In May 2025, I rage quit flight search.
Not because flights were expensive, but because pricing felt opaque, dynamic, and unfair.
Same route, same date, same time window… yet different prices depending on where/how often searched. After checking what felt like a unending OTAs, I was frustrated by what I can only describe as digital discrimination in airfare.
The moment CannyFlyer was born.....
One day, while digging for cheaper fares, we noticed something interesting:
Group bookings can be ~15–35% cheaper.
That triggered a question:
Why is “group intent” rewarded, but individual travelers are stuck playing the refresh-lottery?
So we formed a concept:
That day, CannyFlyer got its first heartbeat......
The team
We were three people building from scratch:
May → October 2025: Roadmap → MVP → deployment → first 100 users
In the first 5 months, we did everything end-to-end:
By October 2025, we reached our first 100 users.
But it was painful. The hardest parts were:
The turning point: “Learn & Apply” (Oct → Nov 2025)
From October 2025, we stopped guessing and started learning.
We took help from industry experts people who understand airline distribution, demand, incentives, and constraints. We adopted one simple operating system:
Learn → Apply → Iterate.
That single shift sped up everything.
Nov 2025 → Jan 2026: 1,000+ users + airline partnerships
From November 2025 to January 2026, we scaled to:
What we learned (the real lessons)
Where I’d love your advice
If you’ve built marketplaces / travel / airline pricing / bidding systems:
How would you design the “supply side” onboarding and incentive model so airlines participate consistently (without harming user trust)?
And what would you track as the one metric that proves real PMF here?
Love how you transformed a personal frustration into a marketplace solution! The "anonymous group demand" model is brilliant - it flips the traditional power dynamic where airlines have all the pricing leverage.
Your point about UI/UX being trust is something I relate to deeply. In my crowd marketing service, I've learned that clients don't leave because of design flaws - they leave when they feel uncertain about what's happening behind the scenes.
The Learn → Apply → Iterate shift you described is exactly what accelerated our progress too. We spent months building features no one asked for until we started talking to real potential clients about their actual pain points.
One parallel I see: just like you're aggregating flight demand, I'm helping businesses aggregate their community presence across forums. Both solve the "one voice is too weak" problem by creating collective impact.
Quick question about your supply-side strategy: Have you considered starting with budget airlines who might be more eager to fill seats vs. premium carriers? They could be your wedge to prove the model before scaling to larger partners.
Also curious - how are you handling the "cold start" problem when groups don't hit critical mass? Do you have a minimum threshold before approaching airlines, or are you testing with smaller groups first?
Great execution on getting to 1,000+ users in such a competitive space!
This is a brilliant application of the Reverse Auction model. The travel industry has spent a decade using AI to squeeze more 'willingness to pay' out of individuals through dynamic pricing. Turning the tables by using Collective Demand as a negotiation lever is the ultimate 'David vs. Goliath' play for 2026. The real challenge with 1,000 users is usually the 'Cold Start'—how are you ensuring the airlines/aggregators take a group of 1,000 seriously before you hit 100k? Great work on the 'Anonymous' angle, too—it removes the friction of data privacy concerns immediately.
Appreciate this a lot — you nailed the core reversal: instead of extracting WTP from individuals, we’re turning collective intent into negotiating power.
On the cold start: we didn’t ask airlines to “trust our future scale.” We made the early groups real, verifiable, and low-friction for them to quote on. A few things that helped:
1,000 users won’t win by “being big” — it wins by being high-signal. That’s what we’re optimizing until we hit true scale.
Curious: from your view, what’s the strongest proof-of-seriousness for airlines early on — deposit-backed intent, repeatable volume on specific routes, or conversion history?
Interesting approach to the demand aggregation problem. The "group intent = better pricing" mechanic is clever — airlines get predictable demand signals, users get transparency.
To your questions:
On supply-side onboarding:
The chicken-and-egg here is real. A few thoughts:
On the one PMF metric:
I'd track "% of groups that receive at least one airline offer." That proves both sides are engaging. Conversion to bookings matters, but first you need proof that the marketplace mechanics work.
Curious: what's your current group-to-offer ratio? And how are you handling the cold start — do users see empty groups, or do you seed with real demand somehow?
Right now offer coverage is still early-stage (we’re intentionally limiting supply to a few routes/time windows), but it’s improving consistently as we add partners and standardize response windows.
On cold start, we’ve designed the product so users never see “empty groups.” We use:
This is gold .... thank you.
Single-airline wedge + off-peak/LCC makes a lot of sense, and I like the “fill unsold inventory” positioning. Also agree that private bidding with public competition is the right balance for trust + airline sensitivity.
On PMF metric: great point. % of groups that receive ≥1 offer is probably the cleanest “market is alive” signal before conversion.
Curious , when you say reverse auction: would you recommend one winner-take-all offer, or multiple offers shown to users with clear tradeoffs (price vs baggage vs timing)? We’re leaning toward multiple to preserve trust.
Glad the framing resonated.
On cold start: Your approach is smart — batched windows + demand pooling + verified counts addresses the "empty marketplace" perception problem well. Users feel momentum without you having to fake it.
On reverse auction (winner-take-all vs multiple offers):
I'd lean multiple offers with clear tradeoffs — here's why:
Trust asymmetry: Winner-take-all feels like "the system picked for me." Multiple offers feel like "I'm in control." For users already skeptical of opaque pricing, that control matters.
Price vs value: Different travelers weight price/baggage/timing differently. Letting users see tradeoffs (even if one is clearly cheaper) validates that the system understands their constraints.
Airline incentive: Multiple visible offers create "soft competition" — airlines see they're being compared, but you're not forcing a race-to-bottom on price alone. That's sustainable.
The risk: too many options = decision paralysis. I'd cap it at 2-3 offers max, and maybe highlight one as "best value" to guide without dictating.
Curious how airlines are responding to the visibility — any pushback on showing competitor offers?
Airlines mostly don’t mind multiple offers they mind full transparency of the auction.
What they’re okay with
What triggers pushback
How we balance it
Show 2–3 offers max with clear tradeoffs (baggage/timing/refund/all-in price)
Keep bidding blind: airlines see rank, not competitors’ exact quotes
Prove fairness via verified group counts + window close + price valid till timer
Give a try to my Reddit Extension. It's a Chrome extension called Pulse of Reddit that basically acts like my own alert system for Reddit.
Anytime someone posts something with keywords I care about like 'looking for a designer' or 'best SEO tool' it pings me right away. It’s saved me so much time and helped me hop into threads while they’re still fresh.
If you’re tired of manual digging and want to catch those conversations early, I’d really recommend giving it a look.
It’s free to start and super simple to set up.
Website:
pulseofreddit.com
Great story, Raghav! I really resonate with your point that 'UI/UX is trust.' I’m currently building a hiring platform (Bossr) that tries to fix transparency in recruitment, and I see the same pattern: users drop off the second they feel like the 'system' is playing games with them.
100% ..... the moment users feel “the system is playing games,” they bounce.
Bossr sounds like it’s tackling the same core problem. What’s the one change you’ve made that improved trust the most so far?
Give a try to my Reddit Extension. It's a Chrome extension called Pulse of Reddit that basically acts like my own alert system for Reddit.
Anytime someone posts something with keywords I care about like 'looking for a designer' or 'best SEO tool' it pings me right away. It’s saved me so much time and helped me hop into threads while they’re still fresh.
If you’re tired of manual digging and want to catch those conversations early, I’d really recommend giving it a look.
It’s free to start and super simple to set up.
Website:
pulseofreddit.com
🫡🫡 great
Give a try to my Reddit Extension. It's a Chrome extension called Pulse of Reddit that basically acts like my own alert system for Reddit.
Anytime someone posts something with keywords I care about like 'looking for a designer' or 'best SEO tool' it pings me right away. It’s saved me so much time and helped me hop into threads while they’re still fresh.
If you’re tired of manual digging and want to catch those conversations early, I’d really recommend giving it a look.
It’s free to start and super simple to set up.
Website:
pulseofreddit.com
Nice work👍
Great work!
Give a try to my Reddit Extension. It's a Chrome extension called Pulse of Reddit that basically acts like my own alert system for Reddit.
Anytime someone posts something with keywords I care about like 'looking for a designer' or 'best SEO tool' it pings me right away. It’s saved me so much time and helped me hop into threads while they’re still fresh.
If you’re tired of manual digging and want to catch those conversations early, I’d really recommend giving it a look.
It’s free to start and super simple to set up.
Website:
pulseofreddit.com
Thank you ! Check out the webpage : cannyflyer.com for group discounts on flight tickets .
I like how you treat CannyFlyer as a trust product first, then a pricing product. The point about users dropping when they feel the system is playing games matches exactly how I behave when booking flights.
Totally with you , we’ve learned the hard way that in flights, trust is the product and pricing is just the outcome.
That “rage quit” feeling you described is exactly what triggered CannyFlyer for us: when prices look like they’re changing based on you (device/search frequency/location), people don’t just think “expensive” they think rigged.
That’s why we’re building around transparent, aggregated intent instead of refresh-lottery. You get offers that feel earned by demand, not manipulated by opacity
Also love your point because it’s a good north star for us: if anything feels like games, users will drop even if we’re cheaper.
If you’re up for it: what’s the one moment in a flight flow that makes you personally go “nah, this is shady”? (price jump, fake urgency, baggage surprises, “only 1 left”, etc.) That’s the exact friction we want to eliminate.
I’ve noticed a significant shift in user engagement since implementing the Learn → Apply → Iterate model. It’s really resonated with me!
I spent eight months building a SaaS that launched with zero paying customers. I didn’t check demand early enough, so that shift to expert feedback and iteration is definitely key.
Here are a few ideas:
How do you balance giving airlines enough control without making the user experience feel complicated or shady?
Interesting take on repositioning demand from a weak individual to a stronger group presence. I like it.
Thank you ! Check out the webpage and register : cannyflyer.com
The insight that 'UI/UX is trust' is spot on. In travel (like in data infrastructure), users don't churn because of bad colors; they churn because of ambiguity.
Users don’t churn in travel because of bad colors they churn because of ambiguity.
When I rage-quit flight search in May 2025, it wasn’t just high prices. It was the opacity: same route/date, different fares depending on how/where I searched. It felt unfair and once trust breaks, users bounce.
That’s why “UI/UX gaps” hurt us early. Not aesthetics .... clarity: what’s happening, why this price, what changes it.
CannyFlyer is basically a trust product: make pricing rules visible by converting “refresh-lottery” into transparent group intent → better offers.
Nice idea — love the group demand angle 👏
1,000 users is a strong start. Excited to see where this goes!
Still very early, but it gave us the confidence that transparent, intent-driven pricing resonates. Excited (and a bit scared 😄) to see where this goes next.
Great insight on CannyFlyer. It strikes me as "Groupon for Flights," tackling the critical issue of perishable inventory (spoilage).
I’ve been analyzing your marketplace dynamics, and it seems the core bottleneck isn't user demand, but the "Trust Mechanism" in the supply chain. You are facing a classic "Cold Start" system trap caused by Delay:
• Users join → Wait for offer (Delay) → No instant gratification → Trust erodes → Churn.
To fix this loop and unlock the Airlines (Supply), you might consider these three layers of intervention:
◦ Psychological effect: It converts "passive waiting" into "active negotiation." It buys you the time needed to close the deal on the backend.
I specialize in designing these kinds of Marketplace Feedback Loops and Trust Mechanisms. I have some specific ideas on how to structure that "Data Verification Model" to win over your first major airline partner.
Happy to chat if you want to geek out on the system design
Your breakdown on anti-cannibalization + blind/opaque auctions is basically the playbook for getting airlines comfortable. If they even suspect business travelers are getting discounted seats, they’ll shut the channel down immediately. So proving incrementality (price-elastic, leisure-ish demand) feels like the real unlock.
What we’ve already started building (not integrated into the product yet):
We’ve begun prototyping AI agents to collect + normalize pricing signals and generate an instant “confidence layer” when someone joins a group — even before any airline offer arrives.
The idea is: the moment a user joins, they see something like:
So instead of “passive waiting,” it feels like “active negotiation” — and it buys us time while the supply-side partnerships catch up.
Why this also helps on the airline trust side:
That same modeling layer becomes the foundation for a demand verification / incrementality profile we can show airlines:
And on the auction side, we’re leaning toward opaque bidding exactly like you suggested bids visible to users, hidden from competing airlines to avoid a public price-war dynamic while still creating competition.
What a great perspective this is! ~
I had never thought that “group intent” gets rewarded while solo travelers get punished, but it is something that makes a lot of sense.
Another interesting point is the shift from guessing to talking to industry people and learning → applying → iterating. You can sense the point where the project quickened.
I enjoy how UI/UX in travel is essentially a trust issue, not a design issue. It is similar to how I act when booking airlines.
Totally ...... and that “group intent rewarded / solo punished” framing is exactly the crack we’re trying to exploit with CannyFlyer.
Today, a solo search is basically a weak signal: “I might book, I might not.” Airlines price that uncertainty. A group, even if anonymous, is a much stronger signal: X seats, same route, same time window, real intent and that’s something airlines can actually plan around (and compete for) without guessing.
Also yes on the shift from guessing → talking to industry folks → applying → iterating. That’s been the biggest acceleration for us. Marketplaces punish “perfect theory” and reward “tight loops.”
And UI/UX in travel being a trust problem is basically our north star. People don’t drop because the UI isn’t pretty they drop when they feel:
That’s why we’re building the “instant confidence layer” (predicted price band + probability score) alongside the marketplace mechanics , so joining a group feels like progress, not waiting in a black box.
If we can make the experience feel transparent + predictable, the supply side gets easier too, because trust compounds on both sides.
Interesting how transparency itself became the value prop. Did you notice any conversion lift once users saw group demand vs just individual pricing?
Yeah.... transparency became the product faster than we expected.
On the specific question: we haven’t measured a clean conversion lift yet from “showing group demand vs individual pricing,” because we’re still early and the full flow isn’t consistently live across enough users to make the data meaningful (especially since supply/offer latency is still the biggest confounder).
That's amazing
This is a really interesting angle. You didn’t just try to “find cheaper flights” — you changed the demand model. That’s smart.
From what I’ve seen in marketplaces, the biggest thing on the supply side is predictability + fairness. Airlines will participate consistently if:
For PMF, I’d track one simple metric:
% of group intents that convert into actual bookings.
If users repeatedly join groups and airlines consistently fulfill them, that’s real signal.
Love the Learn → Apply → Iterate shift. That’s usually where real momentum starts.
Appreciate this , you nailed the core risk: we don’t want to teach users to “wait for discounts.”
We’re aiming to position it as aggregated intent → competitive offer, not “sale pricing.”
And yes, PMF = intent-to-booking conversion. If that number climbs and users repeat the behavior, it’s hard to fake.
Curious: have you seen better supply consistency with time-bound batches (e.g., groups close in 24–48h) or with always-on rolling demand? What worked best in your experience?
That’s a great question.
In my experience (not travel specifically, but marketplaces in general), time-bound batches usually create stronger action early on. Deadlines create urgency on both sides — supply knows when demand locks, and buyers know when to decide.
Always-on demand feels smoother, but it can dilute momentum if there’s no clear “event.”
If I were testing this, I’d start with 24–48h closing windows to force liquidity and learn faster — then layer in rolling demand once volume is stable.
Curious how airline partners are reacting so far to the batching idea?
So far, airlines are more comfortable with batching than always-on because it gives them a clean “decision moment”:
expiry/cancellation rules
We’ll likely stick to 24–48h windows while we build liquidity.
Give a try to my Reddit Extension. It's a Chrome extension called Pulse of Reddit that basically acts like my own alert system for Reddit.
Anytime someone posts something with keywords I care about like 'looking for a designer' or 'best SEO tool' it pings me right away. It’s saved me so much time and helped me hop into threads while they’re still fresh.
If you’re tired of manual digging and want to catch those conversations early, I’d really recommend giving it a look.
It’s free to start and super simple to set up.
Website:
pulseofreddit.com
Where I’d love your advice…..
If you’ve built marketplaces / travel / airline pricing / bidding systems:
How would you design the “supply side” onboarding and incentive model so airlines participate consistently (without harming user trust)?
And what would you track as the one metric that proves real PMF here?