The guest has the answer before they can ask — and you write nothing

The aim is not to hide the automation but to avoid the situation where somebody waits for a reply nobody is writing. Messages go out under your name, in your wording and your tone — they are templates you write once, or that we hand you ready for review. Most of them are messages you would have sent word for word anyway: address, codes, times, rubbish. If a guest asks something no template covers, the question reaches you and you answer it yourself. In practice guests notice no difference, because they get their answer faster than before — and a fast reply is one of the most common reasons a five-star review gets written at all.

Email is the backbone, because it arrives even when the guest is abroad with data roaming switched off. On top of that, the time-critical messages — arrival instructions, codes, the checkout reminder — can also go by SMS, because those need to be findable on a phone quickly and in poor signal. If the booking came from a platform that requires conversation inside its own environment, the thread continues there but still lands in the same view for you, so you are not keeping three apps open. The channel is chosen per message type: substantial guides by email, short time-critical items by text. That way the important information neither drowns in a thread nor gets missed.

Yes. Templates exist in English and Estonian, and the guest receives the one matching the language they chose at booking or that comes from their profile. The same texts can be added in further languages when your guest mix clearly justifies it — Finnish or German, for instance. What matters is not the number of translations but that the instruction is unambiguous: short sentences, concrete times and a photo of the entrance travel across a language barrier far better than a long descriptive paragraph. If you already have your own texts, we use those as the basis and keep your wording rather than forcing you to adopt ours.

The automation covers the repetitive part, not the exceptions. If a guest writes something the templates do not cover — a leak, a broken appliance, a late arrival, a dispute — it reaches you as an alert and the conversation waits for a real person. You see the whole thread in one place alongside the booking, the codes and the door log, so you have the context without hunting for it. The same works in reverse: if a sensor reports a leak or the door log shows something unexpected, that can trigger a message to the guest or the cleaner. The point of the automation is to keep your attention free for exactly these cases, not to pretend they do not exist.

The number of reviews is usually decided not by how much someone wants to praise you but by whether the request arrives at the right moment. Within a couple of hours of checkout the impression is fresh and the phone is in hand; three days later the person is back at work and the review never happens. The automated thank-you always goes out in that same window and always with a direct link, so there is one click involved rather than a search. The same logic works the other way: if something went wrong during the stay, it reaches you before it reaches a public review, and you have a chance to put it right. Together that means more reviews and fewer surprises in what they say.