Pinned straw:
Discl: Held IRL 3.71% and in SM
SUMMARY
Have been thinking a bit deeper about how SDR is embedded into a hotel’s (probably crude) tech stack, I am now clear that:
- SDR’s data moat is a prized and formidable one
- The SDR modules are really an extension of, and hence, very much a part of a hotel’s tech stack and system of record - the SDR intermediation layer is not a thin one at all
- The risk of vibe-coded disintermediation of SDR capabilities on the hotel end is low to zero because minus the SDR-wide data intelligence platform, which is closed, any vibe coded capability will add almost no value to a hotel
- The risk of OTA disintermediation or of OTA’s encroaching into the hotel-end is low to zero - I just can’t see the use case/benefit/advantage any of these will bring to an OTA or an OTA-wannabe. If either happens, is likely to be an OTA-end issue. I expect SDR will end up managing the impact on behalf of the hotels via additions/changes to the Smart Distribution/Channels Plus capability.
Based on the thesis review below and how far the SDR price has fallen, I started nibbling to top up today at $3.18. With the ongoing volatility, there is every chance it will dip closer to $2.85, a 2.5 year support area. Will be topping up again closer to that level if it dips that far.
REVIEW OF SDR’S MOAT
How SDR Fits Into A Small Hotel’s Tech Stack
This is how I am now thinking of SDR from a Hotelier’s perspective:
The barebones basic modules that a hotel would likely have before buying SDR would be a:
(1) Property Management System (PMS) to manage the overall hotel operations - this could well be manual or Excel or the various PMS systems available in the market;
(2) Point of Sale system - this is the system to manage any Food & Beverage outlets;
(3) Room Inventory management - some means of managing reservations, room inventory etc - could be a manual reservations book, Excel or is part of the PMS;
SDR modules are in dotted boxes - the intent is not to map out everything that SDR offers, but to highlight the key capabilities that SDR brings to a hotel that it is highly likely not to have, or can afford, as standalone systems, either because of cost, or the lack of expertise to actually use those systems/capabilities. I signed up to the SDR demo’s on their website which has some walkthroughs of the various modules, which I found helpful.
I thus see:
- SDR modules, base and Smart Platform, are either extensions of, or ARE the hotel’s operating system and operational workflows - the extent will depend on what systems the hotel already has/does not have currently;
- Every transaction flows through SDR and hence, SDR becomes part of, or THE, system of record of the property - that is what SDR has positioned itself more clearly recently;
- SDR Smart Distribution and Channels Plus as the hotels distribution connectivity/pipe to the room distribution capabilities of the Online Travel Agents (OTA);
- Revenue/Yield Management via Dynamic Revenue Plus is uniquely SDR - this is what completely differentiates SDR from any other “distribution” company, which is how analysts categorise SDR - I still think this is the biggest market misunderstanding of the capability benefit SDR actually adds to its Hotel customers.
A good example of this is commentary from Ten Cap’s Jun Bei Liu, which I thought, quite honestly, was disappointing. I have no issue with her not wanting to invest in SDR, but the reason for avoidance reflected a poor/superficial understanding of how embedded SDR is, in a hotel’s tech stack. I have no idea what she means by “customer switching is easier” as SDR’s customers are hotels and their capabilities are actually embedded on the hotel-end.
AI Risk, Impact
From an AI disruption standpoint:
1. Because of SDR’s deep embedment into the core systems and workflows of a hotel, SDR is very much part of the hotel’s workflow and trusted system of record. In many cases, it is likely to be THE workflows and trusted system of record.
2. There is no per-seat licensing issue - licensing is per property and, SDR’s revenue is increasingly transaction-based.
3. The SDR Data Intelligence capability, built from millions/billions of transactions across its entire hotel customer base, is a formidable moat - SDR recognised this years ago and the Smart Platform modules was its strategy to monetise the insights, creating its own data flywheel, completely inaccessible from outside the platform.
4. Given the integration of the modules across a hotel’s tech/manual stack, and given the very basic level of IT in smaller hotels, I do not see vibe coding of any of SDR’s modules as a credible threat - the context and input from the data flywheel which continuously enriches the data intelligence simply cannot be replicated in a vibe coding scenario. I also cannot see what incentive there is for a small hotelier to spend any energy vibe coding as running a hotel is essentially a 24x7 endeavour.
5. SDR IS absolutely the neutral Revenue Manager for the Properties - revenue management is fundamentally challenging/previously non-existent, but SDR has found the way to deliver this, cost effectively, at scale, to small properties so they now have some capability they never dreamed of having. And SDR is performing this role with both its own AI agents over its extensive data around bookings, pricing, yields etc. - I don't think public LLM's will get here as the data remains completely closed
6. I cannot see how any possible disintermediation in and around the OTA’s, if it ever happens, will impact SDR - folk like Google have tried and failed. The booking engine itself should be “relatively easy” to vibe code, but the CONNECTIVTY out of the OTA’s and into the millions of hotels around the world, is not something that is easily (1) replicated (2) implemented (3) supported and (4) maintained, on an ongoing basis, never mind issues of security, payments etc. SDR’s Smart Distribution and Channels Plus capabilities were introduced to simplify connectivity to the distribution channels - that this is growing nicely is indicative of how much of an issue this area of the hotel tech stack is.
7. Any OTA-replacement through disintermediation, must still end up being connected to a hotel, either through a dedicated or ad-hoc pipe - this will all continue to be managed by SDR, as it makes no sense for hotelier to bypass SDR and DIY this. I expect that SDR would would appropriately react to any moves around disintermediation and find news ways of connecting to newly created OTA’s such that SDR does the integration once, and all hotels get access to that via Smart Distribution and Channels Plus
8. Similarly, the OTA's themselves could theoretically bypass SDR via AI agents, but then we go back to today's 1 OTA to a gazillion properties big and small and each property has 1 pipe per OTA - you would essentially unwind everything SDR has done to date. I can't see a business case for the OTA's doing this at all, even with agents, for the exact same reasons as in point 6. For OTA's, they only have 1 pipe into SDR and SDR takes care of everything else on the hotel-end. To undo this just because they have an agent to do this, doesn't make sense to me.
From an AI Productivity standpoint, I expect SDR to continue to use AI to mine its Data Intelligence platform and continue to find ways to add new capabilities that will add value to hoteliers, boost internal development productivity etc.
It feels that SDR will absolutely benefit from AI rather than be threatened by it. And it has not been standing still, doing nothing. I have every confidence in Sankar’s leadership that SDR will be AI-ing everything they possibly can and as fast as they can.
The Lakehouse team gave an interesting take on Siteminder's moat - it is a holding in their Small Companies fund - on a webinar today.
I asked Claude to amplify on what I learnt, and was so impressed by 'his' answer that I've reproduced it in full below.
## SiteMinder's Moat Against AI-Enabled Competitors
The core of SiteMinder's defensive position isn't simply that it has good software — it's that it has become the **contractual and technical middleware** between two sides of a marketplace that neither side wants to rebuild from scratch. Understanding why requires unpacking the structure of that marketplace.
### The Two-Sided Network and Why It Matters
SiteMinder sits between accommodation providers (supply) and booking channels (demand). Its integration network spans OTAs like Booking.com and Expedia, GDS networks that put hotels in front of 600,000 travel agents, wholesalers, and metasearch platforms. [SiteMinder](https://www.siteminder.com/integrations/) Each of these relationships isn't a casual API connection — it is a bilaterally negotiated, certified, technically validated integration built to a specific standard.
This creates a classic two-sided network effect: hotels benefit from every additional OTA that integrates with SiteMinder because it opens them up to additional demand with no additional effort, while OTAs benefit from new hotels joining because they can offer more rooms. [Morningstar Australia](https://www.morningstar.com.au/stocks/ask-analyst-can-this-genius-product-create-next-asx-software-king) Neither side wants to defect because defection means losing access to the other side of the marketplace simultaneously
### Moat Layer 1: The Integration Certification Barrier
The single most underappreciated element of SiteMinder's moat is what a new entrant would actually have to do to replicate it. Each OTA and GDS integration isn't just a matter of writing code against a public API. New partners must execute SiteMinder's certification test scenarios to validate that everything works as expected, including checking edge cases and verifying the integration meets all standards — a process that can take weeks, with pilot environments required before any live production connection is established. [Siteminder](https://developer.siteminder.com/siteminder-apis/integration-process)
Multiply that certification burden across 450+ OTA channels, 350+ PMS systems, and multiple GDS networks, and the picture becomes clear. A new entrant using AI to write the code faster is solving the wrong problem. The bottleneck isn't the engineering time — it's the **bilateral relationship and certification process** that each counterparty must agree to undergo. Booking.com and Expedia have no commercial incentive to undergo that process for an untested new entrant with zero hotel clients.
This is why SiteMinder's scale-based cost advantage over smaller players is a distinct source of competitive edge beyond switching costs alone [Morningstar Australia](https://www.morningstar.com.au/stocks/ask-analyst-can-this-genius-product-create-next-asx-software-king) — the fixed cost of maintaining hundreds of certified, live integrations is enormous, but it doesn't scale with the number of hotels served. A competitor starting from zero faces the full fixed cost but with none of the revenue base to fund it.
### Moat Layer 2: SiteMinder as the Path of Least Resistance for OTAs
From the OTA's perspective, SiteMinder is not a vendor they tolerate — it is a distribution accelerator. Rather than negotiate and certify integrations with tens of thousands of individual hotels and their disparate PMS systems, an OTA can integrate once with SiteMinder and instantly access millions of rooms. SiteMinder's 2.4 million hotel client rooms represent a third more rooms than those operated by the next competitor [Morningstar Australia](https://www.morningstar.com.au/stocks/ask-analyst-can-this-genius-product-create-next-asx-software-king) , which means the inventory density available through a single SiteMinder connection is simply unmatched.
This creates an asymmetric incentive structure. An OTA that chooses to integrate with a new AI-powered competitor instead of, or even in addition to, SiteMinder would be accessing a fraction of the inventory at the cost of a full certification effort. The rational OTA choice is almost always to deepen the SiteMinder relationship rather than fragment it.
### Moat Layer 3: Switching Costs on the Hotel Side
SiteMinder's moat is partly due to switching costs arising when a product is so embedded in a company's daily operations that moving to another supplier becomes more painful and riskier than staying put. [Morningstar Australia](https://www.morningstar.com.au/stocks/ask-analyst-can-this-genius-product-create-next-asx-software-king)
For a hotel, switching away from SiteMinder means temporarily losing clean, real-time distribution to all connected OTAs during the migration. That is not an IT inconvenience — it is a revenue event. Every night of lost or corrupted OTA availability during a transition is a night of rooms that might not sell, or might double-book. For a business where the vast majority of costs are fixed and incremental revenue flows almost entirely to profit, this risk is genuinely material. A revenue manager will resist any change to the distribution stack that introduces even a small probability of such disruption.
### Against AI-Powered New Entrants Specifically
An AI-powered competitor building a new channel manager would need to solve three problems simultaneously: (a) convince OTAs to go through a new certification process for zero hotel inventory; (b) convince hotels to migrate their most revenue-critical system to an untested platform; and (c) do both fast enough that SiteMinder doesn't simply incorporate the same AI capabilities into its existing product, which it is actively doing.The AI advantage — faster code generation, smarter pricing algorithms, better natural language interfaces — is fully available to SiteMinder too. The moat is not in the absence of AI; it's in the bilateral contractual relationships and network density that no amount of AI-assisted coding can shortcut. Most AI travel agents already pull data from existing travel platforms and OTAs rather than sourcing directly from hotels, meaning properties connected to well-integrated distribution platforms are more likely to be included in AI results simply because that's where the models find reliable data. [SiteMinder](https://www.siteminder.com/r/ai-travel-agent/) This is a structural tailwind for SiteMinder: as AI agents become more prevalent in travel booking, the value of being the single, reliable, structured-data source for hotel inventory increases, not decreases.
### Against Hotels Vibe-Coding Their Own Operating System
The "vibe code your own PMS" threat is superficially plausible but fails on inspection. A hotel owner could theoretically use AI to generate a property management system. What they cannot generate is a certified, live, real-time connection to Booking.com, Expedia, Amadeus, Sabre, and 450 other channels. SiteMinder's pmsXchange API is built on Open Travel Alliance specifications with strict message structure requirements, SOAP/REST protocols, WS-Security authentication standards, and timeout and error handling specifications [Siteminder](https://developer.siteminder.com/siteminder-apis/pms-rms/introduction/pmsxchange/integration-requirements) — all of which must be certified against each counterparty's own requirements.
More fundamentally, a boutique hotel owner who used AI to build their own PMS would still need to then also build, maintain, and certify every one of those channel connections — and then keep them updated as the OTAs evolve their own APIs. The ongoing maintenance cost of keeping 450+ live integrations current is not a one-time build problem; it is a perpetual operational commitment. For any hotel smaller than a major chain with a dedicated engineering team, this is simply not a rational use of resources.
What SiteMinder has built, in essence, is an **integration utility** — the kind of infrastructure that becomes more valuable the more embedded it is, and where the cost of the next entrant replicating it is not the cost of writing better software, but the cost of rebuilding a decade of bilateral commercial relationships from scratch.
Great rebuttal @jcmleng
In your investment thesis, what are you worried about?
if you have nothing that you are worried about, does this indicate a significant bias or blinkered thinking?
I hold SDR, and partly because of your great analysis, but in all my strongest conviction holdings, there’s always something to worry about.
@Chagsy, I was focused on answering the "strength of moat" question vs a proper thesis, per se, so I missed out the risks section. Here are my 5 risks.
While the Likelihood is Low, given the current moat, the severity will be High if any of these things did occur, and I suspect, it will dent my conviction somewhat.
