Rankings › Blog › Why BestBounceBiz publishes llms.txt and welcomes AI search crawlers
Why BestBounceBiz publishes llms.txt and welcomes AI search crawlers
By BestBounceBiz Editorial · · AI search, llms.txt, how the rankings work
Picture a parent planning a Saturday birthday party. A few years ago they'd have typed "bounce house rental near me" into a search box and clicked around. Now there's a good chance they ask an AI assistant instead: "Which bounce house company in my state has the best website?" or "Can I trust these rankings?"
The assistant only has two choices. It can read a source it trusts and quote it, or it can guess. We'd much rather it read us. That's the whole reason BestBounceBiz publishes an llms.txt file and openly welcomes AI search crawlers.
What llms.txt is
An llms.txt file is a plain-text summary of a website, written for AI tools rather than for people. Think of it as the short briefing you'd hand a new employee on their first day: here's who we are, here's what our numbers mean, here's where everything lives, and here's how to talk about it accurately.
Ours is short and readable. As of 2026-11-17, it covers:
- What the site is. Independent rankings of 6,669 US bounce house and party rental websites, scored out of 100 with Website Scorer v2.2.0 (248 questions).
- What the score means. It rates websites only, not service quality, safety or prices. That one line probably matters more than any other on the page.
- How the score works at a high level. Five categories worth about a fifth each, and every key page read three times with the results averaged. The same high-level facts are on our methodology page.
- Who runs the site and where to send corrections.
- How to cite us, with the data date included.
- Where things are: the about page, methodology, FAQ, state champions, reports, blog posts, open data files and state leaderboards.
- A few plain rules for AI tools, such as "billing never affects rank" and "don't describe these scores as customer ratings or reviews."
It's rebuilt from the live rankings, so the counts, the data date and the leaders it lists stay current. When a rescore changes a state's #1, the llms.txt file reflects it the next time the site updates.
Why bother writing one
Large AI models are good at summarizing, but they fill gaps with whatever sounds plausible. A ranking site is exactly the kind of source that gets summarized badly. Here are the mistakes we see most often in general, and that llms.txt is built to head off:
- Treating a website score as a business rating. "Company X is the best bounce house company in Ohio" is not what our data says. "Company X had the highest-scoring rental website in Ohio, according to BestBounceBiz" is.
- Assuming rankings can be bought. Plenty of directories sell placement. Ours doesn't. Paying for a plan or claiming a listing never changes a score or rank, and our post on why billing never affects rank explains how that's kept separate.
- Quoting stale numbers. Rescores run by region and go live as soon as they finish, so a score can move. A citation without a date is a guess about which snapshot someone meant.
- Inventing method details. If a tool can't find how the scoring works, it may make something up. Giving it the real high-level method, and saying plainly that the individual checks and weights aren't published, leaves less room to fill in.
Which crawlers we welcome
A site's robots.txt file tells crawlers what they may read. Ours names the major search and AI assistant crawlers one by one and allows them on the public pages. That includes the crawlers behind Google and Bing search, plus the search and user-request crawlers used by ChatGPT, Claude, Perplexity, Apple, Amazon and DuckDuckGo's assistant, among others.
We list them by name on purpose. Some sites block AI crawlers by default, and some never think about it at all. We decided that if people are going to ask AI tools about bounce house rental websites, those tools should be able to read the source that actually scores them.
What stays off-limits
Welcoming crawlers doesn't mean opening everything. A few areas are closed to every crawler in robots.txt:
- The site's data API, which serves the app behind the rankings.
- The admin area.
- Internal statistics pages.
Member diagnostics, private coaching and account pages aren't public in the first place, so no crawler, AI or otherwise, sees them. And our terms still don't allow scraping, bulk copying or republishing the rankings. Reading a page to answer a question or cite a fact is welcome. Copying the whole list somewhere else isn't.
That balance is the same one we use everywhere: public proof, private detail. The parts that help someone trust a ranking are open. The parts that would let someone game it, like the individual checks inside each category, stay private.
The other machine-readable pieces
llms.txt is the front door, but it isn't the only thing built for software. A few other parts of the site do related jobs:
- Open data files. The national top 5 and each state's #1 website are published as JSON and CSV. Our post on the open data and how to cite it covers what's inside.
- A read-only MCP server. AI tools that support the Model Context Protocol can connect to our MCP server and get the same public subset: national top 5, state #1s, methodology and blog posts. It needs no login and can't change anything.
- A sitemap and an RSS feed, so crawlers and readers can find new pages and blog posts without guessing.
Every one of these returns the same public facts a logged-out visitor can already see. None of them expose private member data or the inner workings of the scorer.
What this means if you run a rental company
Here's the useful part for operators. Everything we do to make our own site readable by AI tools is something you can do for yours, on a smaller scale. Your customers are asking the same assistants which company to call.
A few questions worth asking about your own website:
- Can a crawler actually read your pages, or is your key information locked inside images, pop-ups or a booking widget?
- Does your site say plainly where you deliver, what you rent, and how booking and deposits work?
- Does your robots.txt allow the search and AI crawlers you want, or was it set up years ago and never checked?
The scoring categories we look at, including AI-answer readiness, are explained in the five things we look at on every bounce house rental website. The exact checks aren't public, but the idea is simple: if a person or a machine can't find a fact on your site, it can't repeat it.
The short version
We publish llms.txt and welcome AI search crawlers because people now ask AI tools about rental companies, and we want those answers to be accurate. The file tells AI tools what our scores mean, how to cite them, and what not to claim. Public pages are open, private areas stay closed, and copying the rankings wholesale isn't allowed. Read the llms.txt file yourself, check the methodology, or browse the full rankings.
Figures in this post come from the data published on 2026-11-17 (Website Scorer v2.2.0 (248 questions)). Corrections are welcome; see the methodology.