Topics>>A Digital Experience That Enhances Your Physical Travel Product - CarTrawler

A Digital Experience That Enhances Your Physical Travel Product - CarTrawler



  1. Middle East7:48
  2. Green Party8:03
  3. Google2:46, 5:24, 6:36
  4. Quetta15:23
  5. computer programmer0:27
  6. collect data2:25
  7. cost risk15:29
  8. indicators signal10:22
  9. ho chi0:38
  10. neural network3:52
Fri, 19 May 2017|

Bobby Healy, CTO, CarTrawler


Automatically Generated Transcript (may not be 100% accurate)

Everybody. Congratulations die for travel on twenty years actually that now that. I think IPO at every single one for twenty years as well so people mostly boron other mines listened to me. As well correct basis of the cameraman undershirt today well they can decide that there. And so under siege on the computer programmer side com from the different base that Rowland did you know hour I'm there in much smaller company of 600 people the surest fix on the people and bipolar. And the first week in ho chi calories which was that rather limiting name of this type of business Kyra the front of the and I get through that would turn this that the subject is tricky one for me get my head around the fellows point fluffy. I tried to. They aren't subject in important deliver you know. Insight into from a technical point of view with the challenges are to delivering aid did you experience that hanssen's your physical travel products. Obviously car followed don't have a physical travel farther a working with. Over Honda airlines. And many other OTA says I don't need to be partners. We tend to put that lipstick on that car business told you know well we do for anybody who doesn't know as we sell everything around ancillary to. Does customers. Dawn when we look at you know how we've grown our strategy we're about eleven years old now. As a company but will we. You cannot put our strategy the very start you knew we require a very stark work for a six or seven years. We set about to be upbeat view platform that's just technology just as riding cold. He got my linked in profile of these in the first call of ever written some of the best kind of ever written. M you know overbooked from a strategy point of view we look at you know that do the time and the trip timeline. The same way that they do we define it you know in the in the stages of of travel you know from dream on way to experience. The way we participate with regard the wire consumers we don't have any direct consumers ourselves as a business we Europe basically having a lot of people's cost the most. And we look at look had we participate. At all stages in the trip planning crisis which has you know. 330 days. Prior to two to trip and I wanted to trade and we learn about the cost Merkel for behaviour. And only collect data in around that. So if you look at our business in the partners that we. Air work with wiwa who car rental companies we work with you know many many airlines we were met the search companies to work through much of the channel's Hillary in the airline industry sell. We'll see Europe customers. When they gold searching on Google and deep linking T Euro airline website we've seen the same cost amount when they visit in their search channel. And then we've seen staying close harmony gold to cover our brand Dhaka where we power the end in so we see the course of behaviour across multiple channels. And most importantly we see them. Pretty much the entire trip planning timeline. So we learn their behaviour we've been hit really inside. And to customers that we collect we story and the Zahn and we do lots and number pointing out. Fully we aren't used machine learning or a item look at that day that. There's much simpler approach is blue force approaches to learning their cost behaviour. Or more suited to that type of pattern. We do use machine learning just to reference in the area topic. And on pricing and relative ranking of carton parked on the display sell which both the showed which cost the most. And that's an area where or if like job in fifty cards in the display to you know hundred million customers. If I started Skadden clients around display and randomised to results. So they can learn battle in Europe network use in Europe neural network users display its array of different options in the can learn behaviourist. And likelihood propensity basin relative value to clients. So there is the role for machinery and AIN outlets for about earlier. It's far more concise defined. Down yet every news and AI today in machinery today forever in that they do just enough in this nation we need to business and it's not limitation that not ballads. Without a huge day recess. We have 800 billion data points in our our and the Zahn warehouse right now. So we have an absolutely massive amount of traffic that informs and network and even then. It's dinner last of this foresee area cell if you got to make a big investment in ARA. A machine learning deeper meaning a Euro next don't do unless you have very light day recess. And similarly if you working with a small company specialised machine learning there are now empowered to do anything for you unless they have very light day this ad. To actually. Would you need to do versus fact collecting data and building data. N is is this an overview of the way we can just cleared the way we look at the costs amid digitally crossed the journey you know you have your timeline with with flights people book hotels. People book their rental car directly than they do activities and salaam. And then there's the participation on the need that of where the various players are. Where the opportunities our. Where we participate titled this data Barnum. And then the big giant elephant in the room Google FaceBook trip until aren't particularly who goal. And do those. Or we column that effectively as an attack. Aren't the top the final. Two intermediary to capture the top fall to take away the top performer. Or extending monopoly. To take on fairly the top performer from airline operators. And LTA's. So there's a big threat here. To everybody to two or take control or to are on board to. Two to participate in you know control way tie up the trip planning and the dream part of the experience sell. As a brand we believe that our partners and the dollar view in the room that operate physical product. Need to control the top the fall if you know control the a couple of follow the cap understand customer behaviour. In the way that it need to. So again we see that traffic because we're there we're with bear you know in the cases involve the airlines in the mobile apps on the airlines so we see behaviour. And we understand the influence our branding. Top performer participation. Airlines particularly air lines but also put out companies and LTA's how important it is to them. And if you allowed who have controlled the top final Google will keep cost more in Google product Google sites. Forever and the point that extend that further and further. And that's very bad for the industry. We think about red you know again how we participate and how are airlines or our other clients participate with the cost the murdered as one thing about you know timeline relevancy and being here for the cost them with the cost more digitally. Everywhere they goal and every time they goal in their pockets. In flight departure of a rival therefore. In destination. And you know that's timeline that about you know been here beside the cost wrote to be helpful. With information and to be helpful would cross selling a lot of products and only think about product relevancy. That's another thing that's about how the everything that the customer needs such as the sound way to much of the operas the obvious car hire and you know hotel Prague that the top three in insurance. But it's also about you know of the Kosovor arrives in the way you know third of the delta the luggage belt tell what are options are to leave the airport. Do you see there the day appoints the we have we believe that the big survey. You know you're likely Middle East as well on the wireless a parent and so data but. You see that very different forms of transport at this happy believe airport I should say. So in North America most people leave and airports you know. Friends and family. Car versus the Green Party was in Europe obviously public infrastructure Raila boss of the major factor. Taxes are a major factor because of Europe density. Told the we are our mission has been essentially. Colin Tara timeline by. In the timeline and then offer every single product that's relevant for the cost them. And that's not about I mean if you if you see we do for example would with Ryanair or with emirates you know already these costs as we work with. It's another you know vomiting off every single product onto the display. We wish just isn't relevant anymore I mean you cant. You can work with a mobile interface on mobile relationship that cost them and have anything more than one page one finger swipe prior to show them. So the relevancy of the product is absolutely Paramount so where as we build our business title of the business fun car ha. And the unit economic serve car are absolutely fantastic. You know only 5% of cost the most have a need for car you know put systemwide 5%. So the low 95%. Of course the most did leave the airport. Or go to the airport Dodd the need for our product. Now well we want to display our product absolutely everywhere need now sticker. You know one person to case is just not relevant so we have a look at you know what does the customer need. And that is where behavioural understanding of the cost as important. By David backs that up it's important that where real time decision making is important so when we. Trying British not to be in advertising play because technology Khyber. Did do. The way we're successful with our airline partners is we'd comp hide their engineering team. And we handoff a lot of data each other and our analyst day that totally legal but there's a lot of David Shane happen so we wean we see. For example when you travel through the flight purchased on or whatever airlines. Dependent on the airline we'll see the price of the ticket flight ticket that you're about to buy them for eligible we've seen the price of the ticket better by. With you the flexibility of the fear not necessary the fare class we see the flexibility or the option now thirty. That you've gone for under fear did you do things that very indicators signal for a now driven. That would say you know you wave your propensity to go look jury. The goal. You know low cost way ever to there's a lot of signals in the day that the not to do. With customer profile course my history there to do with the behaviour within the track the transaction the current transaction in real time. That inform what I've we show whether or do we show in. Even the size of this so we have a choice of many different supplies we work with you know 2000 car rental companies and in anyone very reports we might have. Farrell 48. Different car companies. Two commanding rather right so why relation L 48 of them so I choose quality level possible rating level and the and the weight versions Sheldon. But most importantly deliver bottom graph. Now we have actually offer you know and we make a lot less money but we lost for a tax laugh and undermanned tactic. You're right with the luggage belt. You get notification and your phone how you get into an if we know your hotel with say how you get to issue over a tale. And would probably laughing attacks the which loyalty points forget that hotel rather in the car hire. Because there's only 4% take off raid at the desperate car are very low propensity to to block. What you there and even then it's a manual Java snuff company. You can intermediate. So we look wider eruptions. We leave enough for coal which. Or we'll offer rail ticket. Both of which the you know economic so horrible just make if you sense on the products. Where she might make tens of euros in the car what it's about given the cost relevancy. And it's about not the particular transaction debated did in relationship. So that if you've given them is actually providing value it's just it's in the Amazon for last thing right. If you offered Iran prior. It's it's enabling up the annoying because it is need to avoid which enough thinking about that where that relationship digitally with their airline or something of value. But if that fills me. What are my options together this airport as quickly as possible. And I'm sure everyone here travelers' love airport one of the biggest problems is had a hell do we get my hotel under preparation. So we make it is easy as possible. And that's tied to do if it's a hard job technically to do just on the standalone basis it's hard work within giant airline companies think Roland you said. You spent 200 million the year on technology. When big companies and smaller companies or type aligns those two opt again there. So as the sit tight you know organically in internally must either entered toward the technology partners well. And separately it's very hard to to contract this shelf. You know we you know we have a contract taxes paid graduates basis solutions in 1600 airports now. That aria airlines fly into cities in. Can't tracking problem and the logic bomb that place them as a deep revenue management and pricing problem because. There today you being in getting the right price point the rightly plasticity measures things like together right of the there's a lot going Ansel I would suggest for a minute that anybody. Would would start off try to do that under Iran I would suggest that you should outsource that effort. Obviously car controllable power rather. You know nearly equal options. M. So again that's just in the view of goods into giving into that indicated scale and I would say the real them. Tipton about our scale is is not to brag to more contended that its own in the last two years I would say that our scale is insufficient. To give us another day here. To make informed decisions around data rather than. You know will be object or subjective evaluation. We used to bring customers in an ass and we used to bring partners now for news survey we used to do with. Human beings alive. And arguing now we're we're now like the airlines that we parker worked exclusively nearly billion passengers PV using year so we've just have very large and diverse Davis said. But only allow us you know really Copley years I would say. That add rhythmically. We do better than just human guessing what's the right thing that we might want to do here. And now out rhythmically we can go much more granular. About things that we wed on top about. And on the web Proehl filed a had agreed to segment of customer types. We we do everything in real time and tiny little pockets and we use triggers signals that the airline partners pass this in real time to make a decision. We can you know as a said that's only been successful for us. You know hands over the top developed partly for ten years but really only makes a material move the needle that work while investing in in the last couple years we got to. Probably over have been in you know annual cost most of we see. So that's you know margin you know because an overview of the of the timeline these products that we. Put together you know we we hit the obvious ones car rental train bus college transfers under man ride which we think that if they're off to 2007 dean. Quetta surveyed 60% of relevancy. So you're giving you have a tried to cover a 60% cost risk needs. Old when they get to the airport when they go to the departure airport and when they're leaving the departure ideas arrive at their point they say. And we think that. You can go to web forum that more than by offering in destination father. So fraud that's on the man taxi heading so we have an undermanned Prague does not compete with Hoover it aggregate dollars guys. But most importantly. For the digital relationship pointed view weekly news loyalty points to near a ride and we can earn points out of favour. Is the extent of the loyalty program in demanded now and we think that's hugely relevant product to offer the cost the most in destination particularly. Other light network airline those customers that may have over when there in the stayed put over doesn't exist in many many countries. And similarly when you go to this data might prefer to use lift. That a married up with TV where ever so do there's an opportunity for a airlines to a a LT aid actually aggregate intermediate. Based on the loyalty program. And then obviously activities and the being really you get has gone at the moment is. Nowhere are we see now there's a big shift to non violent and frost that's represented. It's about 20% we see. About 20% of our carriers. Mob buddies mobile traffic on for two of the very small part of that is native mobile. However want to buy European airlines I'd close to 20% transactions. On mobile apps now flight flights booked on mobile which has really very impressive considering they've gone from 5%. You know only two years ago till what I like about added now we have the blue dart approach to work with cost of rent apart between donate about. We can do notifications we don't have to Allianz do that emails and tracking. And we have their location so we know what hotel they go into even though in my another foot of with us we know their behaviour in destination. And so there's Allah you know obviously that's a bit scary. Well it Lotus could be more relevant with how we communicate the cost has allowed the track them when our interests and that at about relevancy. So we know that there related costs from a weak enough of them activities between our business customer is valued confides as well. So suffer from me again something our. Or to bribing their lives customers' lives and blue sky that's a picture outside my office went on double them. There. You know look it's a tight again did you relationship right number one it's easy to build technology. But very very high deflected huge day recess and you can only really optimise the business forget the fundamentals right of the technology right. That underpins that led to use a huge day this out. And that is for me the key to a property's relationship. Thank you very much.