Part 2: Getting in the Door, and Getting Adopted
Jim Barry
So, to Patrick. How do tech providers get into the consideration set?
4. A Great Idea Isn't a Business Case
Patrick O'Keeffe
Sure. I'd first say it's difficult, we're a skeptical bunch of buyers, because our business is complicated, and since everyone consumes our product, they think they understand how it works behind the scenes, but typically they underestimate the complexity. We're always interested in talking to someone with a better mousetrap, but we need to make sure it fits within our limited set of priorities and financial resources. The things that make people successful: a proven track record in our sector, being able to speak to the complexities of our business, our data, our employee profile. A proven record of stability. We're a twenty-four by seven, three hundred and sixty-five day a year operation. We can't afford significant outages. If your technology stack doesn't have a proven success rate around stability, we won't want to talk to you. And anything you can do to help us improve efficiency, given our low-margin business, is going to be helpful too. But it's difficult to get our mind share if you haven't had recent experience with us or one of our peers.
Bridget Blaise-Shamai
It's not enough to have an in with the CEO.
Jim Barry
Probably too many people abuse that. Next question, very similar. At what point should a technology company engage with airlines?
Patrick O'Keeffe
Back to the mousetrap, I don't think large airlines have been rewarded for taking the first-mover advantage around technology. Bring us something that's proven, proven at scale, and where you've demonstrated legacy integration. If it looks like a super cool tool but you haven't figured out how to connect it with our technology, it's not going to be super helpful. And particularly in the case of American, prove to us that you understand the ecosystem you're connecting to.
Jim Barry
All right, terrific. Jill, you've led operational teams including flight attendants, airports, and lounges. What factors influence tech innovation and adoption for what you've purchased in the past?
5. Adoption Is Won on the Frontline, Not in the Pitch Meeting
Jill Surdek
I guess what I'd say first is everything we just said about why it's complicated, and at the heart of it, there has to be either a very clear ROI to get attention or a very demonstrable operational efficiency or improvement. But let's assume you've passed that bar. What's unique about operations, and what makes it challenging, is then how do you get something actually adopted? Once you build it, do you get adoption across very large, dispersed work groups? That's one of the biggest challenges over my career, and I don't think we invest enough as an industry in training and communication. So many technologies are super well-intentioned and right, and then don't get utilized.
That's partly because if you're going to require employees who are long-tenured in many cases, and working under contractual requirements, you first have to make sure whatever change you're contemplating is in compliance with whatever work rules you have. Some things are disqualified right off the bat because of that. But assuming you've passed that bar, there's the employee angle and the customer angle on how you don't put the complexity of the technology improvement onto your employees, and where you can make it automated in their workflow rather than an additional tool they have to go to. A lot of times historically, we've just communicated with email and memos, because we don't have an extensive history of training besides safety and regulatory in this industry. So the most successful adoptions are when you put money behind frontline management resources and training resources.
Jim Barry
Excellent, and a good ROI is only half the battle.
Jill Surdek
Exactly. One of the things I'd recommend from my operational experience is you cannot make continuous change in an organization the size of a major airline and expect it to be successful. So you pick the couple of things you're really going to focus on with a very large work group, and you relentlessly focus on performance and accountability. You can't change thirty things a year and expect a dispersed work group with very little supervision to hear and embrace it all. Even in an airport, you've got people across gates and terminals, often alone as they work their flights on the ramp. You have crew chiefs, but your managers are dispersed across a terminal. You don't have frontline leaders present in the operation to the same degree other industries do.
Jim Barry
Patrick, you've spent decades leading large-scale tech and business transformation efforts. Were there particular categories of technology that consistently proved harder to deploy than expected, or easier than you thought?
Patrick O'Keeffe
The more enterprise in nature the technology change is, the more challenging it is. And what makes those work is just appropriate levels of governance. I was involved in one where we underestimated the complexity, it was late in our integration program, so we didn't resource it appropriately, and it took twice as long and twice as much money. That's the downside example. We got it done, but the success was post-merger at American, when leadership decided there were five things we needed to get done in the first couple of years. We were going to prioritize those, resource those, and the whole company would know that was the priority. Jill and I worked very closely on the reservation system integration between American and US Airways, that took eighteen months and a lot of money, but it was a great success. Bridget did a similar exercise before us around the loyalty system. So large enterprise programs can be done well, they just require a lot of governance, a lot of business and IT leadership engagement, and good focus from the organization.
Jill Surdek
It was the prioritization, right? And the discarding of other things that seemed like good ideas.
Bridget Blaise-Shamai
Ruthless. Ruthless.
Patrick O'Keeffe
Yeah, we had a change management meeting every two weeks where five or six leaders would sit in the room and managing directors would come in with really compelling issues from their line of business, and we would just say, "no, we're not doing that." You come out of these big programs with a huge backlog of very important initiatives the company wanted to get done, which can feel frustrating, but the good news is it allowed you to focus and get the biggest priority done to the best of your ability.
Bridget Blaise-Shamai
I think one thing Patrick gives detail on, that's really well done at American, is that business and IT work really closely and effectively together. Everybody understood IT a little better, and everybody understood the business a bit better, despite it not being exactly in their remit. That's a positive thing, Jim.
Jim Barry
What separates technologies that get evaluated from technologies that actually get approved and implemented at scale?
Patrick O'Keeffe
A couple points we haven't covered yet. One is strong leadership from that organization, a sense that we can really trust they're committed to this technology over the long term, that it won't just be a vendor relationship, it'll be a partnership that helps us get to the results we wanted, work through the inevitable surprises in a large project, and ultimately make us both very successful.
Jim Barry
What are the characteristics of a company that signals it's going to be more of a partner than a vendor?
Patrick O'Keeffe
I'd say they've staffed their team with some subject matter expertise in the industry. One of the great qualities about American is that we have such strong subject matter expertise in so many parts of our business, it helps to have your partner have some of that expertise too. It shows their commitment to learn, the investment they're making in understanding our business. When they come in with a pitch that's not just the sales guy and the CEO, but a couple of folks who can go into great detail about the business processes their solution will support, that gets them a lot of credibility.
Bridget Blaise-Shamai
I think skin in the game is also a good way to establish trust. Perhaps it's a period of time before the meter starts running on fees the airline incurs, or it's total pay-for-performance, the outcomes promised were realized, and then compensation is provided. Versus the airline taking all the risk under a time-and-materials or licensing arrangement, and then it doesn't manifest as set forth in the model. When you're willing to put a little more of yourself into it financially, that feels a whole lot better, and builds more trust.
Jim Barry
That's a terrific topic. So on one hand, don't put all the immediate risk on the airline, let them test it, evaluate it, prove it out a little. On the other, you might say it's going to save X, charge less up front, and the airline pays for more of the measured value once they see it. But that can get expensive for the airline. How do you manage that?
Bridget Blaise-Shamai
My experience is the airline gets a little nervous about a complete full-on pay-for-performance model, where the risk-reward pricing becomes a little uncomfortable, because we operate at scale, before you know it, the thing has ballooned and your fee has too. So certainly in the conversations I've had, we've always had some sort of floors and ceilings that made everyone comfortable enough, while still feeling like we weren't fully holding all the risk. Tying back to limited time, limited capital, limited human deployment, and then realizing only a fraction of the promise, that's one way we thought about it on the commercial side.
Patrick O'Keeffe
And I think there's also a degree of skepticism about our ability to really measure that, and therefore be comfortable paying out on it. So I think those kinds of deals were certainly more the exception than the norm, at least from an IT perspective.
Jill Surdek
One thing that's changed in the last decade is an ability and willingness to do a little more testing. You see more airport-by-airport solutions now, and that gives you the opportunity to do a proof of concept, whether it's boarding passes, compensation, or reaccommodation, at the airport level, which gives you a venue to do a controlled test and really assess performance.
Patrick O'Keeffe
And because of the way we operate, that approach limits the number of employees who have to be engaged, eliminates a lot of the training and change management burden, and lets you focus on a zone within an airport to really prove out a concept before deploying it more broadly.
Jim Barry
So if a provider can't immediately bring proof from a peer carrier, is testing and evaluation a viable way to build that track record? What makes you smile thinking about this, and what makes you say evals and trials can only go so far?
Bridget Blaise-Shamai
I think American's gotten better at that since we've been gone. The example I'm thinking of vividly is the electronic gate readers that are now part of the renovated C terminal at DFW. They tested the heck out of that, and now they're very comfortable doing it and expanding the deployment. The gate agents are having to learn this new thing under pressure to get the airplane out, but the confidence is there now. That's a really good example of a deliberate testing environment that limits exposure but gives you enough to understand the viability of a new solution.
Patrick O'Keeffe
For things that are more enterprise in nature, I don't know that American historically had a great appetite to test out solutions, because we had so many other large priorities to focus on. So we still tended to rely on: give me a reference that is a company of our scale, with our complexity, with our legacy environment. I want to talk to people who have gone into the war with you and come out successful. That's a far more useful use of my time than going into some pilot program where sixty days in, I realize you're not going to scale for us or you don't understand the legacy complexity.
Jim Barry
But the provider is thinking, "I've got to prove it somewhere, that will have to start as an evaluation."
Patrick O'Keeffe
It can. It might make more sense to go to a smaller carrier that has perhaps greater cost focus than American does. Cut your teeth on a smaller scale, then slowly ramp up in complexity and scale, and eventually make your way to a business case you can present to a major.
Jim Barry
Bridget, you've worked extensively in loyalty, strategy, partnerships, data, and digital initiatives. Why do some companies struggle to gain traction with airlines even when they seem to have meaningful revenue upside?
Bridget Blaise-Shamai
I'd suggest we've really covered the themes here, American doesn't solve anything one-off, because it doesn't work that way in the tech environment Jill described at the outset. The trust factor we've all noted, the build-versus-buy instinct. A lot of the things I was exposed to, which were very cool and modern and I wanted to do, I couldn't, because we were still working on the fundamentals. We hadn't nailed the basics of the customer experience or customer engagement yet. So it was hard to build upon a foundation that wasn't quite ready. And again, the prioritization issue, candidly, even the most promising upside wasn't material enough relative to everything else competing for the same resources.
Jim Barry
A technically impressive solution failed once it hit the operating environment, can you share an example, and what did the company miss?
Jill Surdek
For me, the examples are almost always where we failed in implementation, and it goes back to a theme, don't put the complexity on the customer, don't put the complexity on the employee. If you're going to make a change, you need to integrate the technology so it's a seamless part of their process. One failure is when you do all these add-ons, and people who've done the same workflow for thirty years just don't go use it, you've made the technology integration poor, and no one wants to use it.
A second example: overweight bags. That was a huge initiative for a while. We built out enhanced technology, but it was still an uncomfortable situation for agents to collect overweight bag revenue, so they just didn't do it consistently, we never quite got the data integrated and the frontline management oversight in place, despite having improved technology. That was an education issue. And third is just a management bandwidth issue, if you said you cared about it and built the technology, are you actually following up after the fact? From my experience, it's always been one of those three reasons something hasn't worked.
Patrick O'Keeffe
I'd add, go a little earlier in the life cycle of a project. You really have to get a strong set of subject matter experts at the initiation, both to make sure it's defined well and to think through the usability scenarios Jill mentioned. They also have to be the kind of people who can go out into the operation and sell the project, because buzz about the project arrives much sooner than the actual implementation. There needs to be credibility in the operation that it wasn't just headquarters that came up with this idea. If team members can see their fellow colleagues' hands in the design of something, that increases the odds of success, even if it's not a guarantee.
Jim Barry
How important is demonstrating a measurable ROI? And beyond ROI, what else most influences whether an airline moves forward with a tech investment?
Patrick O'Keeffe
American had a strong culture of financial oversight, and everybody knew we had low margins, you had no chance of getting a technology project across the finish line without a compelling ROI. You also needed both business and IT to stand up for the business case, because there were usually elements of both, perhaps an IT investment, perhaps IT efficiency in the long run, but also a business side. Both parties had to stand up and say, "we can get this done." We always had more projects than resources, financial and human, so making the pitch that what you were trying to do really addressed the strategic needs of the organization is what helped you get attention and priority.
Jim Barry
Curveball, sometimes a purchase helps the valuation of the company but doesn't help cash flow. Supersonic flight, advanced air mobility, the ESG era, there have been commitments or purchases where everyone knew it wouldn't help cash flow but could help the multiple on the stock. Does anyone want to touch that?
Patrick O'Keeffe
I think that falls into the category of strategic investment. We did that with technology, with services, with some of our flying, we flew to places where we weren't making money because we felt it needed to be part of our network. Those decisions are a different kind of prioritization than technology. We had enterprise prioritization for where we'd spend resources, and a portion went to IT. The IT discipline was more traditionally tied to ROI and business cases, but there was certainly strategic investment that was a little more esoteric.
Jill Surdek
Building on that, one of the most important things you can do is understand where a company is in its life cycle. There are times when airlines are much more willing to invest strategically for the long term, and times when an airline is going through cash issues, or there's an industry event where no one's making strategic investments. You need to know who you're approaching and where they are in that cycle.
Bridget Blaise-Shamai
On the multiplier piece, I don't know of any Wall Street investor that's given any airline credit for the value of their loyalty program assets. It's all covered up in the break-even or money-losing core product, which these monstrously valuable loyalty program assets can't live without. The Street gives no credit, despite some airlines trying to get them to.
