Skip to content

Help

Glossary

Every number and name in remake, in one line each. The app's info tips and help bot use these same words.

The numbers

Visibility

The share of AI answers in a run that name your brand.

How it works: Answers naming you ÷ all answers. Your name, a name in Brand settings or your web address counts. Failed answers don't count.

Good: Higher than your rivals. Compare with them, not with a fixed number.

More about Visibility

Share of voice

Your share of all brand mentions in the answers, against your rivals.

How it works: Your mentions ÷ mentions of you plus the rivals shown. Each answer counts each brand once.

Good: Higher than your top rival's share.

Position

Where you're usually named among the brands an answer lists. 1 means named first.

How it works: Your average place in the order brands first appear. Only answers that name you count. A brand gets a rank once 3 or more answers place it.

Good: Close to 1. Lower is better.

Tone

Of the answers naming you, the share that speak about you positively.

How it works: A fixed list of praise and criticism words is checked around your name. No AI judges this.

Good: 65%+ Positive tone, 55–64% Healthy, 45–54% Warning, below 45% Crisis.

More about Tone

Cited as a source

The share of answers that link to a page on your own website.

How it works: Each answer's links are checked against your web address and, if you turn it on, its subdomains. Every answer counts.

Good: Higher than your rivals.

Site score

How readable your website is to AI, out of 100.

How it works: Site checks weighted by importance. A pass scores full, a warning half, a fail zero.

Good: 100 means every check passes.

Overall score

One 0–100 summary of your AI standing, shown on the report.

How it works: A blend: named at all 30%, site score 25%, named early 15%, tone (an AI's overall read) 15%, against your top rival 15%.

Good: 70+ Strong, 40–69 Catching up, under 40 At risk.

More about Overall score

Key insights

Three sentences on what moved in your latest run.

How it works: Built by fixed rules from your stored numbers. No AI writes them.

Where AI gets its answers

Sources

The websites AI answers linked to.

How it works: The links each AI model returned, grouped by site and by kind of site.

Good: Your site, or sites that list you, near the top.

More about Sources

Publishers to pursue

Listing sites the answers mention but don't link to yet.

How it works: Directories, review, map, social and community sites named in the answers. Your own site, rivals' sites and any site an answer already links to drop out.

Good: A shrinking list.

What we ask

Prompts

The questions we ask AI models for you, worded the way a customer would type them.

How it works: Written by you, or written with our AI: press Write with AI on Create prompts with AI.

Price: see Credit & limits.

More about Prompts

Topics

Groups of prompts by subject: six ready-made, plus your own.

How it works: The ready-made ones are Competitive, Best-of, Pricing, Trust & reviews, Problem → solution and About you. They can't be renamed or deleted.

Stage

Where the person asking is in buying: Awareness, Consideration or Purchase.

How it works: Set automatically from the prompt's wording.

Location

A city written into a prompt.

How it works: We don't ask from that place. The city only appears in the prompt.

Rivals

Other businesses the answers recommend alongside you.

How it works: Each run, an AI read of the answers picks out up to 25. Brands you mark Ignore drop out.

Good: You ahead of them on Visibility and Share of voice.

More about Rivals

AI models

The AI assistants we ask, sold as three model sets: Standard (4), Pro (8) and Hyper (12).

How it works: Each run asks every model in your set every tracked prompt. Standard is ChatGPT, Gemini, Meta AI and Google AI Overviews.

Answer

One AI model's reply to one prompt in one run.

How it works: Stored word for word, with its links. Every number counts these.

Running and paying

Run

One round of asking every tracked prompt to every chosen AI model.

Price: see Credit & limits.

More about Run

Budget & Spending

Your schedule: which prompts, which AI models, how often, and an optional monthly cap.

How it works: On Budget & Spending, Start schedule turns it on and Run now runs it once. Expected per month is one run's cost × runs a month.

Good: Your wallet covers the expected month.

More about Budget & Spending

Expected per month

What your schedule should cost in a month, before any monthly cap.

How it works: One run's cost × the runs a month your frequency makes. Weekly is 4.35 runs a month.

Good: Your wallet covers it.

More about Expected per month

Run now

Starts one run of your schedule, right away. The price shows on the button first.

How it works: It counts as the next run of your schedule, so you never pay twice.

Wallet

Prepaid credit that every run and fix check is paid from.

How it works: Each prompt is charged as it's asked; failed ones are refunded. An empty wallet pauses your schedule until you top up.

Good: Enough for your expected month on Budget & Spending.

More about Wallet

Improving

Improve

Your visibility run by run, what to improve, and one click to your AI.

How it works: Measure, create, confirm it's live, measure again. Your AI makes the changes in five sections: Articles & pages, FAQ, Website fixes, Listings & mentions and Corrections. remake confirms each one is live when your AI reports it. Built from your stored runs, so opening it costs nothing.

More about Improve

To-do

Your list of changes to make, highest priority first.

How it works: Prompts where a rival is named and you aren't, sites cited instead of you, failed site checks and page ideas.

Good: A shrinking list.

More about To-do

Ready-made changes

Text and code for your site, written by the brand's first run, that your AI works from.

How it works: Page title, page description, intro paragraph, business information tags, questions and answers and the AI summary file. Your AI makes the change, then remake confirms it's live.

Improve's marks

Where each improvement stands: To do, With your AI, Not on your site or Live.

How it works: The mark shows what remake measured. It turns Live when remake sees the change on your site. It turns Not on your site when your AI reported it and remake did not find it.

Good: Every row Live, in History.

Confirm it's live

When your AI reports a change done, remake reads your live page and confirms the change is there.

How it works: It runs at once, and again at your next run.

Good: Live on your site.

Price: see Credit & limits.

Perception

How AI describes you, and which statements contradict the facts you've entered.

How it works: Each run, an AI read compares the answers with your facts.

Good: No wrong claims.

Your account

Workspaces

A workspace holds one or more brands. A brand is one business with its own prompts, rivals and schedule.

How it works: Add brand on the Workspaces page starts a new one.

More about Workspaces

Activity

A timeline of runs, alerts and fixes as they happen.

More about Activity

Report

A shareable page and PDF of one run, carrying your logo if you add one.

How it works: Anyone with the link can open it, no sign-in needed.

More about Report