What makes Cheapflights tick?

February 12, 2007

0Coincidentally a few days after I posted about the new investment body Howzat Media, I happened to hear David Soskin, CEO of Cheapflights, talking at the recent AlwaysOn media conference in New York via a webcast.

According to the Cheapflights About page information:

Cheapflights is a travel price comparison website. Well, if we were in a bragging mood, we’d tell you we are the country’s leading travel price comparison website (and give you the stats to back it up). But, we’re a nice bunch of people and we don’t want to brag, we just want to help you find the best and cheapest flight we possibly can.

So what does David Soskin believe has made Cheapflights and its advertising policies successful? Here is the essence of his talk:

  • “We have been an innovator in the sector since 1996”
  • “Our market share has been booming in the US and doubled our share in a year”
  • “Our model is quite similar to TRAVELZOO except we focus on flights, whereas TRAVELZOO covers everything.”
  • “Search for the travel industry is very important because of its inherent long tail [lots of individuals who want to book flights I assume cg] – there is more to the travel industry than just Expedia.”
  • “Google is very much the market leader for providing leads to airlines, but there is a growing segment called vertical search. Do people come to vertical search? Yes, because for many people it provides a more satisfying user experience.”
  • “If you want to find a great deal from Boston to Los Angeles and you want to find a great price, Google doesn’t help you. You need to come to a site such as Cheapflights where we have selected the best flight operators and, using our own algorithms, sorted out the very best prices for a whole range of different dates.”
  • “The advertisers on vertical sites are gaining much exposure to would-be customers who are so much closer to a purchasing decision than those who come from a Google search. That’s the motivation for advertisers to appear on vertical sites.”
  • “Why are metrics and analysis important? Because there is a very long tail in travel. We display 200,000 city pairs, 800,000 flight offers daily and we have over 80 travel advertisers on our site.”
  • “We are the innovator in the sector, we invented the model in 1996 and introduced pay-to-click in the UK in 2000. We opened bidding for premium positions in 2002 and in 2003 launched the first multi-booking solution. In 2006 we debuted the Cheapflights partner portal where analytics are so very important. We have developed the first place on the Internet where travel operators could analyse the clicks they were getting on any particular route. They can see in real time where they are getting their clicks and adjust their advertising spend and prices they send us accordingly. So it’s a very powerful tool for the huge US travel business to market on-line.”

David finished with a touch of philosophy:

“Finally, we at Cheapflights follow the philosophy of possibly the richest man in the world, the founder of Ikea, Ingvar Kamprad. He says ‘Only while sleeping one makes no mistakes, the fear of making mistakes is the root of bureaucracy and the enemy of evolution.’ We constantly trial new things, some work some don’t, but we keep trying.”

 

Advertisements

XGate: Protect your child’s on-line experience

February 5, 2007

Having only recently posted a short review of Crisp Thinking, who have taken a network-based approach to detecting possible predatory paedophile discussions in chatrooms, I was interesting in seeing another UK company launching a product in the this product sector – xGATE a division of Global Security One. Like Crisp, ,hey are also based in the north of England, Manchester to be precise.

The xGate is a full blown domestic firewall with additional features that enable parents to monitor, if they so wish, their child’s chat room discussions. To quote their web site:

XGate is the world’s first internet security device that is specifically designed to protect your child’s on-line experience. Even when you are not sat beside them!

  • Monitor your child’s chat room activities even when you are not there.
  • Receive a mobile phone SMS message or an email every time your child enters an unauthorised chat room.
  • Monitor your child’s chat room conversations from your mobile phone or by email.
  • Terminate your child’s chat room conversation immediately from your mobile phone or email if you detect a danger.
  • Prevent grooming by paedophiles – keep your children safe!

xGate take what could be considered to be a more conventional approach by integrating this capabilility into a small box that sits between household computers and the Internet.

It seems to be based on word detection and continuously monitors chatroom conversations looking for selected words that you have entered. I would assume xBox pre-populate the word list for you as this is key for successful detection. Also, can it detect only individual words or can you identify combinations of words which I would have thought to be quite important?

When it finds one, parents will immediately receive an SMS or email to alert to the issue and you can:

a) Terminate and block session
b) Shut down the computer
c) Continue monitoring the chat room conversation
d) De-activate the chat room senso

I do like the pro-active alerting mechanism as it will certainly gain your attention – unless your ‘phone is off!

Just after posting this I came across newly launched anti-grooming software from In Loco Parentis


Mobile apps: Java just doesn’t cut the mustard?

January 30, 2007

When I attended Library House’s MediaTech 2.006 conference in December 2006, I bumped into an old colleague of mine Francis Charig. Francis is CEO of the Tao Group (pronounced ‘dow group’) based in Reading, UK. I think I first met Francis, back in the mid 1990s when they were moving to enter the mobile middleware market. The company was formed in 1992 with Francis as Chairman and Chris Hinsley as the Director of Technology. Their original product was a software development platform that enabled games programmers to develop games that were portable across a number of operating systems.
Over the years this has morphed into intent, a modular software development platform for the creation of rich multimedia services on mobile phones. intent enables applications developers to write binary, portable, native and Java services without the “behavioral inconsistencies that have plagued the rest of the industry.”

In my chat with Francis, we were talking about today’s complexities of developing applications that are able to run on a multiplicity of phones. I mentioned to him that with an application I was developing I was avoiding the use of phone based software (J2EE et al) like the plague because it was so challenging, if not impossible, to develop a portable application. Fortunately, my application is a very simple text based application and the best way to deliver this service is through the phone’s browser. Taking that path, portability is ensured with minimum support costs.

The conversation then moved the conversation on to Java and I exclaimed “I thought so”!

However, it might be best to directly quote from Francis’ paper:

“Many expected Java to be the solution [to multimedia application development on mobile phones]. But now the market has broadly woken up to Java’s true capabilities with the gaming community shouting, “I told you so.”

In its defense, Java wasn’t designed for the delivery of the typical consumer services we want to use today based around powerful multimedia. The only way to deliver rich media on Java, even to a fairly rudimentary level has been to tie each implementation so closely with the underlying hardware that the portability of the platform has been all but eliminated, and the variation in platform has been such that there are huge inconsistencies in running the same application across multiple devices. To illustrate the fragmentation around Java, in the mobile games world the requirement to run across so many handsets has meant that publishers have had to maintain literally hundreds of different SKUs of just of a single Java game, and with significant behavioural differences between handsets.

Java was not designed with 21st century market requirements in mind. It was planned with certain objectives and those objectives, by and large, were fulfilled and fulfilled to a high standard. What it was not designed to do was enable services such as the more advanced 3D mapping applications or competitive casual or console gaming. Java is abstracted too far away from the hardware that powers the device to provide an efficient platform and the ubiquity that the market craves.

Consequently, we have seen application developers move in 2006 to writing in high-level languages that output native codes such as C and C++. This is a return to non-portable platforms with rich media APIs tied to specific hardware/software combinations… Our industry has moved from a Java to post-Java market, where Java still exists as an important part of the platform but not the entire platform itself… ‘Write once, run once’. But native code is required to meet consumer expectations. It’s quite a dichotomy.”

This is really quite profound in its consequences and, as Francis says, presents quite a dilemma to advanced mobile service developers.

Maybe I’ll stick to browser-based WML, XHTML or even AJAX solutions for the moment or just give Francis a call!


Crisp Thinking: a child protection technology

January 29, 2007

When I heard the short presentation from Adam Hildreth of Crisp Thinking at Library House’s MediaTech 2.006 event held at the iMax theatre before Christmas, there was something that really intrigued me about what they were up to.

They were hosted at the event by David Rowe who runs Microsoft’s Emerging Business Team and hosts their Startup Zone.

In fact there were three things that I found interesting (a) The service they offer seemed, to me at least, very innovative that focused on an industry pain not being addressed in a significant way by other companies (b) Their approach to solving the problem from a network and technology perspective seemed remarkably different, and (c) their CEO is only 21! You don’t come across this combination too often in my experience.

What do they do?

Crisp have developed a child protection gateway (CPG) that has been designed to sit in an ISP network to protect children and teenagers from the specific threat of online-grooming and cyber-bullying. The CPG acts as the gateway for the traffic from protected households. Because all of the child protection components actually sit in the network layer, not on the client, the Crisp solution could offer unrivalled levels of protection. It is extremely hard for even a “well tuned” teenager to circumvent their network level controls.

The core of the CPG is Crisp’s Anti-Grooming Engine (AGE™). AGE™ takes a completely different approach to the traditional market approach for Parental Controls. It focuses on protecting children from external threats, rather than on moderating what children can and cannot do on the net. The net result provides an experience which is positive for both Parents and Children.

How do they do it?

Clearly, CPG type functionality could be installed in a domestic environment attached to DSL router to intercept traffiic but Crisp’s business model calls for them to work with ISP’s and embed their CPGs in the ISP’s network.

Although working with carriers / ISPs is always a challenge, it strikes me that Crisp’s technology would be well received as it provided a service that could prove to be very popular with parents. Crisp describe it thus:

“Offering controls within a network that an ISP can ‘switch on’ drives one of the key aspects of an ISP’s revenue model; value added services. With child protection being one of the few mass-market value added services this becomes particularly important in the current broadband ‘price war’.

Network-level protection also gives an ISP huge brand differentiation within the market; a consumer’s perception is immediately changed if their ISP is actually pro-actively protecting their family.”

The people:

Crisp is a young company and is actually Adam’s second start-up, even though he is only 21. Although he might be mightily embarrassed by this, when he was 18 he came number 4 in the RichList 2020 drawn up by the Royal Bank of Scotland. The Internet can be a horror sometimes!

Adam has put together a strong management team and has attracted Andrew Burke to be their Chairman. As Andrew was earlier BT’s CEO, BT Entertainment, I guess Adam has had many meetings with BT at Adastral Park!

Comment:

This certainly seems to be a company that has come with some cute technology and is taking an innovate, but challenging, approach to market. They are certainly addressing a major industry pain and if they are successful with thier trials I would expect a significant take up by the ISP community both small and large.

And, if you want to beat a path the Adam’s door, book your train seat because as they are based in Leeds in Northern England.

Postnote: Coincidentally, I saw that MySpace could be offering child monitoring software.

In January 2006, xBOX was launched – a firewall-based solution.
In February 2006, I came across newly launched anti-grooming software from In Loco Parentis


Sharedband: not enough bandwidth?

January 22, 2007

I noticed a discussion about Sharedband on Vecosys and I couldn’t resist finding out a little more about what they were up to, so I spoke with Keith Collins, who is their Sales and Marketing Director (you can find Keith on LinkedIn). Sharedband have recently started beta trials of their software with a couple of UK ISPs and plan for launch in the March timeframe.

What do they do?

The concept of inverse multiplexing goes back donkey’s years and was applied to traditional E1/T1 lines and to ATM. The term means that individual links are paralleled (or bonded) to multiply the available bandwidth according to the number of links paralleled. Sharedband have extended this concept in a “quite novel” way to the IP layer which enables ISPs to offer a DSL bonding service.

Sharedband is essentially a software company and they provide the tools necessary to manage the service. They license their software which is installed on the ISP’s routers and the ISP provides a firmware update for their customer’s routers.

Because the concept is IP based, it’s claimed to be low cost and an ISP can get the Sharedband service up in a day or so. It’s also simple to install at the customer end as well.

The company believes that there will always be a need to gain some additional bandwidth if it is obtainable at a reasonable price. They are probably right. There will always be individuals who want to increase whatever bandwidth they currently have, turning 1Mbit/s into 2 or more likely 8Mbit/s into 16Mbit/s. According to Keith they “want to look for companies that want to use something now rather than having to wait”.

They are initially focusing on the small business market (SME) and home workers with some money to spend as this seems to be an obvious market.

What is really interesting, is that because the software is network provider agnostic, it could be possible to obtain real provider resilience and even “bond cable and DSL combinations”. For example, the two bonded DSL lines could use different ISP providers although care would be needed to ensure that both are not using BT to provide the DSL connection or the phone line use the same BT street box. This is such an interesting application area!

As Sharedband works at the IP layer, bonding does not have to be limited to being installed on ISP DSLAMs. It could be installed on servers at a datacentre which, I imagine, could lead to network-based services such as Salesforce.com offering some interesting services that could improve performance and resilience directly to their customers.

Finishing off:

Their business model is that the ISP charges for the multiple DSL lines and then adds an additional bonding charge which is split between the ISP and Sharedband.

Sharedband are currently talking to 20 /30 ISPs and hopes to sign up this number within 12 months. One might also imagine, with their BT genesis, that we will see an announcement in this space as well?

They are currently in beta with two providers KeConnect and TeleComplete though I couldn’t find the doubler service on TeleComplete’s web site. You can see details of the service and pricing on the KeConnect site.

Some other vendors offering hardare based solutions:

There are other companies that provide a solution to bonding multiple DSL lines, the Australian ePipe is one example. The ML-IP access concentrator, is a hardware solution an enterprise can bond three DSL links into a multi-link tunnel without the involvement of the ISP.

 

FatPipe is another. “MPVPN enables bi-directional data transmission over multiple VPN paths, providing customers with the confidence that MPVPN will keep their VPNs “up” at all times regardless of router, ISP, line or backbone failures on one or two carriers. MPVPN allows companies to maximize the reliability, redundancy and speed of their VPN infrastructures while only requiring one VPN unit profile life.” Here is a short presentation on MPVPN.