If you've never been involved in the building of a website for a large company then this might come as a surprise.
Getting agreement for so many things would have taken many people months of time. "Lean Startup" where things don't work on day 1 is not an option. Include focus groups, armies of usability testing, all sorts of integration misery and this doesn't seem a crazy big number. They don't get 4 developers who'll work for 2 years for 12 hours a day for no money.
From what I can tell it looks like a fairly decent deal.
The site is snappy, visually impressive and with a well thought out user experience, tasteful upsell and in general clear communication.
At least it launched and managed to avoid all the pitfalls of corporate politics leading to bad architectural choices leading to a very expensive and poor implementation, which is a much more difficult challenge by itself than most people realize.
It's important to remember that:
* very few great developers/designers will work for Four Seasons on the cheap.
* it'll mostly have been well paid consultants earning $100/hr and up
* it is very likely that the project has had multiple iterations to reach it's current state, all hidden behind a single big release.
Seriously. I'm all for being compliant for most users, but accommodating for users who actively gimp their browser is a little ridiculous when they probably represent less than 5% of their users. You have to expect that some sites just aren't going to work.
There are people who deactivate JS, for personal/corporate/whatever reasons. Shouldn't one serve these customers as well? Or does this feature cost another $1.5M ?
Granted, there are people who deactivate JS for whatever reason.
The intersection of "people who deactivate JS" and "The addressable market of Four Seasons Hotels" is likely small enough that the lost potential sales wouldn't even cover the cost of development.
I just moved here, so I'm getting to know the neighbours. One of them had a wi-fi problem. It turns out his TP-Link admin interface relies on JavaScript to function. He uses the No-Script Firefox extension. He's a retired chairman of a PVC company, who moved to the Canary Islands for his sabbatical. Judging by the house, he's part of the addressable market.
I think it makes sense that high-value employees with access to sensitive information get basic cybersecurity education. Therefore, some of them block JavaScript.
Catching deactivated JS is not really big, in terms of development costs. But especially companies like "Four Seasons" are usually the ones that live by "Every missed customer is one too many".
And i dont think that money was an issue in this project.
Except that the cost is near zero if you approach it right. It also means supporting many mobile browsers that usually break js and providing a better experience on slower connections (3g, edge, whatever).
There's a specific touch optimized site that works for mobile. It doesn't look like it or the iPad optimized site are done yet either since they redirect to preview.fourseasons.com, but they're definitely aware of mobile devices.
In general web developers seem to kowtow to the most to ridiculous compatibility requirements--how many still fret about IE6 and IE7 compatibility?
In no other industry do consumers get that kind of luxury. It'd be like the music industry putting new releases on cassette tapes for the people who haven't upgraded to CD or MP3 yet.
Tell that to the corporate IT people who manage thousands of workstations and support internal intranets that were designed around those older browsers.
They can very well keep ie6 for cruftmaster97(tm) but they could provide chrome or firefox to browse the 2012 internet. I know some people call their browser "the internets" but it is in fact just a regular application that you can install as many of as you like. Kinda like if you said we can't use Latex for our yearly report even though you could create all diagrams and tables directly out of the database because we already have a publishing software installed. Word.
Wait, that is exactly what IT would say.
I can't say WHO, but I've been leading the UX on a large newspaper site. There's huge feedback loops, tens of people to sign off per page and a ton of usability testing.
They have their own in-house development team. I can bet you we've not come anywhere close to that and the site is absolutely HUGE.
I suspect somewhere someone was given this as a budget and they had to spend it, or lose it.
The in-house dev team is the secret here. You can pay each person $60k/year and they have management who will say "No, that's not possible with our resources".
Compared to the Four Seasons outsource model where anyone can say "I want that feature" and will be told "Sure, it'll cost an extra $1m of someone else's money".
Most of the costs probably came from expensive consultants/contractors during the analysis phase. I'm going to take a punt and say that Oracle is probably involved somewhere down the chain. Whether it be Weblogic or OracleDB, I bet they had a whole load of consultants to come in long before development started
You forgot various change request, often contradictory by different departments, countless changes to things already signed, specified, implemented, tested and deployed, at least three starts from scratch because of different "visions" ...
20 devs/DBA/IT for 2 years would cost you something like 4M, 8M at max(Super Salaries for everybody) !
The costly part is the "Management" which creates the bureaucracy so other people can hang out in the project controlling excel tables while those 20 bastards work their asses off.
I don't think that includes consultants that helped the design, user experience, and architecture. Since they went with a pretty well known interactive agency, it probably cost them more than three million.
Kinda off-topic, from the article: "The 2012 Four Seasons Luxury Trends Report from Four Seasons reveals that 71% of its customers bring a smartphone with them on their travels and 61% bring a tablet device."
That's a lot. And even so, my only two experiences at luxury hotels meant no wifi available in the room. And daily paid ethernet. I understand being there meant me or my company have money but I'd like to stop being ripped off after paying so much to stay there.
The article talks a lot about the new site being an effort to increase their online bookings. I can see that for a luxury brand hotel, online bookings might be lower than for say holiday inn express. Part of what you're paying for is personal attention, so the people who stay there might want that to extend to the booking experience.
How do you overcome that in an online booking? I don't think they did, but how would you approach it?
The best way to kick the hornets' nest over at HN is to mention lots of money exchanging hands (i.e. color.com). Nothing brings out the self-righteous geek in us than hearing about other developers getting paid.
Right. Developers getting paid. You hear that sound? Listen real carefully. It's the sound of our profession being flushed down a toilet because your buddies are cheating businesses out of their money by charging too much for too little.
You say that, but consider what it takes to build a web site for an organisation like Four Seasons.
You’d probably spend 6 months alone on producing the various specifications and plans, and deal with governance (reviews and revisions, sign-off, acceptance and so on). You’d build out a proper information architecture (a little more than the average SharePoint TeamSite taxonomy).
To show you how realistic that number is, consider the scale of documentation Four Seasons would require you produce. Those docs would probably include, but not be limited to:
Project/team structure
Risk assessment
Vision/Scope
Business requirements
User requirements
System requirements
Operational requirements
Usage scenarios
Conceptual design
Logical design
Physical design
Availability plan
Backup and recovery plan
Budget plan
Capacity plan
Communications plan
Deployment plan
Development plan
End user support plan
Project plan
Project schedule
Pilot plan
Purchasing and facilities plan
Security plan
Support plan
Test plan
Training plan
And after all that you haven’t even written a line of code (other than the inevitable throw-away proofs of concept and demonstrators).
Imagine having to estimate the effort required to produce that set. The analysis alone will take a herd of business analysts months.
Like Steve Sinofsky said – new businesses (start-ups), or even new markets, simply cannot solve the class of problems that is addressed by larger and more established firms.
[Edit] That list up there excludes systems integration (hotel reservations, customer care, etc) and data migration - both of which can add orders of magnitude to a project.
Well, it's probably more like the company charging themselves too much by having a ridiculously mismanaged process for getting the work done.
Incoherent, produced-by-committee specs that don't actually have any explanation about how the soup of buttons and buzzwords is supposed to work, ridiculous management overhead.. it's probably not even the developers getting paid that much. Nobody walked off with that money, it was pissed away in 6-hour meetings.
To be frank, I always thought that our profession was being flushed down a toilet because some guys were cheating themselves by charging too little for too much. Go figure.
You've both made the classic fallacy of assuming that writing some code is all there is to do. For example, perhaps a lot of that money went on paying a photographer and his entourage to fly around the world.
We're talking about Four Seasons here. If this project cost $1m instead of $18m, I'm pretty confident that the other $17m would not be going into our profession.
Unless your metaphor is more about the legitimacy of our profession. If that is the case, I'd point out that we have no inside knowledge on the situation, and then I'd use Four Seasons as a counterpoint -- they charge crazy rates for a fancy hotel room, yet they're not destroying the hotel industry.
Right. And I suppose Mercedes is flushing the automobile manufacturing profession down the toilet because they're cheating customers by charging too much for too little as well?
If the site cost 18M, it does not mean all 18M went to developers (total probably includes other planning & logistics costs). I say "if" because the article from the econsultancy link provided in comments here states 18M without any source to confirm that figure.
Lets assume they received a custom CMS with the website, some custom frontend-plugins, development of flash entities, interfaces to in-house IT-systems, a tracking suite, hosting, QA, send some photographers around the world to get high quality pictures, consulting....Still i cant figure out HOW this can cost $18M (and i am a technical consultant in digital agency, working exclusively for enterprise customers).
I also do that for a living... and they got stolen or their management team sucks badly. With 18M you can build 18 of that site.
For some minutes I thought 18M was 18k(which would be too low)... then I got it, and got even more surprised.
The only explanation I can think of is that they didn't had a IT department and had to Hire and buy every single person/thing they needed including hosting their own site on their own servers and buying their own Backbone :D
Judging by the URL structure, this is built on Adobe (formerly Day)'s Communique 5.x platform.
The CQ Platform gives you a ton of these graphical widgets and social interaction out of the box, though there is still plenty of development to write a full reservation system. It also handle mobile rendering, site templating, and multiple languages quite well.
So while it sounds like people are seeing "a single website for $18m", I'd be willing to wager that was the cost for the entire content platform (CMS/Social/Mailings/Reservation system. So now they would have access to very advanced templating design, multi-lingual components.
The amount of time it would take them to roll out multiple country and language sites would be dwarfed in comparison to any other platform. That $18m may be a big upfront costs, but might well pay for itself if they intend to use for other properties and across multiple regions/languages.
Definitely CQ - pretty impressive in that regard. Having worked with the platform, I can tell you -good- CQ5 developers are very expensive and hard to acquire. Conde Nast is learning this the hard way after they committed to converting all their properties over to CQ.
If the website redesign results in 1% more direct conversions on their website (versus discount "Hotels.com" outfits), then $18M is probably a fairly decent "investment".
Perhaps the developers were smart and made a value proposition rather than a pure hourly / salary proposition. Others on HN have talked about this distinction extensively, so try to avoid that little mental roadblock when considering the amount that they paid for the redesign.
I wonder if photography was rolled into that cost? I'm sure some shots are reused and rights-managed stock, but for those that aren't, the costs of planning, lighting, rigging, stylists, talent, etc. would be significant. And it looks as though each location has 30+ photographs: http://www.preview.fourseasons.com/find_a_hotel_or_resort/
I'm surprised that no one yet said anything about the 18M being a fake number. These guys do this all the time. They spend 10M and release to the press that it was 20M.
I work(ed) for more than one company that uses this strategy. The whole idea is to get everyone to think that you are bigger than you really are.
Of course one can spend 18M on a bad project, but you guys covered it already...
This is exactly it. Not to mention all of the "internal" resources/time they put into that number. I also imagine they did an overhaul of their internal booking system, and possibly increased full-time staff (project managers, support, and sales staff) to handle the new demand.
So? It's a pack of ~90 sites + the main one, content management systems that will probably be used by hundreds, photography.. the cost is not surprising at all.
They could do a better job of graceful degradation for javascript disabled without much effort, but I wouldn't blame them - everyone here seems to support the opposing mindset too.
It's weird, some elements of the design look to be 5-6 years old in style, while other are clearly inspired by Microsoft's Metro UI (which is good).
The small font size and traditional 'luxury hotel' look of the header/footer really ruin it for me (example look how the tagline in the tiny logo is almost unreadable). Where they succeed (mostly) is in the page content layout and design.
Also, shrinking my browser window doesn't dynamically resize or shift the layout so -1 for unresponsiveness. I was on a sub page of the Boston location which may not have that flexibility but for 18M and 2012 ... it should.
"Also, shrinking my browser window doesn't dynamically resize or shift the layout so -1 for unresponsiveness. I was on a sub page of the Boston location which may not have that flexibility but for 18M and 2012 ... it should."
Responsive design also limits design/ux decisions especially on their content pages which have a non-linear grid system and shouldn't be universal. I'm not for or against responsive web design but I think it should be contextual. I haven't seen many sites that have won me over.
Is it just me or is the text at the bottom of each picture (with the location, temp, etc) unreadable in at least half the locations? The white on white is a bad idea.
This must be a "marketing" driven site where "good ideas" trump usability. I'm sure they got an award or three for this cutting edge design.
I've been working on an online reservation CMS for more than 2 years now and this seems quite a good result.
Maybe I would have done some things differently, such as location selection: the drop down is really long, the could have used an autosuggest. But everybody likes different things so I can only assume their usability tests suggested otherwise.
The price is not really something we can judge: it could be either fair or too much.
Who knows, this may be a very flexible framework they'll use for the next X years (good) or a hard to maintain website which will cost them even more to keep alive.
We can't really judge.
I wish I had 18 Millions to develop mine.
Or maybe not, I'm happy with scarce resources, so little goes wasted.
The website itself is just be the tip of the iceberg - it would need to talk to multiple back-end systems. $18m wouldn't be unreasonable if the project required complex legacy integration and/or re-design of underlying infrastructure.
You write a spec sheet, bid low. You want to win the bid. Say 6 mil, maybe even 5. Then half way through the project, say you need another 3 mil to finish, and without that you can't finish it. Sunk costs? don't worry, the flip rate on staff means the sunk costs were the last guys problem.
Continue asking for more until you can't squeeze any more juice out of the orange that is your client. Congratulations, you just won at corporate politics. Ship whatever you've done and say it's the best you could do. Done.
I'd like if someone could please explain why NHS don't negotiate a job price (and maybe even a 20% contingency fund) and then pay that price for the job to be done. 50% up front, waymarked payment points and a refund for failure to complete.
Given the budgets here it seems we should have a MP as minister for NHS IT systems.
I think for $18m they're probably making significantly more than that back. That really is a damned fine website, and you can tell that every piece of it has been double checked and agonized over with the client.
Is there an older story on the internet than a software project overruns budget or doesn't get finished? I'd be more interested to see something like, software project costs $18M and here's where we went wrong, or software project costs $18M and here's why the client got their money's worth.
I suppose only the client can decide wether that's money well spent or not. And while I think the price is high, it's certainly not the worst hotel website I've seen, and is probably one of the better ones.
The site is awesome. The UX is spot on! I think people have seen the price tag and expected something unimaginable. Not quite sure why they chose to use an Adobe application for the cms side of things.
I think it looks great. And, it's in my interest for websites to cost as much as possible, so I have no problem with this.
I don't understand that thing nerds do where they brag about how they could have done it in a weekend for $15 and some Mountain Dew. Well, I do understand it ("look how smart I am"), but it's still obnoxious.
How? I've just started experimenting with web development and I can't figure out how it could possibly cost that much. Maybe there is a whole world I haven't seen yet, but if they had used the latest open source libraries/frameworks that do half the work for you, would it have cost that much?
I work in a digital agency, similar to the one who did this. 80% of the budget is spent in the requirements phase. More than a year doing designs, sending 20 people down to see the client every trip, business requirements gathering, market analysis, and such.
$18M over 24 months is only $750K per month burn rate. At the typical agency rate of $150/hr, this is 35 people working on this project. Only 2 of them are developers. :)
