Stupid Guest Tricks

The user experience on the DIsney website is horrible!

There. I've said it. I feel better now. Maybe. A little.

I've been planning a trip to the World trying to make the most of this whole Diseney Experience. Mind you, we've always enjoyed planning our trips, end even though picking where you'll eat months in advance can be a challenge, we enjoy the discussions about where to make those reservations. Planning is a way to inject a little Disney magic into our every day lives, and heightens the anticipation for the trip. We've even used the online dining reservations for the past few trips, and although it was often slow and a little frustrating, we've never reached the heights of this trip.

This time we were offered the bands, plus we're meeting some good friends down there and touring together. That, and the website has been upgraded (downgraded with more features???? ) have created some real challenges.

First, everything is so very painfully slow, and there is no good reason for that. When I've mentioned it in passing on calls to DVC Member Services they've said it was the number of people doing the same thind as me. I assume this is what they've been instructed to say. I don't debate it with them since they aren't creating the message, but that's just wrong. Disney has had a web presence for many years now, and they have incredible volumetrics on reservations of any and all kinds. That excuse only really works for new websites, or for occasional super spikes created by outlier offers. Not for an existing site with known traffic patterns every time you log on.

Also, just the time it takes to retrieve reservation information after the first part of the page has rendered. We're talking less than 2K of data on a handful of business objects. Any moderately well-designed database can pull that in microseconds.

Don't even get me started on the forms that pop into a browser window and only collect a couple of pieces of information at a time. Sure, your mobile app would want to do it that way, but a website in a full broswer? Just ask me what you want and be done with it! Let's not go through 4 panels to collect 7 pieces of information.

Now, the above complaints can be made about many websites these days. It seems like that is the design trend. To me, it seems like UI's are moving in the direction of a children's book. Bright and colorful with sparse information. I like an aesthetic presentation as much as the next person, but can't we at least get a little sophistication and maybe a 10th grade level?

The biggest complaint I have that is all on Disney relates to coordinated planning. Disney has told us that the parks are places for magical gatherings. Just try and plan one of those. The greatest failing for group planning in the new technology is for Fastpass+. The only way to do it is for everyone to have their own Disney account. We originally had one account with me as the primary and my wife as a member of our travelling party. Our friends were set up the same way. In our account, we had linked both our passes. Again, our friends did the same thing. We then linked the two accounts -- ours and our firends. If I tried to get Fastpasses, I could get them for myself, my wife, and my friend. Even though his wife was a member of his travelling party and had her own media, she was not exposed through the linkage. Same thing was true if they tried to get Fastpasses for the group. Eventually, we had to all get accounts (requiring different emails since they are used as the user ID) and link them all together. I will say, I was able to go into my wife's profile on my account and link it to my wife's new account, and the system sorted out the passes and connected everything up just fine. Still, we would rather not have individual accounts.

I could go on, but I will stop here and hope the catharsis is good enough to get me through the rest of my planning. It would ruin all our plans if I ended up hospitalized for a nervous breakdown.

My wife planned our upcoming trip, and the kids were part of her traveling party and she just had to link my account.

I think the challenge is probably on the backend for Disney, in the DB. I suspect they paid for Oracle, don't have enough hardware on the front and backend, and the DB is not optimized sufficiently. If it is truly a billion dollar system, then it is highly complex that is touching many systems and once, and keep in mind, it is still "beta" at this point.

I'm also moving this thread to the break room.

Some companys can not understand the acronym,K.I.S.S.
Where as Fry yells,"Shut up and take my money",
Disney just tells you,"Shut up and give us your money". "You'll have a great time,trust us".

ktulu wrote:My wife planned our upcoming trip, and the kids were part of her traveling party and she just had to link my account.

But you would not have been able to do it since you wouldn't have been able to select the kids through the linkage. Other members of the travelling party are visible to the account holder, but not to linked accounts.

I think the challenge is probably on the backend for Disney, in the DB. I suspect they paid for Oracle, don't have enough hardware on the front and backend, and the DB is not optimized sufficiently. If it is truly a billion dollar system, then it is highly complex that is touching many systems and once, and keep in mind, it is still "beta" at this point.

I agree completely, but it still isn't an excuse. I make my living architecting large scale IT systems, inlcuding ones with web UI's and significant backends. I have also been involved with optimizing DB activity after a launch, and things like retrieving reservation information has been in place for a number of months now -- pleanty of time to tune, and to scale to hardware.

I'm also moving this thread to the break room.

I debated between the two.

ktulu wrote:I think the challenge is probably on the backend for Disney, in the DB. I suspect they paid for Oracle, don't have enough hardware on the front and backend, and the DB is not optimized sufficiently. If it is truly a billion dollar system, then it is highly complex that is touching many systems and once, and keep in mind, it is still "beta" at this point.

It is still in beta testing. The Company has been quite clear on that point, and that we should all (guest and CM alike) expect continuing changes and development.

At least it's working better than healtcare.gov!

I got to try the system out for a trip the first week of November. Overall, I do like what Disney is doing here; however, I do have to agree that the UI is a little frustrating in that it takes far too many screens and steps to create a Fastpass+ reservation. The biggest issue I had is that it didn't seem to be possible to cancel a single Fastpass+ and create a new one at a different time and different attraction. My choices seemed to be:

1. Change the time only and keep the attraction the same.
2. Change the attraction and keep the time the same.
3. Junk all my Fastpass+'s for the day and start over.

However, the concept overall worked very well for us. We were able to get Fastpasses for parks late in the day when we were planning to park hop or on our first day after arriving at Disney. This never would have been possible before. We were even able to change Fastpass times when we realized we weren't going to get to the park in time to use that 9am pass.

Hopefully Disney's IT cast will get the issues sorted out prior to my next trip :)

WEDFan wrote:But you would not have been able to do it since you wouldn't have been able to select the kids through the linkage. Other members of the travelling party are visible to the account holder, but not to linked accounts.

I think she had to share them or something. I don't recall, but I see them under "My Family" right now.

I agree completely, but it still isn't an excuse. I make my living architecting large scale IT systems, inlcuding ones with web UI's and significant backends. I have also been involved with optimizing DB activity after a launch, and things like retrieving reservation information has been in place for a number of months now -- pleanty of time to tune, and to scale to hardware.

True, but we don't know what their production stuff looks like compared to their test environment. Rolling out a significant patch could be complicated, so they are waiting to cut over a major release. And their stuff could also be so screwed up, that they had to re-code significant parts. I'm sure they are also being overly cautious with this new system. I could only imagine what would happen if they suffered a huge crash. You're talking everything down, room reservations, dining, FP+, admission to the parks, purchasing capabilities, not to mention tracking everyone!

ktulu wrote:I agree completely, but it still isn't an excuse. I make my living architecting large scale IT systems, inlcuding ones with web UI's and significant backends. I have also been involved with optimizing DB activity after a launch, and things like retrieving reservation information has been in place for a number of months now -- pleanty of time to tune, and to scale to hardware.

True, but we don't know what their production stuff looks like compared to their test environment. Rolling out a significant patch could be complicated, so they are waiting to cut over a major release. And their stuff could also be so screwed up, that they had to re-code significant parts. I'm sure they are also being overly cautious with this new system. I could only imagine what would happen if they suffered a huge crash. You're talking everything down, room reservations, dining, FP+, admission to the parks, purchasing capabilities, not to mention tracking everyone![/quote]

FWIW, they have experienced a number of "huge crashes", usually in one module at a time, but still devastating for those needing that module. Some of these outages have lasted several days, a few have lasted weeks.

I expect the difference between test and production environments is a major issue. In particular, I doubt they have a way to simulate the huge number of simultaneous service calls received by the system. Things were much more stable back during Alpha testing when there was only one hotel and no online access. Like many big systems, it has not scaled up as graciously as one might hope.

All that said, I do see steady forward progress. It's just such a massive system (and has to interface with so many other, fundamentally incompatible systems) that it will take much time and patience to get it to where it needs to be.

EpcotFan wrote:I got to try the system out for a trip the first week of November. Overall, I do like what Disney is doing here; however, I do have to agree that the UI is a little frustrating in that it takes far too many screens and steps to create a Fastpass+ reservation. The biggest issue I had is that it didn't seem to be possible to cancel a single Fastpass+ and create a new one at a different time and different attraction. My choices seemed to be:

I believe that has changed very recently (within the past 2 weeks, if I remember correctly), and you can now cancel just one at a time.

As to the complexity of the system, I've heard (unconfirmed) that Disney has over 25,000 different types of ticket media ... Wouldn't surprise me ... just for the Magic Your Way tickets, there are 160 possible tickets that could be sold. Once you add in all the other ticket types, I can see how 25,000+ would be plausible.

The system handles a staggering amount of data. That being said, I don't know that it handles it particularly well, and I'll agree, the guest-facing user interfaces leave something to be desired. (For that matter, so do some of the cast-facing interfaces)

Zazu wrote:I expect the difference between test and production environments is a major issue. In particular, I doubt they have a way to simulate the huge number of simultaneous service calls received by the system. Things were much more stable back during Alpha testing when there was only one hotel and no online access. Like many big systems, it has not scaled up as graciously as one might hope.

Typically with large scale systems you will have several non-production environments (e.g. Integration Testing, User Testing, Regression Testing, Performance Testing.) At least one of those environments should be configured the same as your production environment. Within that environment you can simulate production conditions. There are a number of commercial off-the-shelf (COTS) tools available to simulate production load. The tools are expensive, but in a 100's of millions of dollars project on critical systems, the investment is well worth it (and really only a nit in the overall Bill of Materials for the project.)

On the web experience, with the various kinds of reservations, Disney would have excellent metrics on transaction arrival rates, duration, concurrancy, etc., making it very easy to configure load profiles. Now, the whole FastPass+ portion is new and it would be easy to misjudge the metrics for those areas, but the performance of the standard reservations portions would be well known, predictable and testable.

On the operations end (Front Desks, restaurant kiosks, gates, etc.), those metrics would also be well known and easily modelled. For operations, even FastPass should be well known.

There are a lot of major system implementations that have blown up, particularly in the public sector, but there are also a lot that have gone well. My expectations were that Disney would be among the better ones and it doesn't seem like they made it.

hhsrat wrote: As to the complexity of the system, I've heard (unconfirmed) that Disney has over 25,000 different types of ticket media ... Wouldn't surprise me ... just for the Magic Your Way tickets, there are 160 possible tickets that could be sold. Once you add in all the other ticket types, I can see how 25,000+ would be plausible.

The system handles a staggering amount of data. That being said, I don't know that it handles it particularly well, and I'll agree, the guest-facing user interfaces leave something to be desired. (For that matter, so do some of the cast-facing interfaces)

Permutations can add up quickly, but the attributes should be pretty manageable: ticket class or type, issue date, activation date, expiration date, expiring, park hopping, number of "fun" add ons (initial and remaining), redeemed days and parks (for older media), etc. In addition to the attributes which are needed to determine if entry is allowed, you would need the demographic and possibly biometric information.

The real huge amount of data comes in the transactional end of things. One record per park entry, for example. Even that, though, is not unprecedented in the business world. Consider sales for WalMart, NYSE trades, logistic moves for the military, health care interations for Aetna, etc.

Zazu wrote:All that said, I do see steady forward progress. It's just such a massive system (and has to interface with so many other, fundamentally incompatible systems) that it will take much time and patience to get it to where it needs to be.

I'll be glad for all the progress they make!

Integration is usually the real challenge. Every integration point is a possible point of failure. What's often referred to as "sad path" planning becomes critial along with integrated testing.

In all this, I would say, though, that the DVC Members website seems to be a lot better, so it isn't all gloom.

Like all things, it gets frustrating when the experience falls into an area you are familiar with (or make your living at,) and you know it could be done better.