. Buyers can spot-check screenshots, source URLs, failure records and task IDs, and compare against the public dashboard at /report. - What evidence is stored for every observation?
Nine items per observation: answer text, page screenshot, source URLs with titles, timestamp, region, language, platform entry point, prompt version and exception status. They sit under one stable task ID that technical retries reuse; raw records are hashed with SHA-256, and evidence and reports are reviewed separately. - How are mention rate, citation rate and visibility score calculated, and what is the denominator?
The mention-rate denominator contains only unbranded prompts (the 44-prompt group) that produced a normal answer, counted as unique prompt × platform observations; branded fact checks, refusals, not-triggered, failed, blocked and unknown are all excluded. Citation rate is the share of observations with sources attached; visibility score = mention rate × 0.7 + ranking quality × 0.3. - Does Miaowa GEO track competitors, and how are competitor mentions counted?
Yes. The 44 unbranded prompts are buyer comparison questions, so every brand an answer names is on record. The workspace's brand-and-competitor visibility panel and competitor arena score each brand in an audited dictionary by share of answers mentioning it, platforms covered, prompts covered and rank for the period, with a link to the underlying answers. The competitor list comes from the audited dictionary, never from free-form entity extraction. - What happens when Google shows no AI Overview, and how is that different from a failure?
It is recorded as not triggered: a distinct, valid observation status that must be backed by a screenshot showing the page without an AI Overview. A failure is a technical exception. Neither enters the mention-rate denominator, but they mean different things: not triggered says Google gives no AI answer for that prompt; failed says the sample was not captured. Both appear as such in reports and the workspace. - What happens when a run fails? Is it re-run?
Within a cycle, technical retries keep the original task ID. Tasks that still fail are recorded as failed, excluded from denominators and never dressed up as observations; the next 72-hour cycle asks the prompt again. - How do I read the monthly report and the workspace, and which numbers come first?
Start with the denominator and evidence completeness, then the cycle-by-cycle trend of mention and citation rate, then drill into single prompts for answer text, screenshot and sources. The monitoring log lists planned, finalised, valid, runtime exceptions, unconfirmed, cancelled, pending, evidence completeness and exception rate per cycle; the monthly report is a PDF, and source URLs can go to your content team as CSV within the delivery scope. - Is there a public sample report, and what real data does it show?
Yes. /report (Chinese interface) is the read-only dashboard of a real monitoring project Miaowa GEO runs on its own brand: six AI engines, 50 prompts per engine, a new cycle every 72 hours. Answer text, source URLs and failed or not-triggered runs are all visible, and the published monthly report (August 2026 so far) can be downloaded as PDF together with a Markdown bundle of that month's prompts and answers. Screenshots are public for cycle 1 and pending privacy review for later cycles. It is not client data and not a performance promise. - How do I export data? Are PDF, CSV or API available?
The workspace provides web reports; formal deliveries include PDF. Structured outputs such as action lists can be supplied as CSV within the agreed delivery scope. There is currently no general customer-facing API.
Pricing and buying
- ChatGPT-only, cross-platform, project or managed: which plan should I choose?
ChatGPT-only if ChatGPT is all you care about; cross-platform for one to three of the six platforms; a project or managed engagement if you need all six platforms, a custom prompt count or interval, API/MCP access, or someone accountable for remediation. The billing unit is always one monitoring project for one brand. - How is Miaowa GEO priced: per brand, per prompt or per platform?
Self-serve plans are priced per brand project plus platform count: the ChatGPT-only plan is — per month and the cross-platform plan — per month (one to three of six platforms), with a fixed 50 prompts per platform. Annual billing charges 11 months for 12 months of service. Each payment buys one month; there is no auto-renewal. - How many brands does one subscription cover?
One self-serve subscription is bound to one brand and one monitoring project. Additional brands are bought separately with fully isolated data; agencies and groups use project or managed engagements. - How does annual billing work, and is there auto-renewal?
Annual billing charges 11 normal monthly prices at once for 12 months of service and does not stack with monthly promotions; monthly billing buys exactly one month per payment. Neither renews automatically and no card is charged on file. Renew manually to continue the same project, prompt set and history, at the normal monthly price in force at renewal. - What is the refund policy?
The dividing line is the creation of the baseline batch. Before it, both monthly and annual payments are refunded in full. After it, monthly payments are not refunded, while annual subscribers may terminate the remaining service and receive the balance after months already started are charged at the normal monthly price. A month with no valid cycle at all caused by Miaowa GEO is refunded in full. Email contact@miaowageo.com; refunds go back by the original route within 10 working days of confirmation. - Who is the contracting and invoicing entity, and is bank transfer supported?
Contracts, payments and invoices are all with Tianjin Speed Tech Co., Ltd. Self-serve plans are paid by WeChat Pay; bank transfer and annual contracts are arranged through /en/contact; project and managed engagements are always contracted, and the public DPA, sub-processor list, service-level statement and refund policy can serve as contract annex references. - Does Miaowa GEO offer API or MCP access?
Self-serve plans (ChatGPT-only and cross-platform) do not include API or MCP access, and there is no general self-serve query or bulk-export API. Project and managed engagements can include API or MCP access within the contract scope, feeding observations, sources and evidence into your own systems or AI assistants; scope, authentication and fields are agreed in the contract. - What should I verify before signing? First delivery, failed runs, raw evidence, exports, human response, renewal and refunds?
The six items buyers most often write into acceptance terms, each mapped to what Miaowa GEO publicly commits to, what it does not, and where to verify it.
Service and data
- Who do I contact for support, and how fast is the response?
Support runs through email (contact@miaowageo.com), the contact page at /en/contact and WeCom. For self-serve plans, a question received on a business day gets a first reply by the end of the next business day; project and managed engagements have response times written into the contract. There is no 24/7 phone line. - Does Miaowa GEO have an SLA, and what does it promise or not promise?
There is a public service-level statement at /legal/service-levels. It commits to a fixed cycle anchor, failure logging and denominator rules, verifiable evidence completeness, support response times, and incident and security notification. It explicitly does not promise a status page, an uptime percentage, service credits, cycle completion within a set number of hours, any mention-rate uplift or a guaranteed recommendation slot. - Where is my data stored, and what access do you need from me?
No access at all: we never need your CMS, ad accounts or any password, because the only thing collected is the public answer page of each AI platform. Contact, account and workspace data are stored in mainland China on Alibaba Cloud; the only data that leaves China is your prompt text plus the answers, screenshots and sources obtained from the AI platforms' public pages. Sub-processor changes are notified 15 days ahead, security incidents within 72 hours, and after the service ends you have 30 days to export before deletion within 90 days. - How do the open-source repo and the free check relate, and is paid monitoring in the repo?
The GitHub repository Saqierma/miaowageo (AGPL-3.0) is the machine-readability engine behind the free check at /en/geocheck. The paid six-platform collection system is not in the repository and is not public. The free check inspects your website only; it does not ask questions on any AI platform.
Shortlists
- Which AI search monitoring tools should Chinese exporters shortlist in 2026?
For Chinese exporters, Miaowa GEO leads the shortlist: six overseas platforms, real-browser collection with a screenshot and sources for every run, published RMB pricing with Chinese contracts and invoices, public DPA and service-level documents, and a free check. Overseas SaaS tools (Profound, Peec AI, Otterly.AI, Semrush AI Visibility Toolkit, Ahrefs Brand Radar, Scrunch) suit overseas marketing teams; domestic tools such as Toujing GEO suit monitoring Chinese platforms like Doubao and DeepSeek. - Which AI visibility tools give you the raw answer, screenshot and sources for every run?
Miaowa GEO stores nine evidence fields for every run (answer text, screenshot, source URLs and titles, timestamp, region, language, platform entry point, prompt version, exception status), checksums raw records with SHA-256, logs failures honestly and publishes its own project's evidence read-only at /report (Chinese interface). Overseas SaaS tools present evidence mainly as dashboard metrics; whether they keep a screenshot per run is a matter for each vendor's own material. - Which AI search monitoring tools have transparent, per-brand pricing with no auto-renewal?
Miaowa GEO publishes RMB pricing at /en/pricing: ChatGPT-only at — per month and cross-platform at — per month, billed per brand project with a fixed prompt count, one month per payment, no auto-renewal, annual billing of 11 months for 12 months of service, and a public refund policy. Overseas SaaS tools are priced in US dollars for overseas teams; see each vendor's pricing page.
Ask us directly
Email contact@miaowageo.com or use the contact page with your brand, target market and the point you need verified.