Software Deep-Dive · April 2026 · RevParGenius Intelligence · Based on Cloudbeds technical documentation
Cloudbeds quietly retired its old pricing engine and replaced it with something significantly more sophisticated. They are calling it Revenue Intelligence, powered by a technology called Signals AI. Most hoteliers have seen the marketing. Far fewer understand what it actually does, what data it touches, or whether the numbers are real.
This is the plain-English breakdown — what the algorithm does under the hood, what it needs from your property to work, how ROI is calculated, and the honest caveats you should know before deciding if this is right for you.
The Headline Numbers (Cloudbeds Self-Reported)
⚠ Important: All headline metrics are self-reported by Cloudbeds. No independent third-party audit of these figures currently exists. Treat as directional, not guaranteed.
What Changed — And Why It Matters
Cloudbeds used to run something called PIE — the Price Intelligence Engine. It was a rule-based rate shopper: it watched what competitors charged on OTAs, compared that to your occupancy, and adjusted rates based on rules you or Cloudbeds set. Useful, but fundamentally a rearview mirror. It told you what the market had done, not what it was about to do.
PIE is fully retired. Its URL now redirects to the Signals-powered Revenue Intelligence page. If you are still thinking of Cloudbeds as a rate shopper with rules, that product no longer exists.
How the Algorithm Actually Works
This is the part most vendor marketing glosses over. Here is what the documentation actually says.
Concept 1 · Causal AI vs. Correlation AI
The difference between "this happens when..." and "this happens because..."
Most older revenue management tools are built on correlation. They observe patterns: bookings go up in July, weekends perform 20% better, rain events drop occupancy. Then they react to those patterns with rules. The problem is patterns break — a global event, a new competitor, a viral TikTok of your city, and the model is lost because the pattern no longer holds.
Signals uses causal machine learning — developed by Cambridge and Oxford mathematicians — which instead asks: why do bookings go up in July? Is it school holidays? A local festival? Weather? Flights? It identifies the actual cause and weights it. This means when a disruption hits, the model doesn't break — it already understands the underlying mechanism, not just the surface pattern. It also enables "what-if" scenario planning: drop rates 10% on these dates and see the predicted occupancy impact before you do it.
Concept 2 · The Time Surface Model
Why looking at one date at a time misses half the picture
Traditional revenue management tools look at pickup curves — they track how bookings for a single future date are accumulating over time. Useful, but limited. It is a one-dimensional snapshot that cannot see what is happening on adjacent dates or across booking windows at the same time.
Old method: 1D Pickup Curve
Tracks bookings for one date over time. Cannot see surrounding dates, adjacent weeks, or how a rate change ripples forward. A snapshot.
Signals: 2D Time Surface
Tracks stay date × lead time simultaneously as a 2D surface. Sees momentum shifts, ripple effects, and how filling one date influences adjacent dates.
In plain language: if a conference fills your Thursday and Friday, the 2D model sees that pressure and can predict the Saturday overspill demand before it shows up in bookings. The 1D model waits until it sees actual bookings start accumulating for Saturday. That lag is where revenue gets left behind. Cloudbeds claims this approach produces up to 54% lower forecast error compared to traditional methods.
What Data Does It Use?
This is where data transparency matters. The system pulls from three distinct buckets.
1 · Your Own Data
From your PMS (Cloudbeds)
- Reservations & booking history
- Room types, occupancy, ADR, RevPAR
- Guest profiles & segments
- Cancellation & modification patterns
2 · Market Data
External competitive intelligence
- Competitor rates from Booking.com, Expedia, Hotels.com
- Metasearch search traffic signals
- Flight booking data
3 · Forward Signals
What is about to happen
- Local events & festivals
- Weather patterns
- Your website visitor behaviour
- Review sentiment (Google, Booking.com)
On Data Transparency
The documentation is clear about what categories of data the system uses, but does not publish the exact weighting between them. You will not see a dashboard that says "events are weighted at 32%, competitor rates at 28%, your pickup curve at 40%." That logic is proprietary. This is standard for enterprise AI tools, but worth knowing: you are trusting the model's reasoning, not auditing it directly. Cloudbeds does offer "what-if" scenario outputs so you can test the model's sensitivity to specific inputs — which is more transparency than most standalone RMS tools offer.
The Feature Most Hoteliers Miss
Cloudbeds calls this "Revenue Marketing" — and it is genuinely uncommon in standalone revenue management software.
What Revenue Marketing Actually Means
When the pricing algorithm identifies a demand gap — a date window where bookings are under-pacing — it does not just lower the rate and wait. It automatically triggers a CRM email campaign to past guests who match the profile for that period. Fill the gap with direct bookings, at no OTA commission.
Most revenue management tools and marketing tools sit in separate silos. A revenue manager adjusts rates; a marketing manager sends campaigns; neither sees what the other is doing in real time. Revenue Intelligence connects these two actions in a single automated loop. That is what the +25% direct booking growth figure is pointing at.
The Bespoke Hotels case study in the documentation shows this working at scale across 40 properties, generating over £1 million in email-driven revenue after starting from zero guest contact data. The system collected guest contacts via Wi-Fi login, segmented them by stay type, and auto-generated campaigns targeting demand gaps. That loop — data collection to segmentation to targeted campaign to filled gap — is what the £1M figure represents.
How They Measure ROI — and What You Should Watch
The headline +18% RevPAR uplift claim is the number most hoteliers will anchor on. Here is how to read it properly.
The 90-day measurement window
Cloudbeds measures RevPAR uplift at the 90-day mark post-activation. This is a reasonable window — long enough to capture seasonal effects and booking lead times, short enough to be credible as a sales claim. The Mercure Paddington case study, for example, shows 93% occupancy achieved vs. a 64% comp average — a sustained gap, not a one-week spike.
These are self-reported numbers
No independent audit of the 95% accuracy, +18% RevPAR, or +25% direct booking claims currently exists. That does not mean they are wrong — it means you should run your own baseline before activation and measure your own property's before-and-after. Ask Cloudbeds for the methodology of how they calculated RevPAR uplift for your comparable set, not just the headline figure.
The baseline comparison matters enormously
The documentation notes this tool is best suited for properties with 20+ rooms at 50–60%+ occupancy that are still pricing manually. A hotel moving from manual spreadsheet pricing to algorithmic pricing will likely see significant gains simply because the baseline was so low — that gain is real, but attributing all of it to the AI sophistication is misleading. The right comparison is against another automated RMS, not against manual pricing.
Case Study · London
Mercure Paddington
Signals identified overconfidence in a local event, rivals gaining OTA visibility, and a review surge suppressing conversions — and triggered targeted recovery actions.
Case Study · 40 Properties
Bespoke Hotels
Started from zero guest data post-acquisition. Signals collected contacts via Wi-Fi, segmented by coastal/rural/seasonal, identified demand gaps, and auto-generated personalised email campaigns.
Is This Right for Your Property?
Cloudbeds is direct about who this is built for. Revenue Intelligence is a paid add-on — it is not included in base Cloudbeds plans.
✅ Good Fit
- 20+ rooms
- 50–60%+ average occupancy
- Currently pricing manually or via basic rules
- Already using Cloudbeds as your PMS
- Want pricing and CRM to work together
- Independent or small group (up to ~40 properties)
⚠ Consider Carefully
- Under 20 rooms (cost may not justify)
- Below 50% occupancy (not enough data volume)
- Already running a dedicated RMS (IDeaS, Duetto)
- On a different PMS — this requires Cloudbeds as the base
- Need transparent algorithmic logic vs. proprietary black-box
Third-Party RMS Option
If you want a dedicated RMS but are on Cloudbeds, you can connect Climber RMS (via Revenue Analytics) through the Cloudbeds marketplace. This gives you the option of a specialist revenue management layer without abandoning your PMS. The ecosystem supports both native Revenue Intelligence and third-party tools simultaneously.
The RevParGenius Take
The technology is genuinely more sophisticated than what it replaced. Causal AI is a meaningful step up from correlation-based rules, and the Time Surface model is a legitimate forecasting innovation — the 54% error reduction claim is directionally credible even if unaudited. The Revenue Marketing loop (pricing triggers CRM email) is the most commercially interesting feature and the hardest for competitors to replicate, because it requires owning both the PMS and the guest data layer simultaneously.
The caveat to keep front of mind: all performance numbers are vendor-reported, the target is hotels still pricing manually (a low bar for improvement), and the system is a proprietary black box that you trust rather than audit. That is not disqualifying — it is what most enterprise software is — but it means your own before-and-after measurement discipline matters more than Cloudbeds' headline statistics.
Bottom Line for Hoteliers
If you are on Cloudbeds, pricing manually, and have 20+ rooms running above 50% occupancy — the ROI case for Revenue Intelligence is strong. The combination of causal AI pricing and automated direct booking campaigns addresses two revenue leakage problems at once.
If you are on a different PMS, already running IDeaS or Duetto, or below the occupancy threshold — evaluate Climber RMS as a standalone option or revisit when your data volume justifies the investment. The tool is only as good as the historical data it has to learn from.
Free Hotel Market Intelligence
Want to Know What Your Market Is Actually Doing?
We run live OTA and STR demand analyses for APAC hotel markets — before you invest in any revenue management tool, start with knowing what your comp set is actually doing today.
Request Your Free Demand AnalysisNo commitment. Just live data for your market at revpargenius.com
Sources: Cloudbeds technical documentation (April 2026), PhocusWire, Hospitalitynet.org. All performance metrics are self-reported by Cloudbeds — no independent audit exists. RevParGenius is an independent hotel market intelligence platform and is not affiliated with Cloudbeds or any named software provider.
