AI visibility

NameIO asks AI engines the questions shoppers ask before they buy. It then shows you which answers mention your store, which pages they cite, and which answers name or cite a competitor but not you. This page covers setting up your store, reading the results, running and scheduling checks, how each engine's answers are collected, and what checks cost.

Set up your store

Quick setup is the first step of onboarding after you sign up at https://app.getnameio.com/sign-up, labelled "Website". The same form also appears on any AI visibility or Prompt Explorer page whose project has no website or hasn't been set up for AI visibility yet. It is headed "See how AI recommends your store".

1. Enter your Store website, for example yourstore.com. If the project already has a website, this field is locked.
2. Enter your Brand name.
3. Enter What you sell, for example "trail running shoes". It must be 2 to 80 characters and six words or fewer.
4. Under Competitors (up to 5), enter at least one competitor's website. The name is optional. Select Add competitor for another row, or the X button to remove one. Rows with no website are left out.
5. To read the questions before they run, select See the 20 questions. The list is filled in with what you typed.
6. Select Run my baseline. During onboarding you can select Skip for now instead.

If you leave a competitor's name blank, NameIO makes one from its website: ridgeline-running.com becomes Ridgeline Running. You can't enter your own website as a competitor; the form shows "Please check your input and try again."

On the free plan, quick setup tracks ChatGPT. On Pro it tracks ChatGPT, Gemini, Google AI Overviews and Perplexity. Gemini, Google AI Overviews and Perplexity are on Pro, so on the free plan you can't add them in Tracking settings either.

When you select Run my baseline, NameIO:

1. Saves your domain and, if the project has no overview yet, a short factual overview ("{brand} sells {category} online at {domain}."), plus your homepage title when NameIO can read it. It never copies page body text.
2. If the project has no Prompt Research keywords yet, saves what you sell, "best" followed by what you sell, your brand and your competitors' names, keeping those of five words or fewer.
3. Renames the project to your brand.
4. Creates the tracker with the questions, or adds them to the tracker you already have.
5. Starts the baseline check.

No model runs during setup. If the project already has a saved competitor list, the questions use those names and setup leaves the list as it is. An existing tracker, from an earlier setup or one an agent created over MCP, keeps its own prompts and engines, gains the 20 questions, and adds any engines your plan includes that it doesn't track yet (ChatGPT on the free plan; ChatGPT, Gemini, Google AI Overviews and Perplexity on Pro). A new tracker starts in the project's market, or in the US where an engine can't collect in that market.

A new tracker is created on a weekly schedule, with its first scheduled check at least a week away, so creating a tracker never charges anything on its own.

After the Website step, onboarding continues with Google Search Console and Agent setup (MCP).

The 20 questions

Every store is measured on the same 20 buying questions. They are fixed templates filled in with your brand, what you sell and your competitors. No model writes them, so writing them costs nothing; the baseline check that follows uses credits. Identical questions, for example when a brand is typed twice, are asked once, so you can end up with fewer than 20.

In the list below, {brand} is your brand name, {category} is what you sell, {competitor1} and {competitor2} are your first two competitors, and {year} is the current year (UTC) at setup. If you entered only one competitor, {competitor2} reads "other leading brands".

The five groups become topics in your tracker.

Brand discovery

1. What are the best {category} brands right now?
2. Is {brand} a good brand for {category}?
3. What do people say about {brand} quality?
4. Who are the top competitors to {brand}?

Buying intent

5. What's the best {category} to buy in {year}?
6. I'm looking for {category}. What do you recommend?
7. Which {category} brand has the best reviews?
8. Which {category} brand is worth the money?
9. What should I look for when buying {category}?

Problem-aware

10. What's the best way to find quality {category}?
11. I need reliable {category}. What do you recommend?
12. What do experts recommend for {category}?
13. What's the best alternative to {competitor1} for {category}?

Comparison

14. {brand} vs {competitor1}: which is better?
15. What's the difference between {brand} and {competitor1}?
16. Why do people choose {brand} over {competitor1}?
17. Compare {brand}, {competitor1} and {competitor2} for {category}.

Trust and reputation

18. Is {brand} legit and trustworthy?
19. Does {brand} have good customer service?
20. What are the most common complaints about {brand}?

Questions 2 to 4 and 14 to 20 contain your brand name, so NameIO marks them Branded.

Run the baseline

The baseline collects live answers, the same way Run now does, so results arrive sooner than on the scheduled queue. While it runs, the page shows "Collecting {done} of {total} answers…" with a progress bar. When it finishes, it shows "Updated {date} · Next run {date}".

A check, the baseline included, starts only if your credits (monthly credits plus any top-up credits) cover every answer. If they don't, your setup is still saved, the check doesn't run, the tracker records why, and the form shows "You don't have enough credits for this. Upgrade your plan or buy more credits to keep going."

On the free plan, the whole baseline must also fit within today's free daily budget, unless your top-up credits cover it, or the check doesn't start. If it doesn't fit, your setup is saved and the form shows "Your setup is saved. Free checks are full today. They reset at 00:00 UTC; run your check then, or upgrade to check now."

How answers are collected

NameIO collects every answer through DataForSEO, which passes your prompts and brand names to OpenAI, Google and Perplexity to collect answers. Each answer panel names its source as "{Engine}: {source label}".

DataForSEO has no scraper for Perplexity, so every Perplexity answer, scheduled ones included, comes from the model API.

Scheduled checks wait in DataForSEO's queue, which is cheaper. Run now and the baseline ask live, which is faster.

Google doesn't show an AI Overview for every search. When a results page has none, NameIO records a no-answer result, not a missed brand: the status reads "No AI Overview shown" and the answer reads "Google showed no AI Overview for this search." For the other engines, an empty result reads "No substantive answer".

If DataForSEO's own backend fails an answer, NameIO asks again, up to 3 attempts in all. Timeouts and connection errors are never retried, because they can hide a charge. A failed collection is marked "This collection is excluded from visibility rates."

In Tracking settings, engines are listed in two groups: Scraped responses (ChatGPT, Gemini and Google AI Overviews) and API responses (Perplexity). The country list only shows countries that every selected engine can collect in. ChatGPT can't collect in a few countries, and Perplexity collects only in countries its web search accepts.

Prompt Tracking

In the sidebar, the AI Visibility group holds Prompt Research, Prompt Explorer and Prompt Tracking. Your results are on Prompt Tracking.

The page holds one card with your brand name, the schedule badge, the Tracking settings and Run now buttons, the run status line and the Visibility over time trend. Below them are the tabs Prompts, Competitors and Citations. The tabs show the latest run, which can be a Run now check.

NameIO uses three terms throughout:

• Mentioned: your brand name or domain appears in the answer text.
• Cited (an owned citation): the answer cites a page on your domain.
• Branded: the prompt's text contains your brand name or domain.

NameIO keeps finished runs for 13 months, then deletes them along with their answers.

Visibility over time

Visibility over time compares the latest 7, 28 or 90 days with the same number of days before them. The table below the trend has an All engines row and, when you track more than one engine, a row per engine, each with Brand mentions and Owned citations; when the periods can be compared, it shows the change with ▲ or ▼.

The trend counts non-branded prompts only, since a prompt that names your brand would always mention it. It uses finished runs of every kind: baseline, scheduled and Run now. Notices flag a period where under 90% of the planned answers were collected, a change in what's tracked, or no finished runs in the latest period.

Prompts

The Prompts tab lists every prompt by topic, with the columns Prompts by topic, Engines, Brand mentions and Owned citations. Each prompt is marked Active, Paused or Collecting answers, and Branded or Non-branded. In the Engines column, each engine's name is green when the answer mentions you, red when the collection failed and grey otherwise. Hover it to read the result: Mentioned, Not mentioned, Failed, Running, No AI Overview shown, No substantive answer, Answer unavailable or No result available. Before any results arrive, the tab shows "Run prompts before interacting with results."

The toolbar has Search prompts, Export and Add prompts. Add prompts takes up to 5 prompts at a time, each up to 2,000 characters. A project can track up to 100 prompts in up to 30 topics; archived prompts don't count.

Competitors

The Competitors tab compares you with the competitors you track. Your row is marked You.

These counts use non-branded prompts only. When you track several engines, use the engine menu (it reads "4 engines" with all four selected) to choose which engines the counts cover. Use Manage competitors to edit your competitors on the project's Context page.

Competitor gap lists answers that name or cite a competitor you track and neither name nor cite you. Its columns are Question, Engine, Recommended instead (the competitors named or cited) and Sources (the number of citations). It shows up to 50 rows and, unlike the counts above, includes branded questions. Before your first check starts it reads "Fills in after your first check finishes." If no answer names or cites a competitor without also naming or citing you, it reads "No gaps: every answer that names a competitor also names you."

Suggested competitors shows up to five other sites that AI answers cite most for your prompts, each with the number of answers that cite it and an Add button. A fixed list of common platforms, such as reddit.com, youtube.com and wikipedia.org, is left out, along with your own site and the competitors you already track.

Citations

The Citations tab lists the pages and sites the answers cite. Switch between Pages and Domains, filter by ownership (All ownership, Your domains, Competitor domains, Third-party domains), and, when you track several engines, use the engine menu (it reads "4 engines" with all four selected) to choose which engines' sources are listed. The columns are Cited page (or Domain), Ownership (Your domain, Competitor or Third party), Answers, Prompts and Engines.

One prompt's history

Once a prompt has answers, select its text on the Prompts tab to open its page. The page is titled with the prompt and shows its topic.

• Choose a date range: Last 7 days, Last 28 days, Last 90 days, All available history or Custom range.
• Executions lists each answer by date, Engine, Brand mentioned and Owned citation; baseline and scheduled answers are tagged Baseline or Scheduled.
• Answer shows the answer text with its source label, and tabs for its Citations and Brands.
• Brand comparison shows how often each tracked brand was Mentioned and Cited in the chosen date range, as a percentage and a count of answers.

Export

On the Prompts tab, the Export menu offers Export CSV and Download JSON for the run on screen, branded prompts included. The files are named ai-visibility.csv and ai-visibility.json, and the download link expires after 1 hour.

The CSV has the columns run_id, observation_id, prompt, engine, status, collected_at, brand_mentions, citation_urls and answer.

Run a check now

1. On the Prompt Tracking card, select Run now.
2. Check the Cost and the number of answers in the dialog. The price is fetched again each time the dialog opens, and the button stays disabled until it loads.
3. Select Run now. Run now collects live answers, ready in a few minutes.

If the price rises above the amount you approved before the check starts, the check is refused.

On the free plan, Run now starts only if the whole check fits in today's free daily budget, or your top-up credits cover it. Otherwise it shows "Free checks are full today. They reset at 00:00 UTC; run your check then, or upgrade to check now." A check bigger than the free plan's daily budget can never fit and shows "This check is bigger than the free plan's daily budget. Pause some prompts or engines, or upgrade."

Schedule checks

Scheduled checks are part of Pro: on the free plan you can't turn a schedule on or change it. To schedule a tracker:

1. Open Tracking settings. In Schedule, select Change schedule, or Schedule tracking if the tracker is paused.
2. Choose Daily, Weekly or Monthly (end of month). Weekly is the default. Select change to pick the day of the week (for weekly) and the time of day, shown in your local timezone. If you don't pick a time, a tracker that's already scheduled keeps its current time, and a paused tracker you schedule again gets a random time between 04:00 and 09:59 UTC.
3. Check the cost per run and the monthly estimate, then select Start tracking or Save schedule.

Choosing Daily shows "Daily checks use 7x more credits than weekly". On Pro, select Pause to stop checks. The tracker badge reads Daily tracking, Weekly tracking, Monthly tracking or Paused, and on the free plan Manual checks.

When you save a schedule and the project has no baseline or scheduled check yet, NameIO also collects a baseline right away at the Run now price. If you don't have enough credits, the schedule still saves and the tracker says why no check ran. When a scheduled check finds too few credits, the page shows "You don't have enough credits for this. Buy more credits to keep going, or wait for your monthly credits to refill on {date}." with a Buy credits button.

Scheduled checks from all NameIO accounts share a daily ceiling. If today's ceiling is already spent, your check waits and retries an hour later, and its slot moves by the delay. A single check that costs more than the whole ceiling is skipped, its schedule moves on, and the page shows "The last scheduled check was skipped: it costs more than today's budget for scheduled checks. Pause some prompts or engines, or use Run now."

On the free plan, a tracker runs only when you run it. Its badge reads Manual checks, no next run is shown, and the page shows "Scheduled checks are part of the paid plan. Upgrade to run them on a schedule, or use Run now." After you upgrade, the notice goes away and the tracker's schedule runs.

What checks cost

Checks are paid with credits, shown in the app as dollars: $1 is 1,000 credits. Estimated cost per answer:

For the 20 questions, that comes to about:

• $0.12 for a free-plan baseline on ChatGPT.
• $1.40 for a baseline or Run now on all four engines.
• $1.16 for a scheduled check on all four engines, or about $4.64 a month on a weekly schedule.

Perplexity is most of the cost, because its answers always come from the model API.

The app shows an estimate before you commit: the Run now dialog shows the cost of the check, the schedule dialog shows Cost per run and a monthly figure, and adding or editing prompts shows what each run costs and an estimated monthly total. Monthly figures assume 30 checks for daily, 4 for weekly and 1 for monthly. Before a tracker exists, estimates assume a weekly schedule.

NameIO holds the estimate on your credits before asking DataForSEO, then charges for what DataForSEO actually billed, at the same credit rate as the estimate, instead of the estimate. Scheduled answers are sent in batches under one charge, so the real charge can be lower than the estimate.

Plans and credits

Pro is $54 a month or $499 a year, plus applicable tax. Both include $15 of usage credits every month.

The free plan's credits are a one-time amount and don't refill. When they run out, a banner reads "You've used your free credits. Upgrade your plan to get $15 of credits every month."

Free accounts also share a daily data budget, and each free organization can use up to $0.25 of it a day. The cap applies to your organization, not to each store or project. Budgets reset at 00:00 UTC. The budget only limits calls paid from your free credits. Top-up credits are bought on Pro and never expire. If you still have some on the free plan, calls they pay for don't count against the budget, and when the budget is full, calls move to them if they cover the cost. Otherwise the calls are refused.

If you upgrade with a tracker that tracks only ChatGPT, the Prompt Tracking card offers the other engines that can collect in the tracker's country: "Your plan includes Gemini, Google AI Overviews and Perplexity. Add them to this tracker; checks use your monthly credits." Select the Add button that names them. The next check covers them.

Prompt Research

Prompt Research finds the questions people ask AI about your market, and the sites the answers cite. It is part of Pro; on the free plan the page shows Unlock Prompt Research with View plan and Use Prompt Tracking without upgrading. It needs a tracker; without one it shows Set up prompt tracking.

1. Open Prompt Research in the sidebar.
2. Under Relevant keywords, select a keyword. The list holds the keywords quick setup saved plus your tracker's topics.
3. Read the results under "Prompts about" your keyword. Each prompt shows its Sources, and is marked Tracked if your tracker already has it.
4. Select the prompts you want and use Track prompts to add them to your tracker under a topic.

Each keyword costs about $0.25 in credits. Results are cached for your organization for 24 hours, or 1 hour if the Google search failed.

Prompt Research runs no model. It reads DataForSEO's LLM Mentions data, which holds ChatGPT questions and answers, together with a live Google search for the keyword (top 50 results). A prompt is kept when it asks about a keyword of two or more words as a phrase, or when the sites its answer cited or found in its web search include one that ranks for the keyword, your own site or a tracked competitor's site. It fetches up to 100 questions per keyword, ordered by DataForSEO's AI search volume, which is estimated from Google "People also ask" data rather than counted from AI usage. Near-duplicate questions are merged. When an answer cited nothing, its sources read "The AI answered from memory and cited no sources."

Prompt Explorer

Prompt Explorer asks one prompt across one to three models and shows the answers side by side with their cited sources. It is part of Pro; on the free plan the page shows Unlock Prompt Explorer with View plan and Use Prompt Tracking without upgrading.

Its answers come from each model's API through DataForSEO, not from the chat apps, so the models are labelled ChatGPT (API), Gemini (API) and Perplexity (API).

1. Open Prompt Explorer in the sidebar.
2. In Prompt, enter what your customers might ask AI, up to 500 characters.
3. Optionally, enter a Highlight brand to flag whether each model mentions it.
4. Under Models, choose up to three. ChatGPT is selected by default.
5. Leave Allow web search on (the default) or turn it off, and, with web search on, pick a country in the box next to it if you want one. The default is "No country preference".
6. Run the prompt.

Estimated cost per model: ChatGPT (API) 116 credits ($0.116) with web search or 64 without; Gemini (API) 141 credits ($0.141) or 64; Perplexity (API) 52 credits ($0.052). The charge settles on the real cost. If you asked for web search and a model answered without searching, NameIO makes one more paid call.

Answers are cached for 7 days for each prompt, web-search setting and country in your project, so changing only the highlighted brand doesn't charge again.

What's not done

• Claude isn't a tracked engine: its answers would come from the model API, not the Claude app. It isn't offered in Prompt Explorer either.
• Quick setup, Prompt Explorer and the baseline that a first schedule starts don't show a cost before they run. The estimates on this page apply.
• On the free plan, the daily budget is checked when a check starts but not held. If other lookups use it up while the check runs, the remaining answers fail unless top-up credits cover them.
• Export covers one run at a time. A prompt's history can't be exported.
• Prompt Research only has data in US English.
• NameIO doesn't write prompt suggestions for you. Add prompts yourself or find them in Prompt Research.
• Mentions are matched by name, so a brand with a generic name can be counted as mentioned when an answer uses that word for something else.