Chris: If I told you that most airlines today are making critical pricing and revenue management decisions based on less than 1% of the available market signal, would you believe me?  

You should.  

I'm going to tell you a little bit about why over the next few minutes.  

So, my name is Chris Phillips. I'm the Chief Commercial Officer here at ATPCO and this week, we're here talking about data in motion, offers in action. And right now, I want to focus on that missing 99% of signal that most airlines are missing out on.  

So, here's the reality.  

For 50 years, airline revenue management has been built on three foundational inputs: historical sales data, booking curves, and static snapshots of competitive pricing.  

The model is rigorous, it's widely proven, and it's driven enormous value for the industry for many years. But the market that that model was built for no longer exists.  

There are roughly 17 million commercial airline seats flown every day, and to fill those seats, there's roughly 1.5 billion genuine shopping demand expressions. If you add in meta search and repeat price checks, that number balloons to over 5 billion daily shops before agentic AI even comes into play.  

If your revenue strategy is primarily based on booking behavior, you're working on less than 1% of that available signal.  

And here's what that costs you.  

Could you have sold more high yield products? Held inventory longer? Priced a branded fare more precisely if you had understood what demand actually look like at the moment the tickets were purchased?  

You may never know because that signal is never captured.  

I know airlines have developed unconstrained methods to estimate what's been missed, and that's an important step, but estimating from historical booking patterns is still a rearview mirror strategy.  

Shopping data gives you the actual market intent right now, and this is not an incremental improvement to how airlines can forecast demand. It's a fundamentally different input.  

So, here's the hypothesis I want you to sit with: We need to redefine what demand means in this industry. Booking curves, load factors, yield patterns, those are all lagging indicators. They tell you what happened, but they cannot tell you what is happening.  

The new model is the full picture of what the market expressed, what was searched, what was offered, what was compared, and what was ultimately purchased.  

When your optimization models and your dynamic offer logic are fed that signal instead of historical estimates, everything gets more accurate, not marginally, but fundamentally.  

And for the first time, that signal is becoming visible.  

But here's what's easy to miss in these emerging signals. Shoppers are no longer just comparing prices, they're comparing offers. They're comparing brands. They're comparing the seat they're going to sit in. They're comparing the flexibility of the travel experience.  

Two offers at the exact same fare can mean completely different things depending on whether you're buying a bare bones basic economy product or a fully flexible fare with bags. Same price, entirely different competitive situations.  

And this is where the ATPCO ecosystem becomes the foundation.  

For years we've been building an infrastructure to decode what an offer actually is... beneath the price, down to the atomic construction of the fare.  

Product Catalog, it defines what is being sold.  

Branded Fares tell you how it's packaged.  

The Routehappy product brings the offer to life through visualization.

And then the work underway on brand attributes and product performance, you can increasingly understand not just what an offer contains, but whether it's winning or losing in the marketplace.  

And this is not just a data feed.  

This is a competitive intelligence infrastructure. It's built and trusted by this industry for decades. But one critical piece is missing. Knowing what you and your competitors have built as offers is foundational, but knowing how the market is actually responding to those offers in near real-time across every channel... that's a different capability entirely.  

And that's exactly why ATPCO acquired 3Victors.  

Not to add another data feed, not to build another dashboard, but to close this loop.  

3Victors surfaces what's happening in the market at the speed the market demands. ATPCO decodes what it means, and together, for the first time, you have a complete real-time picture of the competitive reality.  

So, let me make this a little bit more concrete.  

Something shifts in the market. A competitor moves. Your analysts know something's changed, but finding it, understanding it, and acting on it can take hours and sometimes even days.  

Meanwhile, the market's moved on with the 3Cictors price size solution, competitive shifts surface automatically. Not just price changes, offer changes, brand changes. The full competitive picture is decoded in the moment that it happens. You're no longer searching for the problem. You're pointed directly to what's changed and what it means in minutes, not in days.  

Every hour of lag is an hour of missed opportunity. PriceEye eliminates that lag.  

But PriceEye is only half of what 3Victors brings to the table.  

Competitive intelligence tells you what rivals are doing, but it does not tell you what the market is about to do. That upstream signal is where the real advantage lives, and that is DemandView.  

Imagine your models are telling you to close a premium availability. Standard practice... your booking curves look normal, everything's fine. But at the exact moment the shopping data shows a premium cabin search spike more than 40% over the last six hours, travelers are not just browsing, they're comparing, narrowing their search, and getting ready to buy.  

This new data is not a model inference. This is what the market is telling you about the booking curve that you cannot see yet. You need to hold that inventory, adjust your bid price, and act before the moment passes.  

Without DemandView, you're likely going to close that availability and never know what you left on the table.  

With DemandView, you have the evidence to make a better decision.  

Let me show you what that looks like in the real world.  

So, hopefully like many of you, I'm getting very excited for the World Cup, opening up in just a just a few weeks. But even more exciting than the matches that are coming up, we've been studying the data, and one thing that we found after the qualifying matches on the 31st of March, we saw some striking behavior differences.  

We selected three markets here: Sweden, Turkey, the Czech Republic.  

Prior to the 31st of March, demand's steady, nothing unusual happens. Then, the teams qualified for the World Cup. Immediately, demand spikes. Not gradually, not over days, but the moment it happens.  

Fans are planning, they're searching, they're comparing. The shopping signal ignited before a single booking is made. Different markets, same instant response. But none of this is visible if you're only looking at the booking curve.  

The airlines that saw the signal on the evening of March 31st made very different inventory, pricing, and even capacity decisions than the airlines who saw it in the booking reports several weeks later.  

The difference isn't theoretical. It's revenue, captured by one airline and missed by another.  

On the other side of the coin, your team loses, you're going home from the World Cup. The dream's over, but demand doesn't necessarily disappear. It changes.  

Without real-time shopping data, you're guessing.  

With it, you know within hours how many fans are still coming, where they're going, how many are reconsidering their World Cup experience. Difference between reacting and being left holding unsold inventory is in the data.  

The earliest signal for demand is never the booking, it's always the shopping. And the gap between when that signal appears and when your booking curve reflects it is where the yield is either won or lost.  

Eventually the booking data does catch up, but in a dynamic offer environment, eventually will always be too late.  

So, let me close with this: This isn't just a data story. This is a decision quality story. Every pricing decision, every inventory call, every dynamic offer you make is based on the quality of the decision. But the quality the decision is only as good as the signal behind it.  

Better signal means better decisions and this compounds across every market you compete in every day.  

The question is not whether real time shopping intelligence will become the standard in the industry. We believe it already has.  

The only question is whether you're building on that advantage now or whether you are the competitor someone else is out-maneuvering.  

The choice is being made now, whether you were making it consciously or not.  

So data in motion, offer in action.  

And now, rather than hearing any more from me, you're going to hear from someone who's already living this experience.  

So, please join me in welcoming Nick Mirza, Senior Manager of Revenue Management, Data Science at JetBlue.  

Thanks for joining us. Thanks for being our guinea pig on this data journey.  

So, I'd like to start just with some stage setting with you. When you think back to the moment your team realized that the traditional competitive data wasn't enough anymore, was that an inflection point? Was it a specific market event? A miss? Or some just slow-building frustration with the data that you had?  

Nick: Yeah, really happy to be here. I think it was more the latter.  

I think that it's immediately obvious and intuitive that these data are extremely powerful. But what isn't helpful is if we don't trust the data.  

And I think it was a build up of little things here and there where we were really looking for a platform and a provider that could get us the data with reliability and the right context and features. We want rich data, not just more of it that would help us say, "Okay, if I'm seeing this in the market, I can really believe that that's what what's out there and I don't have to check someone's homework."  

I don't want to spend my day price shopping on Google Flights or whatever just to see if the quality of the data in my BI reporting or whatever it is good, right? And as the quality of the data gets better, that's when you can start to lean into it more with things like automation, decision augmentation, and all the fun stuff.  

But it starts with data quality.  

Chris: Trust and confidence in data is  a factor for everybody in this room. I think it's something that Brett and Dan talked about in in their opening chat this morning. A critical part of the decision process eliminates a lot of friction.  

Nick: Yeah.  

Chris: We've talked a lot, and we'll be talking more later today, about real-time data or near real-time data. In your experience, what's the difference between data that's fast and data that's actionable? And where do you think most organizations are failing on that issue?  

Nick: Yeah. I mean I can't speak to other peer organizations, but what I can say is this: We're not moving away from this idea of, of higher velocity data, but with velocity, comes lots of trade-offs, right?  

And not to beat a dead horse...  

The most important thing is, are the data accurate, right? Because if there's inaccuracies in the data, or if there is gaps, incomplete data, then any real-time or near-real time decision making starts to get really dicey.  

Are we chasing noise? Are we chasing measurement error in the data, right? Or is this a real signal?  

So, it starts with a really good solid data feed, call it. And then as you add in velocity, it brings up, I would say, more challenges, but the fun kind, the good kind, if we trust the data.  

Chris: So, speed still wins, but speed with accuracy is critical.  

Nick: Exactly.  

Chris: Okay, kind of building on that, how do you balance the instinct to act fast on a market signal versus the discipline to make sure you're reading it correctly? You're a data scientist, so you lean on the data and the facts, but where does human judgment come in as part of that decision making process in this environment?  

Nick: Yeah, it's a really good question. And you touched on it earlier in your opening remarks when you were talking about things like load factors and bid prices, and the more that you start to ingest high velocity data and start to build things and scale them, it's really critical.  

And this is sort of an obvious thing, but you have to be marrying these data sources together. There's no single oracle of "this is exactly what we should do," right? It's never going to. We all know that the industry and revenue management practice is too complex for that to work.  

So, when we think of sort of building systems on top of it, it really comes down to marrying data sources to triangulate in context this high velocity signal, what does it really mean?  

Chirs: Let me build on that a little bit. So, shopping data sits at this interesting intersection of pricing, revenue management, data science. You've talked about the trust factor, the confidence factor, but how do you navigate when you're marrying all this data? How do you navigate the internal question of who owns it? What does that mean for how the decisions actually get made so it doesn't become siloed?  

Nick: It's another good question and I might have a little bit of small carrier bias here because it's not something that we've really struggled with. Very lucky to work in an organization, even if I'm technically not in the revenue management group, that's who I sit next to in the office, right?  

I would say that data ownership comes down to how you want to set up some of the engineering work, right, for like who owns pipes. But I think the more interesting part is decision ownership. And to me, that's something that lives more apparently, not necessarily with a data science group, but really who are we supporting, right. If our job is to support the pricing and revenue management teams, they're the decision makers and we can do our best to try to help them make better decisions and more accurate, faster decisions, all that. But at the end of the day, they're the ones who are accountable to what we're putting out in the market.  

So, that's sort of, that's our framework that we work with. And if there's things that they're not getting from the data, we have really open lines of communication. We can help fix that.  

Chris: I'd like to build on the decision making and the trust topics that you brought up. As you're bringing this new data together, you're marrying it to you'll make the right decisions. What's the change management process like, within JetBlue to actually use the data at scale to make better decisions?  

Nick: Yeah, so change management's fun. It's something I have a lot of experience getting wrong, frankly, and I can kind of speak on what not to do.  

The worst thing you can do is just throw together a dashboard or some sort of automation and not give your stakeholders the right context and the right framing to explain why it's important, what it can answer, and then what it can't answer. And, so, I think that one of the most effective ways to combat that is being really intentional when you're designing products to have a specific use case in mind and then scenario building, right?  

And I think that's the fun part. And again, it helps to work with a relatively small team where we can tap shoulders and say, "Hey, here's some exercises, Here's some scenarios that you might want to consider. If you're seeing this in your market, if you're seeing this in the shopping data, and this in your load factors," and we connect those dots, here's how you might want to action something.  

And one of the most rewarding things is when someone has that light bulb moment of, "Okay. I saw this in my load factors or whatever you're calling traditional RM signals," and then I went and I looked at the shopping data." And then they're like, "It all made sense then that I was missing this bit of context. I was missing this signal. And here's how it helped me come up with a new decision, and I believe it was the right one for this reason." And that's a really fun and rewarding thing because, again, the data are just so intuitive. If you're thinking like a customer and you're able to ingest that data at scale to help get in front of people... like what are normal people buying plane tickets? What are they seeing? It's the missing link, right?  

Chris: So, the delivering those competitive insights to your team is an important role for what you do for them. But we talked to a lot of airlines where competitive data becomes a crutch for them.  

Everyone's watching each other. Sometimes some are leading, some are reacting.  

How do you make sure that those teams that you're delivering this data to are using the market intelligence to differentiate and not just following?  

Nick: I would say that's one of the fun parts. On the more abstract, sort of call it systems design side, having that mentality of  what are we really trying to get out of these data? And how does that fit into a system, sort of as a first principle of... we don't want to open the door to be just sort of whipsawing and reacting to everything if we don't think it's worth reacting to, right?  

And again, that comes back to marrying data sources together, right?  

If your shopping data live in a silo, and then you're booking data live in a silo, and your systems data live completely siloed, you're not going to have... you're going to struggle because there's going to be friction there.  

So, when we work with the team, it's really driving home that element of we're sort of examining this cube that is the real demand and shopping world, but we have to look at it from different angles. And the more we can make it easy to do this and not have to reach here, here, and here, the more empowered the team is going to be to make the right decisions.  

Chris: Okay, that helps them make the decision to really kind of lead their strategy versus being a market follower.  

Leveraging that data, what questions do you think you and your team are more confident in answering today, but you couldn't answer two or three years ago before you really started unlocking the power of the shopping data?  

Nick: It's a very simple one that we can answer now and it's "Really?" Right! It really comes down to, "Okay, I I trust these data, I can start to use them in a more, call it high-leverage way when I believe that the data are showing me something that's intuitive." And it's not just really as in like, is that really the fare? It's "Is that the fare brand that I would expect to see that fare associated with?"  

And the more I can say, "Oh, aha, that that actually makes a lot of sense or that actually doesn't make sense so I'm going to disregard it because someone's doing some strategy that I don't know, fine, I'm going to choose to ignore it." But having a better understanding and more trust in the data that says, "Okay, I don't think this is measurement error." It makes it much easier to sort of call a shot and use those data to help inform my own strategy.  

Chris: So, in a world of increasing dynamic offers, in a world of agentic AI, we threw out some provocative ideas that the old world of static data is no longer fit for purpose on its own. Blending in some of this more real-time shopping data is the future.  

I'd like to get your opinion on how it's really changed your approach in understanding what was shopped, not just shopped, but understanding what was the demand? What does your team think about that progression of really understanding the market signals as they're happening versus reacting to what's already occurred?  

Nick: Yeah, it's a really good question and I'm lucky to have a really fantastic data science team of really, really sharp scientists who work with me. And that's one of the things that they spend a lot of time thinking about, right?  

And I don't want to divulge any secret sauce here, but I would say that there are ways to use the data where it goes back to that "aha" moment of, if I'm looking at my historical bookings and I can marry up with the price was at the time, then you can start to tease out cause and effect in a better way, right?  

And, ultimately, in revenue management, there's what do I think is going to happen? And there's what do I want to happen? And getting those two things to agree with each other is where these data become really powerful, even when we're starting with a historical-looking view of, well, what did I want to happen at the time? What really happened? Okay, what other signals were out there that maybe I hadn't considered? And here's how I can sort of reduce that tension and use what I've learned going forward.  

Chris: That's where you unlock the power of the data.  

You talked earlier about ingesting this data is a journey. It's a learning curve. We're all bound to make mistakes.  

If you could give one piece of advice to your peers in this room who are just starting this journey of really kind of taking and ingesting real-time shopping data, what's the thing that they need to hear that nobody tells them?  

Nick: I'm not sure that no one tells them, but don't silo the data. Don't keep it under lock and key so that you can't learn from it and encourage your people to look at the data and come up with creative use cases.  

At the end of the day, you're going to be better off organizationally if you have teams of people who are curious and thinking critically about the data and coming up with their own insights than if it's this super sort of "this lives in data science and we control it" You're not going to do yourselves any favors doing that.  

Chris: I think we'd all have to be honest, the airline industry is very powerful on data silos. So, the opportunity to break those down is going to put you in a much better position.  

Nick: Yeah, and with all of the investments that the world is making in technology, right, it's not just an airline thing... It's never been easier to marry these data sources together with modern cloud computing and all that stuff. To me, there's really no excuses at this point.  

Chris: So, it's getting easier, it's getting faster, there's more and more data inputs that we're taking advantage of. We talked about speed and confidence as well. I'd love your thoughts on whether you think there's a risk that an airline becomes too reactive, that you're constantly chasing a signal and you're attempting to over-optimize over the long term.  

Nick: Yeah, absolutely. And this is one of those things that's sort of a hot topic of discussion for the folks in this room who I know from, the AGIFORS world, right? How do you use these data in a way that's not going to make you dizzy and give you decision regret, right?  

And I think that's where, if you understand that risk and that potential trade off, sort of at the ground level and you keep that in mind as you build out systems, processes, practices, whatever it is, you should have ways to naturally safeguard against that.  

And I would say being reactive is not as fun as being proactive. And I think all of us here are in the airline industry in no small part because it's really fun. I think that's what keeps people coming back to this type of work. And I think one of the easiest ways to keep it fun is maintain some ownership and control of your strategy, your differentiation, and how you're revenue managing and creating offers, right?  

So, don't remove that thing. Don't take that away from yourself. Use the data to inform your strategy, but don't make a singular data source be your only strategy. And that's true for any signal in the revenue management space.  

Chris: That's good advice. Alright, last question. We've talked earlier today a little about AI. We can't predict what AI is going to do with us tomorrow, but if you had a crystal ball, what would winning look like for you or for any airline that gets this data story right five years from now?  

Nick: Feeling like you're in the driver's seat. There's a lot of things that we cannot control, but getting a better sense of "here are the things that I can control," and one of those things is if I have a good understanding of how the market might react to my offers. I get back a lot of control and then I can do the fun part, which is come up with strategies and think more broadly about how I want to merchandise my product.  

Chris: One thing I love about that is we got to keep it fun.  

Nick: Yeah.  

Chris: Well, Nick, thank you very much for sharing your thoughts and insights on this. We appreciate it and we're looking forward to working with you on this journey.  

Nick: Alright. Thank you very much.  

‍