Umetna inteligenca za neodvisne hotele · Slovenija

Vemo, kako poln bo vaš hotel, preden se to zgodi.

Naša umetna inteligenca vsak dan prebere vaše rezervacije, pretekle sezone, dogodke v vaši destinaciji in trg. Iz tega napove zasedenost in pravo ceno za vsako noč do 18 mesecev vnaprej ter vam vsak ponedeljek pove, kaj to pomeni za vas. Za hotele z 20 do 80 sobami in apartmajske ponudnike.

Brezplačna analiza preteklih sezon Kaj napovemo

Pisna analiza v enem tednu. Brez obveznosti.

Odločitve brez podatkov

Vsak hotel ima podatke za napoved. Skoraj nihče jih ne uporablja.

Zgodovina ostaja v PMS

Tri leta rezervacij povedo, kdaj se vaš hotel polni, od kod prihajajo gostje, koliko dni vnaprej rezervirajo in kdaj odpovedo. Ti podatki ležijo v sistemu, odločitve pa se sprejemajo po občutku.

Dogodki presenetijo vsako leto

Koncert, sejem, kongres, kolesarska dirka ali šolske počitnice v Nemčiji napolnijo destinacijo vsako leto ob istem času. Brez merjenja njihovega vpliva se prodaja po isti ceni kot deževni torek.

Zasedenost je znana šele, ko je prepozno

Ali je 41 % za november dobro ali slabo? Odgovor je odvisen od tega, kje ste bili lani ob istem dnevu in kako se je ta mesec na koncu končal. Brez primerjave hitrosti polnjenja je številka brez pomena.

Popusti na kanalih se seštevajo neopazno

Mobilna cena, Genius, tedenski popust in promocija za zadnji hip skupaj odnesejo 15 do 25 odstotkov cene. V podatkih se to vidi takoj. Na računu šele konec leta.

Kaj napovemo in analiziramo

Tri analize, ki jih dobite vsak teden, iz vaših podatkov in trga

Napoved zasedenosti

Za vsak datum do 18 mesecev vnaprej: koliko sob bo prodanih in po kakšni povprečni ceni. Napoved temelji na vaši hitrosti polnjenja v preteklih letih, trenutnem stanju rezervacij, odpovedih in tržnih podatkih destinacije. Vsak teden jo preverimo za nazaj.

Analiza dogodkov

Koledar dogodkov za vašo destinacijo: koncerti, sejmi, kongresi, športne prireditve, prazniki in počitnice v državah, od koder prihajajo vaši gostje. Za vsak dogodek izmerimo, koliko povpraševanja je prinesel v preteklih letih in koliko je vredno dvigniti ceno.

Analiza preteklih sezon

Kje ste prodajali prepoceni in kje predrago. Kateri kanali prinašajo goste in po kakšni neto ceni. Kako dolgo vnaprej rezervirajo vaši gostje, koliko odpovedo in katere dolžine bivanja se najbolje obnesejo. Vse v jeziku, ki ga razume lastnik, ne analitik.

Tedenska napovedPonedeljek, 12. oktober 2026Hotel · 34 sob

November se polni hitreje kot lani. Dve priložnosti, ena nevarnost.

Zasedenost za november, po dnevih pred mesecemta teden: 41 % na knjigah, napoved konca 68 %

letos, na knjigahnapoved za letoslani, isti dnevi

November: 41 % na knjigah, lani ob istem dnevu 36 %, lani končna zasedenost 61 %. Napoved za letos je 68 %. Cene za vikende lahko dvignete za 8 %, med tednom jih pustite.
Konec tedna 20.–22. 11.: sejem v Ljubljani. V zadnjih dveh letih je prinesel 23 % več povpraševanja v vaši destinaciji. Predlagamo +15 % in minimalno bivanje 2 noči.
Silvestrovo: rezervacije prihajajo 14 dni prej kot lani, največ iz Italije in Avstrije. Priložnost za paket s tremi nočmi namesto dveh.
Nevarnost: odpovedi za december so pri 19 %, lani 11 %. Vzrok je brezplačna odpoved do 1 dan prej na Bookingu. Predlagamo 3 dni.
Od 28. 12. naprej cene niso objavljene. 7 % lanskih rezervacij za januar je prišlo v oktobru. Te rezervacije zdaj gredo drugam.

Primer. Takšno napoved prejmete vsak ponedeljek, s podatki vašega hotela.

Kako deluje

Od vašega izvoza rezervacij do prve napovedi v treh tednih

11 teden

Brezplačna analiza preteklih sezon

Pošljete nam izvoz rezervacij iz vašega PMS za zadnji dve ali tri leta. Dobite pisno analizo: kdaj in kje ste prodajali pod ceno, kateri kanali in dogodki so prinesli goste, kako hitro se polnite v primerjavi s trgom. Ne glede na to, ali nadaljujemo skupaj.

22 tedna

Model za vaš hotel

Povežemo vaš PMS ali uredimo redni izvoz. Zgradimo napoved za vaš objekt: zgodovina, hitrost polnjenja, koledar dogodkov za vašo destinacijo, tržni podatki. Napoved najprej preverimo na lanskem letu, da vidite, kako natančna je, preden ji zaupate.

3vsak ponedeljek

Tedenska napoved in priporočila

Napoved zasedenosti za naslednjih 18 mesecev, dogodki v naslednjih 90 dneh, priporočene cene in omejitve. Spremembe izvedete vi v svojem sistemu ali jih na vašo željo vnesemo mi. Ko se zgodi kaj nenavadnega, pokličemo.

4vsak mesec

Mesečna analiza

Zasedenost, povprečna cena in RevPAR v primerjavi z lanskim letom in s trgom. Kako natančna je bila napoved prejšnjega meseca. Ena stran, ki jo lahko pokažete banki, partnerju ali lastniku.

Zakaj mi

Smo ekipa, ki vodi Juliana Collection. Napovedi, ki jih prodajamo, uporabljamo vsak dan za lasten denar.

17 enot

Bled, Ljubljana, Kranjska Gora, Portorož in München. Vsako ceno v portfelju določa naša napoved, 365 dni v letu.

90–93 %

Zasedenost naših apartmajev na Bledu v septembru 2026 ob tržnem povprečju 84 %. Vir: PriceLabs tržni podatki, 2. 10. 2026.

vsak dan

Naši AI agenti vsak dan preberejo rezervacije, odpovedi, dogodke in nastavitve kanalov ter pripravijo predloge. Vsako priporočilo pred pošiljanjem pregleda človek.

RentlioMewsApaleoBooking.comAirbnbExpediaExcel izvoz

Vašega PMS ni na seznamu? Zadostuje redni izvoz rezervacij. Dinamično orodje za cene ni potrebno; če ga imate, ga upoštevamo.

Cenik

Prvi korak je brezplačen, ceno nadaljevanja poznate vnaprej

Korak 1

Analiza preteklih sezon

0 €

Zadnja 2–3 leta rezervacij iz vašega PMS
Kanali, dogodki, hitrost polnjenja, odpovedi
Pisna analiza v enem tednu

Korak 2

Model za vaš hotel

Po analizienkratno

Povezava s PMS ali redni izvoz
Koledar dogodkov za vašo destinacijo
Napoved, preverjena na lanskem letu
Odvisno od števila sob in kanalov

Korak 3

Tedenske napovedi

Po analizina mesec

Napoved zasedenosti in cen vsak ponedeljek
Dogodki in priporočila za naslednjih 90 dni
Mesečna analiza s preverjanjem napovedi
Telefon, ko se zgodi kaj nenavadnega
Po želji: vnos cen in samodejna sporočila gostom
Ali namesto tega: delež od rasti prihodka glede na lansko leto

Ponudbo za korak 2 in 3 dobite skupaj z analizo. Odvisna je od števila sob in kanalov, brez obveznosti. Cene brez DDV.

Za koga

Najbolje delamo tam, kjer cene danes določa lastnik po občutku

Za vas, če ste

neodvisen hotel z 20 do 80 sobami brez revenue managerja,
apartmajska hiša ali aparthotel z 10 do 50 enotami,
manjša skupina z dvema do petimi objekti pod istim lastnikom,
objekt z vsaj dvema letoma zgodovine rezervacij.

Ne za vas, če

že imate revenue managerja in sistem RMS z napovedmi. Tega ne bomo izboljšali.
imate manj kot 10 enot. Podatkov je premalo za zanesljivo napoved; raje vam pokažemo, kako do dobrega orodja.
odpirate nov objekt brez zgodovine. Lahko pomagamo z analizo trga, napovedi pa še ne.

Pogosta vprašanja

Kar nas vprašajo pri prvi kavi

Katere podatke potrebujete?

Izvoz rezervacij iz vašega PMS za zadnji dve ali tri leta: datumi, cena, kanal, datum rezervacije, odpovedi. Brez imen gostov. Pri Rentliu, Mewsu in Apaleu se povežemo neposredno, drugje zadostuje mesečni izvoz v Excel.

Kako natančne so napovedi?

Vsako napoved najprej preverimo na vašem lanskem letu, preden jo začnete uporabljati. Potem vsak mesec v poročilu pokažemo, kaj smo napovedali in kaj se je zgodilo. Napoved, ki se ne preverja, ni vredna nič.

Uporabljate umetno inteligenco?

Da. Napovedni modeli berejo vašo zgodovino in trg, naši AI agenti pa vsak dan pregledajo nove rezervacije, odpovedi, dogodke in nastavitve kanalov ter pripravijo priporočila. Vsako priporočilo pred pošiljanjem pregleda človek. Tako upravljamo tudi lastne enote.

Ali potrebujem PriceLabs ali podoben program?

Ne. Napoved in priporočila delujejo z vašim obstoječim sistemom. Cene vnesete vi ali na vašo željo mi. Če orodje za dinamične cene že imate, ga upoštevamo in nastavimo tako, da sledi napovedi.

Kdo spreminja cene?

Vi, na podlagi priporočil, ali mi, če nam date omejen dostop. Spodnje in zgornje meje vedno potrdite vi. Nič ne gre pod vašo spodnjo mejo.

Kako je s podatki gostov?

Za napoved ne potrebujemo osebnih podatkov. Delamo z datumi, cenami, kanali in državo gosta. Če dodate sporočila gostom, je pogodba o obdelavi po GDPR del sodelovanja.

Začnite z analizo

Pošljite nam ime hotela in izvoz rezervacij za zadnji dve leti. V enem tednu dobite pisno analizo.

Zahtevaj brezplačno analizo info@predictionlabs.si +386 31 792 365

Odgovorimo v dveh delovnih dneh. Če izvoza ne znate narediti, vam pokažemo po telefonu, traja pet minut.

AI for independent hotels · Slovenia

We know how full your hotel will be before it happens.

Our AI reads your reservations, past seasons, the events in your destination and the market every day. From that it predicts occupancy and the right price for every night up to 18 months ahead, and tells you every Monday what it means for you. For hotels with 20 to 80 rooms and apartment operators.

Free analysis of your past seasons What we predict

Written analysis within one week. No commitment.

Decisions without data

Every hotel has the data to forecast. Almost nobody uses it.

History stays in the PMS

Three years of reservations tell you when your hotel fills, where guests come from, how far ahead they book and when they cancel. That data sits in the system while decisions are made by feel.

Events surprise you every year

A concert, a trade fair, a congress, a cycling race or German school holidays fill the destination at the same time every year. Without measuring their effect, those nights sell at the same rate as a rainy Tuesday.

Occupancy is known only when it is too late

Is 41% for November good or bad? It depends on where you were on the same day last year and how that month ended. Without comparing booking pace, the number means nothing.

Channel discounts stack unnoticed

Mobile rate, Genius, weekly discount and a last-minute promotion together take 15 to 25 percent off the price. In the data it shows immediately. On the bank statement, at year end.

What we predict and analyse

Three analyses you get every week, from your data and the market

Occupancy forecast

For every date up to 18 months ahead: how many rooms will sell and at what average rate. Built on your booking pace in past years, the current state of reservations, cancellations and market data for the destination. Checked against reality every week.

Event analysis

An event calendar for your destination: concerts, fairs, congresses, sports events, public and school holidays in the countries your guests come from. For each event we measure how much demand it brought in past years and how much a price lift is worth.

Past-season analysis

Where you sold too cheap and where too dear. Which channels bring guests and at what net rate. How far ahead your guests book, how often they cancel, which lengths of stay work best. All in language an owner understands, not an analyst.

Weekly forecastMonday, 12 October 2026Hotel · 34 rooms

November is filling faster than last year. Two opportunities, one risk.

Occupancy for November, by days before the monththis week: 41% on the books, forecast finish 68%

this year, on the booksforecast for this yearlast year, same days

November: 41% on the books, 36% on the same day last year, last year finished at 61%. Forecast for this year is 68%. Weekend rates can go up 8%, leave weekdays alone.
Weekend 20–22 Nov: trade fair in Ljubljana. In the last two years it brought 23% more demand to your destination. We suggest +15% and a 2-night minimum stay.
New Year's Eve: bookings are arriving 14 days earlier than last year, mostly from Italy and Austria. An opening for a three-night package instead of two.
Risk: December cancellations are at 19%, last year 11%. The cause is free cancellation up to 1 day before on Booking. We suggest 3 days.
No prices published from 28 Dec onward. 7% of last year's January bookings arrived in October. Those bookings are now going elsewhere.

Example. You receive a forecast like this every Monday, with your hotel's data.

How it works

From your reservation export to the first forecast in three weeks

11 week

Free analysis of past seasons

You send us a reservation export from your PMS for the last two or three years. You get a written analysis: when and where you sold below value, which channels and events brought guests, how fast you fill compared with the market. Whether or not we continue together.

22 weeks

A model for your hotel

We connect your PMS or set up a regular export. We build the forecast for your property: history, booking pace, an event calendar for your destination, market data. We first test the forecast on last year so you see how accurate it is before you trust it.

3every Monday

Weekly forecast and recommendations

Occupancy forecast for the next 18 months, events in the next 90 days, recommended rates and restrictions. You apply the changes in your own system, or we enter them for you if you prefer. When something unusual happens, we call.

4every month

Monthly analysis

Occupancy, average rate and RevPAR against last year and the market. How accurate last month's forecast was. One page you can show your bank, your partner or your owner.

Why us

We are the team behind Juliana Collection. The forecasts we sell, we run on our own money every day.

17 units

Bled, Ljubljana, Kranjska Gora, Portorož and Munich. Every price in the portfolio is set by our forecast, 365 days a year.

90–93 %

Occupancy of our Bled apartments in September 2026 against a market average of 84%. Source: PriceLabs market data, 2 Oct 2026.

every day

Our AI agents read reservations, cancellations, events and channel settings every day and draft recommendations. A person reviews each one before it is sent.

RentlioMewsApaleoBooking.comAirbnbExpediaExcel export

Your PMS is not on the list? A regular reservation export is enough. No dynamic pricing tool is required; if you have one, we work with it.

Pricing

The first step is free, and you know the price of the rest up front

Step 1

Past-season analysis

€0

Last 2–3 years of reservations from your PMS
Channels, events, booking pace, cancellations
Written analysis within one week

Step 2

A model for your hotel

After analysisone-off

PMS connection or regular export
Event calendar for your destination
Forecast tested on last year
Depends on room count and channels

Step 3

Weekly forecasts

After analysisper month

Occupancy and rate forecast every Monday
Events and recommendations for the next 90 days
Monthly analysis with forecast accuracy
A phone call when something unusual happens
Optional: rate entry and automated guest messages
Or instead: a share of revenue growth over last year

You receive the quote for steps 2 and 3 together with the analysis. It depends on room count and channels, with no obligation. Prices exclude VAT.

Who it is for

We do our best work where the owner sets prices by feel today

For you if you are

an independent hotel with 20 to 80 rooms and no revenue manager,
an apartment house or aparthotel with 10 to 50 units,
a small group of two to five properties under one owner,
a property with at least two years of reservation history.

Not for you if

you already have a revenue manager and an RMS with forecasting. We will not improve on that.
you have fewer than 10 units. Too little data for a reliable forecast; we would rather point you to a good tool.
you are opening a new property with no history. We can help with market analysis, but not yet with a forecast.

Questions

What we get asked over the first coffee

What data do you need?

A reservation export from your PMS for the last two or three years: dates, rate, channel, booking date, cancellations. No guest names. With Rentlio, Mews and Apaleo we connect directly; elsewhere a monthly Excel export is enough.

How accurate are the forecasts?

We first test every forecast on your last year before you start using it. Then every month the report shows what we predicted and what happened. A forecast that is not checked is worth nothing.

Do you use AI?

Yes. Forecasting models read your history and the market, and our AI agents review new reservations, cancellations, events and channel settings every day and draft recommendations. A person reviews each one before it is sent. This is how we run our own units.

Do I need PriceLabs or a similar tool?

No. The forecast and recommendations work with your existing system. You enter the rates, or we do if you prefer. If you already have a dynamic pricing tool, we work with it and set it to follow the forecast.

Who changes the prices?

You, based on the recommendations, or we do if you give us limited access. You always approve the floors and ceilings. Nothing goes below your floor.

What about guest data?

The forecast needs no personal data. We work with dates, rates, channels and the guest's country. If you add guest messaging, a GDPR data processing agreement is part of the engagement.

Start with the analysis

Send us your hotel's name and a reservation export for the last two years. You get a written analysis within a week.

Request a free analysis info@predictionlabs.si +386 31 792 365

We reply within two working days. If you don't know how to make the export, we show you over the phone, it takes five minutes.