American Shuns Sticks, Tries Carrots to Woo Agencies Toward a Better Way of Booking

American, NDC, Technology

American rolled out a major new program on Friday that’s intended to encourage travel agencies/travel management companies to start booking flights differently than they do today. Airlines have long wanted to shake up the way their tickets are distributed, but this is the first time I can remember a major airline making a push to do so using a carrot instead of a stick. Not to play this up too much, but this could be the spark to effect big change, and it’s a good change at that.

Airline distribution is complicated, so I consider this a personal quest to try to make this post understandable. I don’t want to get into too much backstory, so to get a full understanding of how airline distribution has evolved, I’d suggested reading my three-part series on the topic. But I still need to get into some history….

Longer Backstory Than I Hoped
There has been a near-constant uneasiness over the last couple of decades regarding how airlines have sold tickets. If we flash back 25 years ago, the internet wasn’t for e-commerce. Airlines sold tickets directly via phone, ticket counter, or city ticket office. They also sold a tremendous amount through travel agents. Travel agents were great ambassadors for the airlines, because they could sell tickets without the airlines having to employ them to do the work. Instead, the airlines paid a commission to travel agents who sold tickets as well as a booking fee to the computerized reservation systems (now called Global Distribution Systems, or GDSs) that the travel agents used.

E-commerce changed everything, and airlines were very early adopters in the mid-1990s. Since flights were primarily chosen based on a mix of price and schedule, they were easy to sell online both through the airlines directly and through third parties that became the online travel agents we know today like Expedia. It was cheaper to sell tickets that way, and so airlines started offering discounts to people who booked directly. The strategy worked.

While this was going on, your standard travel agent became almost instantly irrelevant. Airlines ditched published commissions, and travel agents who booked air disappeared rapidly. Only those who could justify adding enough value to get travelers (instead of just airlines) to pay for their services survived. Oh, and of course, the big Travel Management Companies (TMCs) that handle corporate accounts were fixtures as well.

The GDSs, meanwhile, had become fat and happy. They were paid for every flight segment booked by the airlines, and it was a nice business to be in. Then they paid some of that money to travel agencies to use their systems as a way of preventing those agencies from looking at alternatives. But as soon as the airlines could sell lower fares directly, the GDSs woke up. They engineered a brilliant (if mildly-nefarious) plot to dramatically lower the fees airlines would pay in exchange for the airlines promising to always make their lowest fares available through the GDSs. You probably remember all those web fares back in the late 1990s. They disappeared because the airlines made this deal.

Over time, the airlines worked hard to get away from selling just by price and schedule but the GDSs and online travel agents didn’t really do much to adapt to the new world. The airlines unbundled and wanted people to buy a base fare and then add on everything they wanted. The model wasn’t a problem as long as people could compare apples to apples, but they couldn’t. The GDSs weren’t very good at supporting these ancillary sales and so the experience stunk.

A Standard is Born
Eventually, the International Air Transport Association (IATA) realized that this new way of selling required a new standard. And so, the not-so-creatively-named New Distribution Capability (NDC) was born. Once adopted, this provided a data transmission framework that all airlines would be able to use to sell their tickets, including ancillary services after purchase. This standard was hugely important. Previously, individual airlines started to develop their own systems for so-called “direct distribution,” but that meant it was expensive and difficult for third parties to develop around it. And if they did, it would only be good for one airline. With NDC, third parties could develop to the standard and then add additional airlines with ease.

At first, many saw this as a way of getting around the GDSs. If they could sell directly, they’d be able to eliminate a middleman, keep costs low, and be able to sell more ancillary services. But it became evident to many that trying to get around the GDSs was not the right strategy, especially as booking through them became less costly (especially in the US). The GDS today is an essential part of the business. The important travel agencies/TMCs use them, and there is no solution to replace them fully. There probably won’t be for years because of all the functionality that will be hard to replicate.

Some airlines in Europe have continued to fight this fight against the GDSs. Notably, Lufthansa Group and British Airways-parent IAG both have or have announced additional fees when their flights are booked through the GDS. But it’s a poor strategy since it penalizes the airlines’ best customers.

American Kills Them With Kindness
I realize this post is already long, but it was important to build the story. Now, let’s talk about what’s happening. American is saying, in effect, “we want everyone to use our NDC connection.” If you’re a GDS, great, use it. If you’re a new technology company? Great, use it. If you’re a travel agent, perfect, please use it. It’s pushing this in two ways.

First, American is making the technology available for any agency that wants to use it. Agencies can use the standalone SPRK desktop application to book with American if they don’t want or need to do any technical integrations. They can also work with a number of third parties which, now that NDC is catching on, have developed applications to interface with other travel agent systems. And lastly, they can develop something in-house if they’re big enough, and then just build it into their tool. Concur, for example, is working with American to integrate NDC.

Why would agencies do this instead of just sticking with a GDS? Well, there are benefits, especially for TMCs.

  • They can create bundles by company so that certain attributes (seats, etc) can be always included. This can even include sub-groups for different parts of the company that get different things.
  • The waivers and favors system can automatically integrated so there’s no need to toggle back and forth between the GDS and American’s system.
  • They can improve the duty of care by showing travel agents not only if someone was booked but also confirming if someone boarded the aircraft.
  • Agents will be able to sell all kinds of ancillary services in their systems. That includes upgrades on day of travel, same day changes, and same day standby. They can pre-order meals as well, with pre-purchase of paid food coming. And they can sell paid seat assignments, the only ancillary service Sabre (a GDS) supports for American today.

That all sounds nice, but American has finally realized that’s not enough. That’s why part two of American’s plan is to actually pay the agencies the same way the GDSs pay them today. They need to make them whole financially. For now, American says it will pay $2 per flight segment booked, and any agency is eligible. That’s the metric GDSs use as well, so it’s an apples-to-apples comparison. And at least for American, this is a competitive payout. For a small to mid-size agency, it’s probably a very favorable payout.

Sure agencies have agreements with the GDSs and have to meet their targets to get their payouts, and that’s why American is treating this so casually. There are no commitments and no minimums. Just book what you can using NDC and you’ll get paid.

American, naturally, will only pay this out for American-marketed flights. Agencies can book some other airlines via the NDC connection, but they probably wouldn’t. They’d do that through the GDS. In the meantime, if agencies have hit their goals with the GDSs, they can start booking through NDC so they can get all those added features that don’t come with GDS bookings. Sure, some development work is required (or at least a contract with a third party), but every airline is going to move toward this standard. Once it’s set up, it’ll be easy to plug in additional carriers.

This is a fantastic approach since it basically just opens the door and tells agents to waltz on in anytime they want. There’s no pressure to make the switch, but the combination of the financial and commercial benefit to doing so means agencies are going to be interested.

This kind of pressure is key to making GDSs change their ways. The threat of lost business is the only thing that will make them move. And American is effectively telling them to start using NDC or other methods of aggregating and selling tickets will pass them by. This isn’t about killing the GDS. It’s about dramatically improving functionality for all who use them or any other system. And that’s why it seems like this approach may be a winner.

My apologies to all of you who suffered through that post only to realize you don’t care about distribution… at all. It’s pretty fascinating stuff, but it can get wonky. Any questions? Hit the comments.

Get Cranky in Your Inbox!

The airline industry moves fast. Sign up and get every Cranky post in your inbox for free.

20 comments on “American Shuns Sticks, Tries Carrots to Woo Agencies Toward a Better Way of Booking

  1. My dear late brother used to say, years ago, he would gladly fly if he could only figure out where and how to buy a ticket, and he didn’t feel like he was getting screwed every time he paid for a ticket!

    CF, you explain all of this distribution stuff better, more interestingly, than anyone, but this is all part of why I, and I’m sure others, have such problems with the airlines. Fare structures that seem so complex, complicated, convoluted, service and distribution of product, the same, and all for what–not really to serve a customer well, but to figure out how to make the last cent of every stinkin’ seat, the control of every means by which tickets are sold, ultimately the ability to kill off or cripple, almost every last competitor, whatever. For what?

  2. I appreciate the effort to explain everything but I’m still confused by all the insider terminology. The internet killed the travel agent, that I get, but the rest of the acronyms just make my head spin. What I’d like you to explain better is the part about the GDS’s having super cheap web fares in the late 90’s. As someone who used to get “name your price” tickets on Priceline for insane cheap and then suddenly one day that window slamming shut I’m curious. It made sense to me that an airline would rather sell the seat than let it go empty, and often I’d fly on a $50 ticket and have empty seats all around so it fit. Always figured the change was airlines getting smart about seat management and down-gauging after 9/11. Was there more at play?

    1. It wasn’t the GDS’s that had super cheap web fares. The GDS’s were selling the normal fares. The airlines realized they could by-pass the GDS’s using the internet and introduced the super cheap web fares directly to the consumer if they booked directly on the airline’s website. The airlines were trying to incentivize their customers to book on what was the least expensive distribution platform for the airline (the airline’s own website) rather than through a travel agent using a GDS (with all the associated fees and commissions). The GDS’s realized this would kill their business, so they offered discounted fees and commissions to airlines who would agree not to undersell them on a different platform (airline’s website).
      I don’t think this had anything to do with Priceline’s model though. Maybe Cranky can explain what happened there.

      1. I reckon Priceline bought seats in bulk at a flat rate on a range of routes and priced them as they saw fit.

        So “name your price” was you competing against other customers, Priceline had already paid for the seat and it was their job to sell it on.

        That can be attractive to airlines as someone else is doing all the hard work but ultimately as an airline you are handing over control to a third party, who might not share your (sorry for this….:) brand values.

        Am being snobby but “CheapOAir”? No thanks.

    2. Airfares are fascinating (and complicated!).

      I know that working with travel agents meant that each airline had to have a support network in place which was expensive. And they were paying commission on top. For a while that worked and I still think that airlines don’t appreciate all that travel agents do. Customers got a wake up call during the ash cloud disruption and there were those whose travel agents helped get them home and those who had to rely on dealing with the airline directly (how long are you on hold to your airline of choice, usually?). So there’s still a (much smaller) support network but not so much commission being paid.

      Anyway, I digress.

      Why was the Priceline name your price dropped?

      Because airlines became averse to allowing a third party control their inventory. Relying on a third party to sell your stock isn’t ideal for most airlines in the internet age. Additionally, it creates a situation where customers hold off booking because perhaps Priceline will be able to offer a cheaper name your price flight. And airlines really don’t want to be relying on that to fill planes. To them that would be a failure of their own marketing and pricing strategy.

      Airlines also realised that last minute seats could be sold at a premium if people were desperate to get somewhere. How much easier to change the narrative to “book early” (airlines can plan better) rather than taking a chance on a last minute deal. Sending the message that if you wait you can get a $50 flight is a terrible message from the airlines point of view. It devalues the seat (and Basic fares do that nicely!!! ;-) ) and airlines prefer overbooking at higher prices to optimise capacity rather than scrambling to fill the plane at lower prices.

      I realise this this post is all over the place (!) – am slightly jet lagged!!!

      1. I would add: these days there are more travel agents that pre-internet. At least in the U.K. Yes we have fewer high street shops but they have survived.

        1. billy – The thing about travel agents is they now have to be much better than they were. There are a lot of basic things that can be done better via self service, so agents had to learn how to add value. Most do that on the land side (tours/cruises/etc). But then there are companies like mine, Cranky Concierge, that came out after the commissions went away. The focus is on providing a service that’s worth paying for, and people have responded to it. It just required travel agents to up their games.

          1. Right on! Agencies have to specialize, and be experts at their given specialty. I was stuck in a bed and breakfast near JNB immediately after the 911 attacks. As soon as SA said they were flying again, all the game hunters in the B&B called their travel agents (not SA). Those agencies specializing in big-spending pursuits, such as big game hunting arrangements, will not just survive–but also prosper.

    3. A – I try to define every acronym I use, so if there’s something you’re missing let me know and I’ll try to clarify it. It’s complex, but I don’t want acronyms to make it even worse.

      As David mentioned, the GDSs didn’t have web fares. The airlines offered those directly. (I remember at America West, we had Surf ‘n Go fares.) But when they agreed to give all their fares to the GDSs to sell, it took away the incentive to put them out there. So they quietly went away.

      As for Priceline, the airlines just became much more skilled at filling seats. Load factors were far lower back then, so it was more common for airlines to give Priceline aggressively low rates that it could sell. As the airlines figured out how to fill those seats at their own, higher fares, they started giving Priceline less availability and at higher fares. Eventually the volume was so low they just walked away.

  3. Ah, yes, the good old days of Eaasy Saabre, et al. My travel agent loved me, because I’d do all the work and she’d just have to issue the ticket and collect the commission. And all was well in the peaceable kingdom, until commissions went away… and so did printed tickets. So now we have peaceable kingdom version 2.1, and while there used to be as many as 4 travel agents to choose from in my county of 19,500 people, now there are none. but I get along fine without a travel agent, but there is one available 40 miles away — I think, still — when I need one.

  4. Thank you, Cranky.

    It is not a sexy post, like you admitted, but it is fascinating to read and to learn a little bit more about things that happen behind the scenes. It may not be the most interesting subject, but it is useful and now I know a little but more about distribution. Keep them coming!

  5. Great! Now how do individual travelers get the $2 payback? Probably not by booking on, because the payback would go to which is part of American itself. But if I could write my own program that interfaced directly with American… just waiting for some open-source libraries to show up that will make this easy :-)

    1. In the U.K. we have Quidco and TopCashBack. Which is essentially the commission from online travel ads being directed back to you, the customer.

      I think there’s a US version.

    2. Ron – Well you’d have to become a registered travel agency first. I’m pretty sure it wouldn’t be worth the investment just to save $2 a flight segment.

      1. I could see someone creating a non-profit registered travel agency with the objective of taking the $2 per flight segment and directing it towards a travel-related charity (e.g., towards airfare costs for the families of kids who need specialized medical treatment). I’m not sure that American would like it much, but it would be a huge PR black eye if they tried to shut it down…

        Amazon (through Amazon Smile) and some supermarket chains (Kroger among them, IIRC) direct a small percentage of shoppers’ spend to charities of their choice. This wouldn’t be that different.

        Alternatively, I could see an airline do a promotion to the effect of, “Book directly on our web site and we’ll donate 1,000 miles towards the airfare of [worthy cause],” to encourage travelers to book directly. I know I’ve seen hotels do something similar with drop boxes for room keys by the front desk, and I’m sure this has been done by airlines before.

  6. That was a great article!

    I used to be a travel agent and we used Worldspan but in the U.K. there were also two pretty basic “viewdata”‘systems. One called BA-Link and the other ABC Easyres.

    They could book basic point to point flights but nothing complicated like open jaw tickets (at least not that I can recall).

    BA-Link became quite infamous as a dumping ground for capacity, and we would often get prices that weren’t on Worldspan (our GDS of choice). Then there was a golden age for our little agency when BA axed the 7.5% and 9% commissions and replaced them with a £7 per sector fee. For us, we would make £28 on say, a GLASGOW > MANCHESTER > PARIS round trip which could be as cheap as £99 rtn. Good times!

    ABC Easyres looked similar to BA Link, and covered many other airlines.

    Both systems were incredibly simple to use (none of this A12SEPGLACDG1200 etc.) but also didn’t have the likes of Worldspans Power Pricing.

    I don’t work in the business now but really miss it. I used to actually READ the ABC/OAG Airline Guide and BA’s Fares Fortnightly!

  7. Cranky, how do services such as (who direct you to,, and similar sites) offer lower-than-public fares? Are they traditional consolidators? Or do they obtain their inventory some other way? (Right now, for travel to London on specific close-in dates, I’m seeing sub-$1000 fares on bookingbuddy sites, vs. over $2,000 on and itasoftware. It sounds too good to be true….)

    1. UAPhil – That’s not Booking Buddy doing anything. It’s the third parties that it links to that are getting consolidator fares. Hard to know if it’s legitimate until you ticket it, and then of course, you don’t know what kind of hornet’s nest you’re stepping into if you need to change or cancel.

  8. Congratulations on a very clear article. It was very informative without being too “wonky.”

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.

Cranky Flier