The Guest Risk Pre-Screener,
explained.

Every damage story starts with an approved booking. Once the guest is inside, your options shrink to damage control: documentation, claims, deadlines. Screening is the only step that happens while you can still simply say "tell me more first".

The problem is what most screening actually looks at. Instinct reads the profile photo, the name, the writing style, signals that predict nothing and discriminate plenty. The facts that do correlate with parties and damage are dull and behavioural: a guest who lives ten minutes away, a same-day booking for Saturday night, a group that exactly fills the house, a month-old account that skipped the conversation. Paid screening services check these automatically, but they start at real money per month. Most independent hosts and VAs check nothing at all.

The Pre-Screener puts the structured version of that check in front of everyone, free, and keeps the discriminatory signals physically out of it: there is no field where a name, a photo, or an origin could even be entered.

Run it when a request arrives and before you approve, especially on request-to-book, where you have a real decision window. On instant book, run it the moment the confirmation lands: you cannot un-approve, but you can still send the welcome message, confirm the guest count, request ID verification, and prepare the property accordingly.

It is most valuable on the bookings that already feel slightly off: the last-minute Saturday request, the group that exactly matches your maximum, the profile with no history. It is also worth running on bookings that feel fine, precisely because a consistent routine applied to everyone is what makes your screening fair and defensible.

Two design rules keep the sum honest. The combination patterns, full house on a weekend, one night on a weekend, last minute for a weekend, only fire when both halves are actually answered, never on partial data. And answering "not sure" adds zero points anywhere: missing information is surfaced as a lookup to do, not treated as hidden risk.

The bands are fixed: 0 to 24 reads low, 25 to 49 medium, 50 and above high. The theoretical maximum stack, every pattern at once, caps at 100.

The score measures how strongly a booking matches known incident patterns. That is all. It is not a probability that this guest will cause damage, not a background check, and not a judgment of any person. Most flagged bookings turn out completely fine; the score exists to tell you which requests deserve five minutes of verification instead of a reflex approval.

That framing is also the legal one. Platform non-discrimination policies, and in many countries the law, prohibit refusing guests over protected characteristics. The screener is built so those characteristics never enter the process, and every result carries the same reminder: verify booking facts, communicate warmly, and decline only within platform rules when the facts and the conversation justify it. The Guest Screening Guide covers those lines in detail.

The result page shows the score and band, then every triggered pattern as its own card: what fired, why it matters, and the specific pre-approval action that answers it, down to the message worth sending. Below that, any "not sure" answers are listed as quick lookups with a hint on where to find each on the platform.

From there you can share the result with the owner as a link, save it as a PDF, or request the optional AI pre-approval plan. The plan is emailed to you free and includes a ready-to-send guest message written as hospitality rather than suspicion, the checks in priority order, and the specific replies that should make you escalate. A privacy toggle lets you share the result with the AI plan hidden.

Have the booking request and the guest profile open next to the screener. Everything it asks is visible there:

• The request shows the dates, the group size, and whether it was instant book.
• The profile shows the join date, the review count, the verification badge, and usually a home city.
• The message thread shows whether the guest introduced themselves or answered your pre-booking questions.

If your listing settings allow it, turning on pre-booking questions permanently is the single best screening upgrade: it converts every future request into one that arrives with answers already attached.

The score gives you a lane; the follow-through decides the outcome:

• Low risk: approve, send the welcome message with the house rules, and move on. The routine already did its job.
• Medium risk: run the lookups it lists, send the friendly purpose-and-headcount question, and read the reply before approving.
• High risk: verify everything, involve the owner if you manage for one, require ID verification where the platform allows it, and treat a failed conversation as a legitimate reason to decline within platform rules.
• Whatever you decide, document the property before check-in. Screening lowers the odds; the check-in inspection is what turns a bad outcome into a paid claim instead of a loss.