Digital Compendium
Hotels print their guest guides. Guests never read them.
I spent 18 months designing the digital replacement — now the hub 175,000 guests open every month, and a $1M+ product line for Canary.
- Year
- 2024–2025
- Role
- Product designer
- 100% design ownership
- Scope
- CMS builder
- Guest experience
- Compendium Lite
The Problem
Hotels have relied on printed compendiums — booklets in guest rooms listing amenities, restaurant menus, Wi-Fi info, and local attractions — for decades. These are expensive to produce, impossible to update quickly, deteriorate with use, and require sanitization. Competitors like Duve, Hudini, and LoungeUp had already gone digital, and several offered features Canary didn't match: in-room dining, activity booking, and analytics.
Canary needed a digital compendium that did more than replicate a PDF. The product had to become the guest's home screen during their stay — a hub connecting reservation info, messaging, upsells, tipping, and eventually food ordering. Without it, Canary was selling point solutions while competitors sold a guest experience. The initiative target: increase average revenue per room per month from $5.10 to $10.
The Solution
I designed the full Compendium platform end-to-end — a hotel-facing CMS builder for managing content, and a guest-facing mobile web experience for browsing it — then kept extending it for 18 months, from MVP through enterprise-wide release.
The guest experience — sections, dining, and property info configured by the hotel in the builder.
Key Design Decisions
1. Builder-first architecture
I designed the CMS as a structured builder rather than a free-form editor. Hotels configure sections (restaurants, amenities, custom content) through a guided interface with real-time preview. This was deliberate — hotel staff aren't web designers, and a “builder” metaphor with clear sections and drag-to-reorder maps directly to how they think about their property's offerings. The patterns would be reused across registration cards, auth forms, and other editors.
2. Custom sections as the extensibility layer
Rather than hard-coding every content type, I designed custom sections — hotels create their own categories with items, images, descriptions, and call-to-action buttons. One pattern now covers spa services, local attractions, transportation, house policies, whatever a property needs, with no product work per use case. It became the most-adopted feature in the product: 82% of active properties use it.
3. Compendium as the guest journey endpoint
Every guest touchpoint leads here. Check-in completes? You land on Compendium. Arrival SMS? Links to Compendium. QR code in the lobby? Compendium. That routing is the product strategy: with Compendium as the hub, every other Canary product — messaging, upsells, tipping, food ordering — gets organic distribution to guests who are already looking at their phone.
4. Compendium Lite as a freemium growth lever
When we needed to drive cross-sell and deepen the Wyndham relationship, I designed Compendium Lite — a free tier capped at 2 sections and 4 items each. The hard part: where do the limits live so the free product still feels useful? I put the “Request Upgrade” prompts at the exact moments a hotel outgrows the cap — trying to add a third section, a fifth item — because that's when they're already wanting more. Getting that balance right took 22 iterations.
Research & Discovery
Research methods
- Full design QA audit of the guest experience in preparation for Best Western conference — documented all bugs and filed 6+ engineering tickets
- Competitive analysis of digital compendium solutions (Duve, Hudini, LoungeUp, STAY, HiJiffy) — mapped feature gaps across amenity display, concierge, in-room dining, and analytics
- Design jams with engineering on data model decisions
- Live feedback from COMO property tours, Omni pilot hotels, Best Western rollout, Eurostars (270 properties), and Wyndham
- COMO brand requirements workshops — outlined UI requirements including custom fonts, gradient treatments, and brand-specific styling
Key insights
Hotels think in categories, not pages. Ask a front-desk manager about their property and you get restaurants, amenities, policies, local attractions — never “pages.” So the builder is organized around typed sections instead of a generic page editor. Staff recognized the model immediately.
Content quality varies wildly between properties. Some hotels have professional photography; plenty have phone photos taken in bad hallway lighting. The image handling had to flatter both — bulk upload without a forced crop step, galleries, layouts that survive mediocre photos. The bulk upload flow alone took 19 rounds to get right.
Chains and boutiques pull in opposite directions. Best Western wants brand consistency across hundreds of properties; a boutique wants its property to feel like nowhere else. Brand-level defaults that individual properties can override served both, along with branded QR templates for the chains.
The polish itself was selling. Before we had feature parity with competitors, prospects were reacting to how the guest experience looked and felt in demos. Craft was doing sales work — which validated putting the design bar first.
“It's sleek — this is exactly what I'm looking for.”
— Prospect feedback, recurring across 10 sales demos
Research methods
- Full design QA audit of the guest experience in preparation for Best Western conference — documented all bugs and filed 6+ engineering tickets
- Competitive analysis of digital compendium solutions (Duve, Hudini, LoungeUp, STAY, HiJiffy) — mapped feature gaps across amenity display, concierge, in-room dining, and analytics
- Design jams with engineering on data model decisions
- Live feedback from COMO property tours, Omni pilot hotels, Best Western rollout, Eurostars (270 properties), and Wyndham
- COMO brand requirements workshops — outlined UI requirements including custom fonts, gradient treatments, and brand-specific styling
Key insights
Hotels think in categories, not pages. Ask a front-desk manager about their property and you get restaurants, amenities, policies, local attractions — never “pages.” So the builder is organized around typed sections instead of a generic page editor. Staff recognized the model immediately.
Content quality varies wildly between properties. Some hotels have professional photography; plenty have phone photos taken in bad hallway lighting. The image handling had to flatter both — bulk upload without a forced crop step, galleries, layouts that survive mediocre photos. The bulk upload flow alone took 19 rounds to get right.
Chains and boutiques pull in opposite directions. Best Western wants brand consistency across hundreds of properties; a boutique wants its property to feel like nowhere else. Brand-level defaults that individual properties can override served both, along with branded QR templates for the chains.
The polish itself was selling. Before we had feature parity with competitors, prospects were reacting to how the guest experience looked and felt in demos. Craft was doing sales work — which validated putting the design bar first.
“It's sleek — this is exactly what I'm looking for.”
— Prospect feedback, recurring across 10 sales demos
Design Process
The prototype sold the product
Before any of it was built, I made a Figma prototype of the whole idea — a CMS builder with a sticky mobile preview that scrolled alongside it, clickable through Wi-Fi, service requests, restaurants, and room upgrades. Sales started showing it to prospects. Deals closed on it. Months later an engineer asked, half-joking, whether we'd sold Compendium entirely over Figma prototypes — and the honest answer was: at first, yes. That set the bar for the real thing.
Building it for real
The builder MVP was the whole CMS designed from scratch — section management, restaurant and amenity editors, photo management, hours, QR codes, and a translation picker, with the data model worked out alongside engineering (the design doc we wrote together broke down into 27 implementation tickets). The guest side had the opposite brief: feel like a consumer app, not hotel software. Then, weeks before Best Western would put the product in front of hundreds of properties, I ran a full QA audit of the guest experience myself — documented every rough edge and filed the bugs. Enterprise scale means nobody's there to explain away a misaligned label.
Extending the platform
The second year was proving the architecture. Custom sections shipped and became the most-used feature in the product. Carousels and groupings layered onto the same pattern without new plumbing. Compendium Lite turned the platform into a growth lever. And bulk image upload — unglamorous, 19 iterations — removed the single biggest friction point in hotel onboarding by killing the mandatory crop step. None of these required rearchitecting the shell, which was the point of the shell.
Constraints I designed around
- No native app — Everything as mobile web via QR code, SMS link, or Wi-Fi portal redirect
- Multi-language from day one — Translation management built into every content field
- Varied technical sophistication — Hotel staff range from tech-savvy to barely comfortable with email
- Platform foundation — Every pattern needed to be reusable across Canary's product suite
The prototype sold the product
Before any of it was built, I made a Figma prototype of the whole idea — a CMS builder with a sticky mobile preview that scrolled alongside it, clickable through Wi-Fi, service requests, restaurants, and room upgrades. Sales started showing it to prospects. Deals closed on it. Months later an engineer asked, half-joking, whether we'd sold Compendium entirely over Figma prototypes — and the honest answer was: at first, yes. That set the bar for the real thing.
Building it for real
The builder MVP was the whole CMS designed from scratch — section management, restaurant and amenity editors, photo management, hours, QR codes, and a translation picker, with the data model worked out alongside engineering (the design doc we wrote together broke down into 27 implementation tickets). The guest side had the opposite brief: feel like a consumer app, not hotel software. Then, weeks before Best Western would put the product in front of hundreds of properties, I ran a full QA audit of the guest experience myself — documented every rough edge and filed the bugs. Enterprise scale means nobody's there to explain away a misaligned label.
Extending the platform
The second year was proving the architecture. Custom sections shipped and became the most-used feature in the product. Carousels and groupings layered onto the same pattern without new plumbing. Compendium Lite turned the platform into a growth lever. And bulk image upload — unglamorous, 19 iterations — removed the single biggest friction point in hotel onboarding by killing the mandatory crop step. None of these required rearchitecting the shell, which was the point of the shell.
Constraints I designed around
- No native app — Everything as mobile web via QR code, SMS link, or Wi-Fi portal redirect
- Multi-language from day one — Translation management built into every content field
- Varied technical sophistication — Hotel staff range from tech-savvy to barely comfortable with email
- Platform foundation — Every pattern needed to be reusable across Canary's product suite
Impact & Results
A new product line
Compendium passed $1M in cumulative CARR by December 2025 — a brand-new SKU designed from zero, with zero gross churn in the tracked period. And the platform bet paid out: F&B Mobile Ordering launched as the first interactive module inside the same guest experience, no shell redesign required.
“Huge congrats to the entire team for bringing a brand new SKU to market from scratch... This is now the second SKU launched by just this pod in 2024. Most companies are lucky to launch 2 SKUs across their entire company in a year.”
— Aman, CEO (launch announcement)
Guests and hotels actually use it
175K guests open Compendium every month, averaging 2.5 minutes per session. On the hotel side, 82% of active properties have built custom sections and 72% use custom CTAs — averaging 8.7 items of their own content per property. Hotels aren't just enabling the product; they're building on it.
“It's like having your own digital concierge.”
— Live hotel customer
Enterprise traction
- Best Western: The largest rollout, driving the majority of Q4 activations; one property put Compendium QR codes on every guest room TV
- Wyndham: Compendium Lite delivered through their Wi-Fi captive portal, scaling to 400+ sites
- COMO Hotels: Multi-property deployment across Asia-Pacific
- Omni Hotels: Pilot expanded to a full deal, now exploring geofenced QR codes and loyalty integration
A new product line
Compendium passed $1M in cumulative CARR by December 2025 — a brand-new SKU designed from zero, with zero gross churn in the tracked period. And the platform bet paid out: F&B Mobile Ordering launched as the first interactive module inside the same guest experience, no shell redesign required.
“Huge congrats to the entire team for bringing a brand new SKU to market from scratch... This is now the second SKU launched by just this pod in 2024. Most companies are lucky to launch 2 SKUs across their entire company in a year.”
— Aman, CEO (launch announcement)
Guests and hotels actually use it
175K guests open Compendium every month, averaging 2.5 minutes per session. On the hotel side, 82% of active properties have built custom sections and 72% use custom CTAs — averaging 8.7 items of their own content per property. Hotels aren't just enabling the product; they're building on it.
“It's like having your own digital concierge.”
— Live hotel customer
Enterprise traction
- Best Western: The largest rollout, driving the majority of Q4 activations; one property put Compendium QR codes on every guest room TV
- Wyndham: Compendium Lite delivered through their Wi-Fi captive portal, scaling to 400+ sites
- COMO Hotels: Multi-property deployment across Asia-Pacific
- Omni Hotels: Pilot expanded to a full deal, now exploring geofenced QR codes and loyalty integration
Reflection
Worked well
- Designing the builder as a structured, section-based CMS rather than a free-form page editor was the right call. It maps to how hotel staff think and ensures content quality without requiring design skills.
- The platform architecture paid off. Designing Compendium as an extensible hub meant that when F&B Ordering launched 12 months later, it slotted naturally into the guest experience without redesigning the shell.
- Full design QA before the Best Western conference caught issues that would have been visible to hundreds of properties. Taking ownership of quality was critical for a product launching at enterprise scale.
- Close collaboration with engineering on the data model meant the frontend architecture matched the design architecture. Decisions made in that initial design doc carried through 18 months of development.
The customer we lost
COMO deployed Compendium across their Asia-Pacific properties. I ran their brand workshops, designed around their custom fonts and gradients, and property staff genuinely liked the product. It didn't matter: their decision-maker saw a native-app gap against competitors, called us “behind the curve,” and in early 2026 COMO started evaluating a competitor. Two lessons I'm keeping. Staff sentiment isn't buyer conviction — the people who liked it weren't the person deciding. And when the retention conversation came, we had no usage data to argue with, because analytics kept losing the prioritization fight. Polish doesn't beat a roadmap gap.
Would change
- Analytics from day one — see above. A property once asked if they could check how their compendium was being used, and we didn't have a good answer for them either.
- Owning discovery instead of observing it. I joined 20+ customer calls but they were PM-led; I should have been running my own conversations with the hotel staff who actually maintain this content.
- Pushing further ahead of the roadmap. For a platform product, a 6–12 month vision mockup aligns a team faster than any spec — I made one eventually (the North Star brainstorm), but a year later than I should have.
What I learned
- Platform design requires thinking in layers. The guest experience, CMS builder, data model, and API are all part of the design — not just the screens.
- Enterprise products need polish at a different standard. When Best Western is rolling out to hundreds of properties, every pixel matters.
- The freemium model (Compendium Lite) is a design problem, not just a business decision. Where you place constraints, how you communicate limits, and what triggers the upgrade conversation are all UX decisions.
Worked well
- Designing the builder as a structured, section-based CMS rather than a free-form page editor was the right call. It maps to how hotel staff think and ensures content quality without requiring design skills.
- The platform architecture paid off. Designing Compendium as an extensible hub meant that when F&B Ordering launched 12 months later, it slotted naturally into the guest experience without redesigning the shell.
- Full design QA before the Best Western conference caught issues that would have been visible to hundreds of properties. Taking ownership of quality was critical for a product launching at enterprise scale.
- Close collaboration with engineering on the data model meant the frontend architecture matched the design architecture. Decisions made in that initial design doc carried through 18 months of development.
The customer we lost
COMO deployed Compendium across their Asia-Pacific properties. I ran their brand workshops, designed around their custom fonts and gradients, and property staff genuinely liked the product. It didn't matter: their decision-maker saw a native-app gap against competitors, called us “behind the curve,” and in early 2026 COMO started evaluating a competitor. Two lessons I'm keeping. Staff sentiment isn't buyer conviction — the people who liked it weren't the person deciding. And when the retention conversation came, we had no usage data to argue with, because analytics kept losing the prioritization fight. Polish doesn't beat a roadmap gap.
Would change
- Analytics from day one — see above. A property once asked if they could check how their compendium was being used, and we didn't have a good answer for them either.
- Owning discovery instead of observing it. I joined 20+ customer calls but they were PM-led; I should have been running my own conversations with the hotel staff who actually maintain this content.
- Pushing further ahead of the roadmap. For a platform product, a 6–12 month vision mockup aligns a team faster than any spec — I made one eventually (the North Star brainstorm), but a year later than I should have.
What I learned
- Platform design requires thinking in layers. The guest experience, CMS builder, data model, and API are all part of the design — not just the screens.
- Enterprise products need polish at a different standard. When Best Western is rolling out to hundreds of properties, every pixel matters.
- The freemium model (Compendium Lite) is a design problem, not just a business decision. Where you place constraints, how you communicate limits, and what triggers the upgrade conversation are all UX decisions.