Your visitors are asking questions your site never answers.
And you never find out which ones. Paste your address: we read your public pages and show you the exact questions that come back empty — plus a working assistant built from those same pages.
- No account, nothing installed
- Public pages only, robots.txt obeyed
- Yours to keep or ignore
How much does a first appointment cost?
related pages exist, none answer it
Do you take walk-ins on Saturday?
nothing on the site is close
Is there parking at the clinic?
nothing on the site is close
A real report, from a real crawl of a real site. Yours takes about a minute.
Analytics shows you a bounce. It never shows you the question.
- ·How many people came, and from where.
- ·Which pages they opened, and for how long.
- ·The questions patient enough to reach your inbox.
- ×The question they typed into your search box and got links for.
- ×The one they scrolled three pages looking for and gave up on.
- ×The one that decided them against buying, unwritten anywhere.
That last group never contacts you. They are not annoyed enough to complain and not patient enough to hunt — they close the tab, and in your dashboard it looks exactly like someone who found what they wanted. This page reads your own pages and puts the missing questions in writing, before you install anything.
Two things on your site, and a list of what is missing from it
An assistant that answers from your pages
It replies in the visitor’s own words, using what HootyBot read on your site and nothing else — and names the page it took the answer from, so anyone can check it.
A search that returns answers, not links
The same index behind a search field in your page. Ranked pages of your own site with a preview, and a click goes straight there. No conversation.
The four things people ask us first
It cannot invent an answer
The assistant is restricted to the pages we read. When they do not cover a question it says so instead of guessing — and that refusal is exactly what the report is built from.
Every answer is traceable
Each reply names the page it came from. A visitor can open the source; so can you, when you want to know why it said what it said.
You decide what we may read
HootyBot fetches robots.txt first and follows it, one page at a time. Anything it may not fetch never reaches the index, and we do not train models on what we read.
Nothing is installed to try it
The demo runs entirely on our side. No script on your site, no account, no card — if the report is not useful, close the tab and nothing has changed.
How long does the crawl take?
The assistant is answering within seconds of the first page. Reading the rest — up to twenty-five pages for the demo — takes a few minutes, because we deliberately request one page at a time with a pause between them rather than hammering your server.
Will this slow my site down or change anything on it?
No. The demo installs nothing and touches nothing. HootyBot requests public pages the way any visitor’s browser would, one at a time. It never submits a form, logs in, or touches a cart.
What happens to the pages you read?
They become the index behind your demo, and that demo deletes itself after a week if you do nothing. We do not train models on what we read.
Do I need a developer to install it?
One line of HTML before the closing body tag, or the WordPress plugin. If you can paste a tracking code, you can do this.
What does it cost?
Nothing to start. The free plan reads up to 200 pages of one site and answers 100 questions a month, with no card. Pro is $39 a month — 1,000 pages, 2,000 answers, five agents, and your site without our badge on it.
Find out what your site is failing to answer.
About a minute. No account, no card, nothing installed.