Data Mining Disney -- A Magical Experience

[Pemeberly]

While I fully agree that Disney wants every bit of money you have (and don't have) it's next to impossible to derive analytics about why people didn't buy something unless you have direct interaction with them.

True, at the individual level. But data mining usually isn't about the individual. It's about taking data from a group of individuals to find the pattern.

 

[rteetz]

I still don't think that this is the "real" reason that Magic Bands came into existence. That's way too simplistic. A contributing factor? Sure. He just misses out on the moat it provides and the ease of use side of things. I think those two factors are bigger than data mining.

Also... He called Walt Disney World "The Happiest Place on Earth." It is not the happiest place on earth... That's Disneyland. Walt Disney World is "The Place Dreams Come True." Pet peeve of mine.

We share a similar pet peeve. Walt Disney world also has the name of most magical place on earth.

 

[rteetz]

True, at the individual level. But data mining usually isn't about the individual. It's about taking data from a group of individuals to find the pattern.

Exactly one of the advantages of this. I think this will help Disney in certain aspects. I also think similar systems will come into place eventually at places like universal.

 

[Yellowstonetim]

What most people don't know is with your credit and debit cards, phones, cams, facial ID, information brought by many information collections groups, all this data is already available on the markets to anyone or company. I lot ore then anything the MB is collect. As a post poster said............Disney is just catching up to what companies like Amazon and others already are colleting.

What stuffed Disney character or items you buy in the shops is really very minor and a non issue.

AKK

You couldn't be MORE wrong. No this is NOTHING like what Amazon, or the local grocery store, or any other retailer does. It would be equivalent if these places had all your demographics and tracked you walking through the store, knew how many wer in your group and their ages, and knew how long you stopped and at what products.

You have to be kidding!? Knowing what you buy a non-issue?! Grocery stores will place a stand with junk widgets in a pink display right in the middle of an isle specifically because X% more widgets are sold and Y% more when it is pink instead of blue. Important stuff! Grocery stores offer you that discount card so they can track your spending. They would LOVE to know everything about you and track every movement if they could figure out a way.

Data mining isn't the only reason Disney did this. It is only 98% of the reason. You have to understand what they can do with this. Stretch your imagination a bit:

Not only do they know a person went into a store, spent 12 minutes, and didn't use a credit card (In the cash example), but they will know exactly how many people spent cash, how long that group spent in the store, what their age, home place, and party size were. They will know if this party had a credit card and just chose not to use it. They will know if this trip in the store the entire party was there or if no kids were there. They will know how many days, what parks, how long this group spent in the park, whether or not they stayed on property, how often they visit, etc., etc., etc. AND they will know how long they were in the park before they visited the store. AND they will know what the weather was that day to know at what temperature people go into the store and how that affects spending both cash and credit card!

They will know if they move a planter where people sit 30 feet from its current location outside the store, how that will affect spending. They will know if older patrons spend less because they sit further from the store.

They will know which pavilions in World Showcase take the longest to walk by for people from certain countries and certain ages. They will know how a 55/51 year old couple move and spend money when by ourselves and how we move and spend money when the kids are with us. And how we change travel patterns when it rains and is warm versus when it rains and is cold. They will know how much people that attach a credit card to the band spend on the credit card and how much those that don't attach it to the band spend on the credit card. And their ages, stay length and home state/country. They will know what line lengths we will tolerate and what we won't and how much weather affected this and which rides. AND they will know in each scenario whether or not and how this affected our spending!

I could sit here and list at least 100 different things Disney can track and cross reference with each or all of the other items. There is an endless amount of information that is extremely, extremely, EXTREMELY, valuable. Human beings are fascinating creatures. While they have free agency and make individual decisions, they also have relatively predictable response to stimuli if you can study the effect. Changing the color and design of a store exterior may affect the money spent. This is something measurable. Maybe it makes a difference to 50 to 60 year olds but not 20 to 30 year olds.

There are two levels to this value that will make Disney money:

1. They will know the impact every change has on every demographic group. They will be able to make changes that cause spending changes. For example: Placing hats in the window of a particular store attracted a particular age group without children and resulted in 8% more hats sold, but placing Frozen merchandise in the window made a higher percentage of younger families stop in the store and sold 12% more Frozen merchandise at twice the profit of the hats. OK, change the window display back to Frozen. Oh and this is true in the summer but not the winter so change it back in the winter.

2. They will know how changes affect our habits and will know how to improve our experience to keep us happy. This is a more indirect relationship to their profits but still very important. This might mean they find that the placement of walkways and planters makes people go left instead of right and slow down in a crowd, stop, and neither shop or ride at one particular point. They rearrange the walkway and find people move differently and whether or not it is an improvement. AND they will know the demographics of the groups whose behavior changed and how much and in what manner each groups behavior changed.

This cannot be overstated. IT IS PURE GOLD! No one has every been able to track this number of peoples movements and actions in a theme park before and tie it back to specific data about those customers.


P.S. Disney has given up on the official Happiest Place on Earth and The Place Dreams Come True because the public won't follow it. It is still official but they no longer stick to it. For some time Disney has been calling WDW the happiest place on earth in various venues because 98% of the people do. And it wouldn't do for Disney to run around correcting everyone, "No! It isn't the happiest place on earth! No, it isn't the happiest place on earth!"

 

[wallrock]

You couldn't be MORE wrong. No this is NOTHING like what Amazon, or the local grocery store, or any other retailer does. It would be equivalent if these places had all your demographics and tracked you walking through the store, knew how many wer in your group and their ages, and knew how long you stopped and at what products.

You have to be kidding!? Knowing what you buy a non-issue?! Grocery stores will place a stand with junk widgets in a pink display right in the middle of an isle specifically because X% more widgets are sold and Y% more when it is pink instead of blue. Important stuff! Grocery stores offer you that discount card so they can track your spending. They would LOVE to know everything about you and track every movement if they could figure out a way.

Data mining isn't the only reason Disney did this. It is only 98% of the reason. You have to understand what they can do with this. Stretch your imagination a bit:

Not only do they know a person went into a store, spent 12 minutes, and didn't use a credit card (In the cash example), but they will know exactly how many people spent cash, how long that group spent in the store, what their age, home place, and party size were. They will know if this party had a credit card and just chose not to use it. They will know if this trip in the store the entire party was there or if no kids were there. They will know how many days, what parks, how long this group spent in the park, whether or not they stayed on property, how often they visit, etc., etc., etc. AND they will know how long they were in the park before they visited the store. AND they will know what the weather was that day to know at what temperature people go into the store and how that affects spending both cash and credit card!

They will know if they move a planter where people sit 30 feet from its current location outside the store, how that will affect spending. They will know if older patrons spend less because they sit further from the store.

They will know which pavilions in World Showcase take the longest to walk by for people from certain countries and certain ages. They will know how a 55/51 year old couple move and spend money when by ourselves and how we move and spend money when the kids are with us. And how we change travel patterns when it rains and is warm versus when it rains and is cold. They will know how much people that attach a credit card to the band spend on the credit card and how much those that don't attach it to the band spend on the credit card. And their ages, stay length and home state/country. They will know what line lengths we will tolerate and what we won't and how much weather affected this and which rides. AND they will know in each scenario whether or not and how this affected our spending!

I could sit here and list at least 100 different things Disney can track and cross reference with each or all of the other items. There is an endless amount of information that is extremely, extremely, EXTREMELY, valuable. Human beings are fascinating creatures. While they have free agency and make individual decisions, they also have relatively predictable response to stimuli if you can study the effect. Changing the color and design of a store exterior may affect the money spent. This is something measurable. Maybe it makes a difference to 50 to 60 year olds but not 20 to 30 year olds.

There are two levels to this value that will make Disney money:

1. They will know the impact every change has on every demographic group. They will be able to make changes that cause spending changes. For example: Placing hats in the window of a particular store attracted a particular age group without children and resulted in 8% more hats sold, but placing Frozen merchandise in the window made a higher percentage of younger families stop in the store and sold 12% more Frozen merchandise at twice the profit of the hats. OK, change the window display back to Frozen. Oh and this is true in the summer but not the winter so change it back in the winter.

2. They will know how changes affect our habits and will know how to improve our experience to keep us happy. This is a more indirect relationship to their profits but still very important. This might mean they find that the placement of walkways and planters makes people go left instead of right and slow down in a crowd, stop, and neither shop or ride at one particular point. They rearrange the walkway and find people move differently and whether or not it is an improvement. AND they will know the demographics of the groups whose behavior changed and how much and in what manner each groups behavior changed.

This cannot be overstated. IT IS PURE GOLD! No one has every been able to track this number of peoples movements and actions in a theme park before and tie it back to specific data about those customers.


P.S. Disney has given up on the official Happiest Place on Earth and The Place Dreams Come True because the public won't follow it. It is still official but they no longer stick to it. For some time Disney has been calling WDW the happiest place on earth in various venues because 98% of the people do. And it wouldn't do for Disney to run around correcting everyone, "No! It isn't the happiest place on earth! No, it isn't the happiest place on earth!"

Sorry, a good portion of what you’re saying depends on input and feedback from the customer. The magic band isn’t that ‘magic’. Other items you state are either pure conjecture or something that can already be done without tracking your every move. Specifically, knowing how many people use cash vs credit cards is something that companies have known for fifty-plus years. Nothing new here. Disney has been in possession of much of this information ever since people tied their credit cards to their room key. They’ve been optimizing store locations and product location within those stores for decades.

The personal data you speak of is voluntary, and not everyone is willing to give all or some of it up. This cannot be understated.

Your travel patterns outside of Disney will be completely unknown to them, so data for offsite guests will be limited at best.

As I’ve stated before, none of this matters if you visit once and are done. Everyone’s reaction at a first time visit is going to be different.

And how exactly will they know your threshold for tolerance when it comes to an attraction line? Just because you choose to not get into a line because it’s a 30 minute wait doesn’t mean that you’re impatient. You could feel tired, have something else on your mind to do or suddenly have a phone call to take. Tracking may say that it takes someone X number of minutes to walk from A to B, but that walk may take half that time or twice as much by someone else who shares similar personal characteristics (again, only information that you’ve provided). Disney doesn’t know why one person covered the distance quickly while the other did it slowly, especially if the difference can’t be tied back into time needed to make a purchase. Bottom line is Disney can know the where and the time consumed but they won’t know the why. Magic Bands don’t read minds.

To your point humans are rather unique but one thing that no one has ever been able to do is determine what someone will do in the next five minutes regardless of how much information you have collected. The only pattern they will see is that the pattern is constantly changing. The biggest benefit they have is studying the complex system of a crowded park so they can manage the system as a whole and staff accordingly. Without a doubt there will be some benefits, but nowhere near the complexity you indicate.

 

[DDLand]

@Yellowstonetim Good post above. I agree with much that you said. I think the potential here is huge. The subtle changes can be the difference between millions of dollars of increased revenue. The possibilities (as you pointed out) are nearly endless. Especially in this theme park setting it present different circumstances. With a brand people inherently trust, Disney is in a unique place where they can push the limits. They're positioned where they can get away with a lot of things because everything is closed and proprietary. It's like miniaturized world where all the retailers and restaurants are owned by one company. They can release a new technology standard and get 50% marketshare in a matter of years. This is in contrast to, for example, mobile payments that are still struggling to get off the ground. Disney has so much room to play in the physical world, smart phone interfacing, and also with longterm data collection.


With that said, I still feel like we should temper expectations slightly.

Data mining isn't the only reason Disney did this. It is only 98% of the reason.

I want to believe that Disney already has this infrastructure in place or is quickly developing it, and that Disney is about to start exploiting this. I just am not sure they are.

Let's think about this from Disney's standpoint, you've just spent a billion on a new project, and you want to prove that it's worth it for shareholders and for guests. Wouldn't an awesome way of proving that be by taking about improving the longterm customer experience (and by boosting sales)? Data mining, in a more PR friendly description, would be a perfect way to brag. Those Fast Company and Wired articles should've been loaded to the brim full of the benefits of mass data collection. As far as I can remember there was basically no mention to this in either of them.

Any use of the long range scanner has been primarily for very specific purposes such as the photopass or Be our Guest. We don't know where the long range scanners are currently deployed, and we certainly don't know if they're being used for the purposes mentioned above. That's not to say there isn't potential, there is. I'm just not ready to say Disney is fully utilizing the potential. In fact, if the data collection was 98% of the reason Disney launched MyMagic, it could be deemed a failure of massive proportions.

More likely reasons Disney created MyMagic+
1) Allows for preemptive bookings, and puts ([for the sake of being non controversial] the majority of) guests in the drivers seat. They no longer have to criss cross the Magic Kingdom 20 times a day if they don't want to.
2) Keeps guests on property. By allowing guests to book farther in advance, guests no longer feel that they can drop out for a day when there are dinner and FP reservations waiting. Also, due to the seamlessness of purchasing and using other functions of MyMagic+ leaving becomes more unattractive because you can't use that feature off property.
3)With MyMagic+ Disney united multiple clunky services into one streamlined experience, eliminated ugly and bothersome turnstyles, got rid of the annoying paper slips, eliminated the need to carry multiple cards, and they finally made PhotoPass less of a pain.
4)Attraction Interactivity-So far nothing to report

Those to me represent the main reasons Disney created MyMagic+. Data mining may be on the list, but I don't see why we don't know about anything they've done to back that up.

P.S. Disney has given up on the official Happiest Place on Earth and The Place Dreams Come True because the public won't follow it. It is still official but they no longer stick to it. For some time Disney has been calling WDW the happiest place on earth in various venues because 98% of the people do. And it wouldn't do for Disney to run around correcting everyone, "No! It isn't the happiest place on earth! No, it isn't the happiest place on earth!"

Interesting. I don't think I've ever seen any official Disney materials call any other Disney Park THPOE. Though that problem you mentioned has been something I've run into myself. Very valid indeed. WDW could have a worse slogan attached to it, so I guess they can be grateful for that...

Still I'm never going to not cringe when I hear WDW called that. It's too much.

 

[Mouse of Mystery]

...Specifically, knowing how many people use cash vs credit cards is something that companies have known for fifty-plus years. Nothing new here. Disney has been in possession of much of this information ever since people tied their credit cards to their room key. They’ve been optimizing store locations and product location within those stores for decades...

...Your travel patterns outside of Disney will be completely unknown to them, so data for offsite guests will be limited at best...

..Magic Bands don’t read minds...

Yes, customer data has been available since the stone age, but this is an entirely new level of tracking. The ability to take customer data and marry it with their actual interactions in a setting like this is the next level. Credit cards are only tied to a single person, whereas magic bands consist of entire families as defined by the customer. Yes, merchandising has been around for quite some time. It's science is based on human behavior and data as well. This is meta data that far surpasses that which is available to almost any other retailer.

You said yourself that customer data has been around for decades. All Disney has to do with offsite guests is take available information and match that up with a magic band, regardless of its origin of purchase. It may not be as much data, but does it need to be? Think about the number of guests that stay on property already.

Magic Bands don't read minds -- but they don't have to. You can be sure that they will not abandon other methods of collecting information. There will still be eyes and ears on the ground. They're going to try to predict what you want -- which is exactly what they do now -- they'll just be better at it. Not perfect, no, but better.

Is it because they didn't have large enough databases with sufficiently detailed inputs related to age, income and shopping habits? Or is it because they became infatuated with massive investments in IT data collection and analysis to the point where nobody responsible for making the big decisions ever bothered to visit the stores any more, walk around the neighborhood, and talk to the people who might or might not want to shop there?

You're assuming an awful lot there. Who said data replaces anything? It's another layer of information. The decisions you make based on all of the information available can be good or bad.

My question is…what are they gonna SELL with it?

A lot of that has been covered already, but I'll add a few things for emphasis: there is a very sophisticated science to retail and this kind of information can be added to the arsenal to help them make better decisions -- on product mix, product placement, prices, proximity to other items... in other words, just about everything that involves selling can be enhanced greatly with this kind of data. But it's not a magic pill. Disney has to read the data and respond correctly.

What are they going to sell? They are hoping, I'm sure, to sell more soft and hard goods, as well as food and beverages.

The Magic Bands don't just help with retail, but also staffing and crowd control. In a park like WDW, the possibilities are endless.

 

[wallrock]

Yes, customer data has been available since the stone age, but this is an entirely new level of tracking. The ability to take customer data and marry it with their actual interactions in a setting like this is the next level. Credit cards are only tied to a single person, whereas magic bands consist of entire families as defined by the customer. Yes, merchandising has been around for quite some time. It's science is based on human behavior and data as well. This is meta data that far surpasses that which is available to almost any other retailer.

You said yourself that customer data has been around for decades. All Disney has to do with offsite guests is take available information and match that up with a magic band, regardless of its origin of purchase. It may not be as much data, but does it need to be? Think about the number of guests that stay on property already.

Magic Bands don't read minds -- but they don't have to. You can be sure that they will not abandon other methods of collecting information. There will still be eyes and ears on the ground. They're going to try to predict what you want -- which is exactly what they do now -- they'll just be better at it. Not perfect, no, but better.

Again those customer definitions are inputs that Disney has to rely on to maximize the benefit. To get that information Disney is going to have to sell confidence to guests and show their data is secure. Right now selling that is hard in any industry.

Another hurdle is system complexity. Metaphorically speaking, what the tracking is going to give Disney is a series of snapshots within a confined space (say the Magic Kingdom). That snapshot is going to have thousands of ‘dots’ (a dot representing a guest) scattered all over that confined space. Those dots are going to be moving at various speeds; some will sit idle while others will move rapidly. Some dots will be within the space for an hour or two while others may be there for twelve or more. Some dots will exhibit a certain behavior (spend money) while others will exhibit something else (a five year old throwing a tantrum) which cascades into additional behavioral changes. The velocity, direction and behavior of each dot is subject to change without notice and will not be repeated in the next cycle (a cycle meaning a new day in park operation). A system that complex isn’t going to generate any sort of valuable trending data for years. I don't doubt that they'll do what's possible to capture the data, but tangible benefits, if any, are well into the future. Not sure if Wall Street and the stockholders will be willing to wait that long.

Regarding offsite guests....There’s an assumption that offsite guests will use the magic bands to make purchases. Given that they already feel they’re at a disadvantage to those that stay onsite (mainly due to fastpasses) the magic band may be nothing more than an afterthought once they pass through the gate and instead make their purchases with cash. No real reason to use the band if you don’t share in the perks offered up to onsite guests. No telling if this behavior will manifest itself or not, and it's too early in the game to know which direction it's going.

 

[Josh Hendy]

A lot of that has been covered already, but I'll add a few things for emphasis: there is a very sophisticated science to retail and this kind of information can be added to the arsenal to help them make better decisions -- on product mix, product placement, prices, proximity to other items... in other words, just about everything that involves selling can be enhanced greatly with this kind of data. But it's not a magic pill. Disney has to read the data and respond correctly.

I'm sure it's very scientific. Or as scientific as a bunch of computer, mathematics, accounting and psychology boffins can be, who are sitting in an office many miles from the place where the lowest underlings in the workforce are supposed to be helping them to make money.

Let's assume a nearly 100% lack of any individual entrepreneurialism, business savvy, or profit motivation in any Disney employee working in the retail sector - which is I suppose a pretty safe bet. There is as a result a total lack of any meaningful or purposeful human observation of shopping behavior or interaction with any customer by anyone who is actually motivated to give a d)mn what the customers buy.

Given that situation, then I suppose that a large company's most powerful executives ... with the most control over how the stores are run ... are many layers above the front-line employees in the org chart. In that case, they have no way of observing and interacting with customers except by electronic eavesdropping and by implementing behavior-modification techniques which were pioneered with laboratory animals and then refined by running simulations on spreadsheets.

Do you follow me? The family that runs your nearest corner store doesn't need their customers to wear tracking bracelets, and they don't need to be told where to position the higher-margin candy displays and where to stack the cases of soda ... by someone working 8 layers above them in an org chart. But a store staffed with near-minimum-wage kids who aren't allowed to change the position of a single stuffed toy without permission ... what the people working at HQ really need to make the store successful is a really huge amount of data and a really large staff of experts to help them decide whether the goofy hats should go to LEFT or the RIGHT of the entrance as you walk in from Main Street. This mighty arsenal of managerial know-how is called ... SCIENCE !!!

Let me summarize very briefly: I understand why a really large and unavoidably blind and dumb company would want to track everyone, every second of the day, with electronic thingies. But I don't see anything to like and admire about it. In fact it's a joke.

 

[lockedoutlogic]

A lot of that has been covered already, but I'll add a few things for emphasis: there is a very sophisticated science to retail and this kind of information can be added to the arsenal to help them make better decisions -- on product mix, product placement, prices, proximity to other items... in other words, just about everything that involves selling can be enhanced greatly with this kind of data. But it's not a magic pill. Disney has to read the data and respond correctly.

What are they going to sell? They are hoping, I'm sure, to sell more soft and hard goods, as well as food and beverages.

The Magic Bands don't just help with retail, but also staffing and crowd control. In a park like WDW, the possibilities are endless.

This isn't "retail"...

Retail doesn't involve a $5,000
Investment and travel to "enter the store"... So the simple "it provides valuable data for product placement, offerings, trends, blah, blah, bleh..." Doesn't quite compute.
Retail implies choices and a competitive market and that is not really the case within the compound - which is the vacuum the bands exist in.

What I'm referring to is that they have systematically streamlined and gutted the appeal of their merchandise with a growing segment of their clientele (locals and DVC definitely) by eliminating variety... And now they have a data tracking system that may be obselete...

When I ask what are they going to sell?... Its semi rhetorical in that they have already sold it and its an up hill climb to recapture a segment that had moved off it.

They put the same stuff in every shop...isn't it fairly marginal to analyze where they put it at this point?

And I do agree that crowd and especially LABOR control is a prime goal of the bands... It's an operational tool at least as much as it is a "guest interface" system.

 

[MrJobson]

Specifically, knowing how many people use cash vs credit cards is something that companies have known for fifty-plus years. Nothing new here. Disney has been in possession of much of this information ever since people tied their credit cards to their room key. They’ve been optimizing store locations and product location within those stores for decades.

This is true. The new things behind the "big data" movement of late is largely that computer processing has become powerful and cheap enough, and this new capacity has spurred development of more sophisticated algorithms and data analysis techniques. The data isn't new. The technology has just made it much, much easier to collect and analyze, and as a consequence humans have figured out better ways to do and use this analysis.

Magic Bands don’t read minds.

To your point humans are rather unique but one thing that no one has ever been able to do is determine what someone will do in the next five minutes regardless of how much information you have collected. The only pattern they will see is that the pattern is constantly changing. The biggest benefit they have is studying the complex system of a crowded park so they can manage the system as a whole and staff accordingly. Without a doubt there will be some benefits, but nowhere near the complexity you indicate.

I do agree that one of this biggest benefits will be in improving efficiency in crowd management, and that this is one of the less publicized and less sexy reasons why Disney was willing to invest so much in the next gen project. But I would argue that you are underestimating how much information can be gleaned from large amounts of data and overestimating how unique individual human beings are.

For example, one of the earliest in depth articles I am aware of examining how companies were leveraging data was in the New York Times magazine in 2012 and principally discussed how Target was trying to mine the data it had available to identify women who had just become pregnant, because earlier research had identified that people's shopping habits were hard to change but that when they became new parents was one of the few times their brand and store loyalties actually could actually be changed and captured with appropriate marketing. So Target began mining their shopper data to try to predict which women were pregnant so they could market to the soon to be new parents before anyone else. The whole article is well worth reading on the general subject of this thread, but it has a fun anecdote about the power of algorithms:

About a year after Pole created his pregnancy-prediction model, a man walked into a Target outside Minneapolis and demanded to see the manager. He was clutching coupons that had been sent to his daughter, and he was angry, according to an employee who participated in the conversation.

“My daughter got this in the mail!” he said. “She’s still in high school, and you’re sending her coupons for baby clothes and cribs? Are you trying to encourage her to get pregnant?”

The manager didn’t have any idea what the man was talking about. He looked at the mailer. Sure enough, it was addressed to the man’s daughter and contained advertisements for maternity clothing, nursery furniture and pictures of smiling infants. The manager apologized and then called a few days later to apologize again.

On the phone, though, the father was somewhat abashed. “I had a talk with my daughter,” he said. “It turns out there’s been some activities in my house I haven’t been completely aware of. She’s due in August. I owe you an apology.”

Now, this story is a little too perfect, which means it is almost certainly not true. Or at least not completely true. However, there is plenty of actual science to back up the value of big data. Take, for example, a couple of recent studies looking simply at what could be predicted just by analyzing a person's "likes" on facebook.

In the first study, researches found that just by analyzing the facebook likes they could accurately predict factors including "sexual orientation, ethnicity, religious and political views, personality traits, intelligence, happiness, use of addictive substances, parental separation, age, and gender." So, with no more information about you than what posts on Facebook you had hit that little like button about they could determine your sexual orientaiton with 88% accuracy, your political party with 85% accuracy, and even obscure things like whether your parents were divorced when you turned 21 with 60% accuracy. These are things that a human being could not regularly intuit about another human being from casual conversation, yet could be predicted with relatively high accuracy from just one bit of relatively minor information people published online.

In another study (from some of the same researchers), they asked a large number of facebook users to fill out a personality survey. They then evaluated how strongly they were able to predict the psychological self-assessment from those users using only their facebook likes. Finally, they asked some of their facebook friends to describe their friend's personality. They found that "computer-based models are significantly more accurate than humans in a core social-cognitive task: personality judgment. Computer-based judgments (r = 0.56) correlate more strongly with participants’ self-ratings than average human judgments do (r = 0.49)." Basically, by feeding their computer algorithm your facebook likes it was better able to predict how you assessed your own personality than your friends and relatives (i.e. other actual human being who had multiple previous social interactions with you) could. The difference was not huge, but the computer was significantly better even with relatively minimal information.

These were computer-based judgments using just one relatively basic data point. Disney will be able to analyze what a statistician might call a ridonkulous number of data points in their attempt to predict how to be more efficient, what really makes guests happy, and, probably most of all, what will make them the most money.

Big data has already delivered some really impressive results in everything from medicine to making sure I have a cheap and convenient ride home from the bar, and there is reason for much optimism about what else can be done with these techniques. Disney could figure out what kinds of things specifically makes you and your family the most happy on vacation and make sure that you get those things even while they are simultaneously serving tens of thousands of other quests. There are also many, many terrifying aspects of what can be done in the error of big data. But at the choice between optimism and pessimism I am going to lean towards optimism for now.

So I for one welcome our new computer overlords. And I'd like to remind them that as an occasional internet message board poster, I can be helpful in rounding up others to toil in their underground data caves.

 

[Tonka's Skipper]

You couldn't be MORE wrong. No this is NOTHING like what Amazon, or the local grocery store, or any other retailer does. It would be equivalent if these places had all your demographics and tracked you walking through the store, knew how many wer in your group and their ages, and knew how long you stopped and at what products.

You have to be kidding!? Knowing what you buy a non-issue?! Grocery stores will place a stand with junk widgets in a pink display right in the middle of an isle specifically because X% more widgets are sold and Y% more when it is pink instead of blue. Important stuff! Grocery stores offer you that discount card so they can track your spending. They would LOVE to know everything about you and track every movement if they could figure out a way.

Data mining isn't the only reason Disney did this. It is only 98% of the reason. You have to understand what they can do with this. Stretch your imagination a bit:

Not only do they know a person went into a store, spent 12 minutes, and didn't use a credit card (In the cash example), but they will know exactly how many people spent cash, how long that group spent in the store, what their age, home place, and party size were. They will know if this party had a credit card and just chose not to use it. They will know if this trip in the store the entire party was there or if no kids were there. They will know how many days, what parks, how long this group spent in the park, whether or not they stayed on property, how often they visit, etc., etc., etc. AND they will know how long they were in the park before they visited the store. AND they will know what the weather was that day to know at what temperature people go into the store and how that affects spending both cash and credit card!

They will know if they move a planter where people sit 30 feet from its current location outside the store, how that will affect spending. They will know if older patrons spend less because they sit further from the store.

They will know which pavilions in World Showcase take the longest to walk by for people from certain countries and certain ages. They will know how a 55/51 year old couple move and spend money when by ourselves and how we move and spend money when the kids are with us. And how we change travel patterns when it rains and is warm versus when it rains and is cold. They will know how much people that attach a credit card to the band spend on the credit card and how much those that don't attach it to the band spend on the credit card. And their ages, stay length and home state/country. They will know what line lengths we will tolerate and what we won't and how much weather affected this and which rides. AND they will know in each scenario whether or not and how this affected our spending!

I could sit here and list at least 100 different things Disney can track and cross reference with each or all of the other items. There is an endless amount of information that is extremely, extremely, EXTREMELY, valuable. Human beings are fascinating creatures. While they have free agency and make individual decisions, they also have relatively predictable response to stimuli if you can study the effect. Changing the color and design of a store exterior may affect the money spent. This is something measurable. Maybe it makes a difference to 50 to 60 year olds but not 20 to 30 year olds.

There are two levels to this value that will make Disney money:

1. They will know the impact every change has on every demographic group. They will be able to make changes that cause spending changes. For example: Placing hats in the window of a particular store attracted a particular age group without children and resulted in 8% more hats sold, but placing Frozen merchandise in the window made a higher percentage of younger families stop in the store and sold 12% more Frozen merchandise at twice the profit of the hats. OK, change the window display back to Frozen. Oh and this is true in the summer but not the winter so change it back in the winter.

2. They will know how changes affect our habits and will know how to improve our experience to keep us happy. This is a more indirect relationship to their profits but still very important. This might mean they find that the placement of walkways and planters makes people go left instead of right and slow down in a crowd, stop, and neither shop or ride at one particular point. They rearrange the walkway and find people move differently and whether or not it is an improvement. AND they will know the demographics of the groups whose behavior changed and how much and in what manner each groups behavior changed.

This cannot be overstated. IT IS PURE GOLD! No one has every been able to track this number of peoples movements and actions in a theme park before and tie it back to specific data about those customers.


P.S. Disney has given up on the official Happiest Place on Earth and The Place Dreams Come True because the public won't follow it. It is still official but they no longer stick to it. For some time Disney has been calling WDW the happiest place on earth in various venues because 98% of the people do. And it wouldn't do for Disney to run around correcting everyone, "No! It isn't the happiest place on earth! No, it isn't the happiest place on earth!"

Really?.........I never said Disney wasn't tracking with MBs..............BUT WHO CARES.............I really don't...........whether I buy a Pluto for the grandkids or eat here or there?......when I book FP+.............WHO CARES....its meaningless. Let them arrange and maximize their sales, that is what ALL companies do and it doesn't make Disney EVIL...................If you don't want something....DONT BUY IT! That's the individual guests choice to buy or not!

What scares me is all the others stores/companies , Amazon, car dealerships, LL Bean, Macys , Target, Banking ,CREDIT RATINGS, when I buy a House or rent a apartment, buy a boat (major purchase), tracking us by our cell phones, cams, facial recon. on public street for other then criminal matters. When the details like this are brought and sold between them they are actually EFFECTING peoples lives and often in a negative way. This has been going on for a number of years now, years before the MBs...............his is what we need to worry about and be scared about.

AKK

 

[bbarba]

I work in secondary education. I am a big proponent of using the data in schools. Most teachers I know disagree, they feel the human element is gone from a bunch of numbers and that we forget about the individual. I argue that it is all important and that the data can be used to make better informed decisions.

For example: we should be able to identify a subset of 400 sophomores who struggle in writing document based essays (a social studies exercise). Our teachers can identify them, but often preconceived notions can get in the way. If we use the students scores on said DBQs over a period of several years coupled with teacher feedback, we can create a class where these students are together and can be given added support to improve their skills in their regular class, not needing a special class for this.

Disney constantly receives feedback from CMs and guests, this coupled with data, helps them tailor the experience so everyone benefits (not just about making money today, but ensuring people return in the future).

 

[Mouse of Mystery]

Forgive me for chopping up your posts, but I wanted to quote mainly the points to which I wished to speak:

Again those customer definitions are inputs that Disney has to rely on to maximize the benefit. To get that information Disney is going to have to sell confidence to guests and show their data is secure. Right now selling that is hard in any industry.

A system that complex isn’t going to generate any sort of valuable trending data for years. I don't doubt that they'll do what's possible to capture the data, but tangible benefits, if any, are well into the future.

I think you overestimate the resistance people have to sharing their information. Right or wrong, they give it away freely whenever they feel like it benefits them. You need look no further than Facebook to see a snapshot of what is given away. Now people sure love to raise a stink when their trust is violated, so there's a great responsibility on the part of the data gatherer. Disney should be well-aware.

I believe you underestimate the sophisticated software that is available to make sense of those blips and dots you say Disney will be seeing. In fact, I think you'd be shocked at how accurately they can make predictions. For everyone? Of course not. But they don't need to be right 100% of the time to make it pay off.

I'm sure it's very scientific. Or as scientific as a bunch of computer, mathematics, accounting and psychology boffins can be, who are sitting in an office many miles from the place where the lowest underlings in the workforce are supposed to be helping them to make money.

Given that situation, then I suppose that a large company's most powerful executives ... with the most control over how the stores are run ... are many layers above the front-line employees in the org chart. In that case, they have no way of observing and interacting with customers except by electronic eavesdropping and by implementing behavior-modification techniques which were pioneered with laboratory animals and then refined by running simulations on spreadsheets.

You paint an odd picture of a retail environment in which there are unknowing overlords and mindless servants. I don't have knowledge of the inner-workings of Disney, but they are a savvy retailer / entertainment corporation, and I doubt they are 100% buffoons at every level. You can be sure that there is a middle layer of supervisors who are keenly interested in obtaining human feedback and who are required to provide their own on a very regular basis.

This isn't "retail"...

Retail doesn't involve a $5,000
Investment and travel to "enter the store"... So the simple "it provides valuable data for product placement, offerings, trends, blah, blah, bleh..." Doesn't quite compute.
Retail implies choices and a competitive market and that is not really the case within the compound - which is the vacuum the bands exist in..

Of course it's retail. They are in the business of selling and that involves not only your $5,000 entry fee, but capturing every possible dollar you have to drop along the way. And it's not at all the vacuum you perceive it to be. Disney tests merchandise at all of its parks and works with vendors from all around the globe. You can also be sure that they observe other retailer / entertainment companies.

You sound very jaded with their current merchandising direction. Don't allow that feeling to confuse the issue that they now have a very powerful system at their fingertips. If there are smart people behind the curtains pulling the levers, then things should get better at some point. Again, they aren't going to please everyone, just more people, statistically speaking.

 

[George4]

Sorry, a good portion of what you’re saying depends on input and feedback from the customer. The magic band isn’t that ‘magic’. Other items you state are either pure conjecture or something that can already be done without tracking your every move. Specifically, knowing how many people use cash vs credit cards is something that companies have known for fifty-plus years. Nothing new here. Disney has been in possession of much of this information ever since people tied their credit cards to their room key. They’ve been optimizing store locations and product location within those stores for decades.

The personal data you speak of is voluntary, and not everyone is willing to give all or some of it up. This cannot be understated.

It actually isn't voluntary. If I collect all of this data about 50 million people per year, and I can get even 20 of them (yes, just 20) to give me some feedback on why they did some things, I can extrapolate from that all kinds of fairly accurate descriptions of behavior. If I get even 20 PER DAY, the accuracy of that information becomes phenomenally accurate. I am then also able to forecast behavior of the other 49.99 million visitors without ever talking to them and understanding the "why" of their behavior. Besides all of that, you can ask me why I do certain things, and I can tell you. But it may not be true, because I may not really know why. If you look at all of the decisions that 50 million people make time and again, you can draw reasonable conclusions about the why of the behaviors across that many people, and be fairly accurate without ever talking to them. You can say it isn't true, but it is. Data is a powerful machine to determine what the behavior patterns are, how we can affect the patterns, and we can change them to what we want. That may mean something as simple as retail inventory and staff deployment, and something as complex as availability of information at home about potential future vacations.

Data allows companies to predict who would be interested in a 5 day vacation versus a 10-day vacation (or longer), who would be interested in a cruise, who would be interested in a park in Timbuktu, who wants Disney stuff available at home and who wants to purchase what, where and when, and that doesn't just pertain to the $5 plush toy - it also covers a $25,000 DVC purchase or vacation trip. Disney can accurately determine who to market to, and what should be marketed. They can determine where to deploy cruise ships from and to, in an effort to increase visitor count.

If this data system only "extracts" $10 per customer per year, that is still around 3/4 of a billion dollars. That $10 plush toy is now a big deal.

 

[Lewisc]

Do you have to mine data from wrist bands to find out that people like to go on exciting theme park rides without too long of a wait, they enjoy eating high quality food at a reasonable price, and they like to shop for interesting, unique and reasonably priced clothes and souvenirs? And that they'll leave early and go watch TV or swim in their hotel pool if they don't get those things? And then next year maybe vacation somewhere else.

I"

You're missing the big picture. Disney is continuing to learn how to sell mediocre food at higher then reasonable prices. How to sell high priced clothes and souvenirs. The goal of FP+ seems to be rationing the top attractions. You'll get reservations, spaced during the day, for 2 or 3 things you want to see. You'll be free the rest of the day to shop.

Magicbands tells Disney what guests are doing. Are they shopping during the gaps? Are they waiting in long lines for E rides. Are they spending time on secondary attractions.

 

[Rileygirl]

Yellowstonetim- I agree and share your sense of 'holy crap - this is almost terrifying with the applications and ramifications'.

I keep thinking of an ant farm analogy. See, imagine Walt is holding on to an ant farm. And how big is this farm - well, I imagine that Disney resort guests alone number somewhere between 50,000 and 100,000 every day, 365 days a year. (maybe way more - I don't actually know). But anyways, imagine 50,000 ants in an ant farm, and every single one of them is tagged. They are tagged with a name, an age, a home address, and quite possibly a visa card as well. And they are also tagged in 'family groups' - so its not all individual information, but also pertinent group information. And while in this ant farm, every time they open a hotel room door, purchase an ice cream, shop, eat, ride an attraction, go to the bathroom, even their traveling paths, time spent in locations, everything is recorded. 24hrs a day, 7 days a week, 365. 2 days a year. Also imagine that all of these ants are not the same, some of them are local, some national, and some international. Hmm. Some of them are married, some of them are teens, some of them are kids, some are grandparents, some of them are school friends, some of them are partners. All sorts of family groups. Hmmm.

Now, imagine all those ants, 50,000 of them, on any given day, going about their business. Some will be here, some will be there, but all are 'contained' in the ant farm called Walt Disney World. This is not a small sample folks. Its huge. Its accurate. Its consistent. And the potential for this data is crazy.

And now I keep thinking about that 5th theme park Disney World is opening. You know, that one that has no entrance fee, but what you do in there costs money. I think its called 'High End Shopping Mall at Downtown Disney' or something like that. So instead of the plush mickey toy, think about that Michael Kohrs purse, or those Tory Burch boots, or that Gucci bag being involved. I wonder if these high end stores would like to participate in the data mining of the ant farm? I wonder if they would like to get data on every single person who walks through the door, their age, Where they are from, how long they spend in the store, if they buy anything, what time of day it is, what day of the week, maybe even what they are looking at and how long they looked at it for. I wonder if that information is valuable.

Also, I want to talk of Mutual of Omaha's Wild Kingdom analogy - imagine Marlin Perkins and well, actually, his side kick Jim, being able to tag all the ants, and decipher data on feeding patterns, sleeping patterns, movement patterns of certain family groups, how they react in the rain, or intense summer heat, or when the water hole is crowded and hard to get at (xmas, easter, spring break). They can look at time of day data, time of year data, age of the ant data, and get some pretty amazing information. Can they come up with models on how to make the ants eat at certain times of day, or be in certain places at specific times, or keep the ants longer at the ant farm?

Imagine Barney Miller and the gang being able to get data on how many ants are in one place at one time, with relevant information on the total number of ants, their movements, and ALL OF IT associated with day of the year, day of the week, weather that day, type of group, where they are from. Maybe they could staff the police station on not only PAST DATA, but SELF REPORTED actual plans of those darn ants! You not only know what city the ant intends to be in, 60 days before the ant even gets to the ant farm - but you know what time of day that ant plans to be at specific geographical points - eating or riding rides.

Imagine Jack Bauer, and all those corrupt government officials on 24, well - just imagine how they could decipher and manipulate the information to control those ants. And its not all past, or even real time information. The mind blowing part of it - its also hard core future numbers that are self reported by those ants!

If I could imbed pictures and photoshop - I would end this post with a picture of Marlin Perkins directing Jim to tackle and tag Jack Bauer while Fish and Barney look on holding a pair of Mickey Band handcuffs. :>)

'edited to add the word handcuffs - ' don't get excited, Fish is the one holding them.... :>)

 

[wallrock]

It actually isn't voluntary. If I collect all of this data about 50 million people per year, and I can get even 20 of them (yes, just 20) to give me some feedback on why they did some things, I can extrapolate from that all kinds of fairly accurate descriptions of behavior. If I get even 20 PER DAY, the accuracy of that information becomes phenomenally accurate. I am then also able to forecast behavior of the other 49.99 million visitors without ever talking to them and understanding the "why" of their behavior. Besides all of that, you can ask me why I do certain things, and I can tell you. But it may not be true, because I may not really know why. If you look at all of the decisions that 50 million people make time and again, you can draw reasonable conclusions about the why of the behaviors across that many people, and be fairly accurate without ever talking to them. You can say it isn't true, but it is. Data is a powerful machine to determine what the behavior patterns are, how we can affect the patterns, and we can change them to what we want. That may mean something as simple as retail inventory and staff deployment, and something as complex as availability of information at home about potential future vacations.

Data allows companies to predict who would be interested in a 5 day vacation versus a 10-day vacation (or longer), who would be interested in a cruise, who would be interested in a park in Timbuktu, who wants Disney stuff available at home and who wants to purchase what, where and when, and that doesn't just pertain to the $5 plush toy - it also covers a $25,000 DVC purchase or vacation trip. Disney can accurately determine who to market to, and what should be marketed. They can determine where to deploy cruise ships from and to, in an effort to increase visitor count.

If this data system only "extracts" $10 per customer per year, that is still around 3/4 of a billion dollars. That $10 plush toy is now a big deal.

Actually it is. There is no personal data that you are obliged to provide Disney other than what’s been provided for years when you make reservations.

Again, to get the feedback you seek you need to directly interface with the guests. Most of the conversation has been around passive scanning of thousands of people all doing different things and somehow that’s going to generate valuable data almost immediately that they can use for financial gain.

And I’d be curious as to how you can take information from a small handful of people and somehow extrapolate that into forecasting behavior for millions. Complex systems simply don’t work that way. If you know of one please provide the example.

And I’m not following you regarding someone asking you why you do something (in this example, probably why you made a certain purchase or decided to tour the park a certain way) and you’re saying the answer may not be true because you don’t know why you did that something. Given the planning required to visit Disney nowadays plus the money spent people don’t blindly buy a Mickey ice cream bar or a Goofy hat and not know why they did it.

 

[George4]

And I’d be curious as to how you can take information from a small handful of people and somehow extrapolate that into forecasting behavior for millions. Complex systems simply don’t work that way. If you know of one please provide the example.

This is what we do with statistics, and it is what can be done with any sort of data with someone that knows how to read it. It is a simple process, and is quite accurate (like >95% of the time if you do it correctly).

And I’m not following you regarding someone asking you why you do something (in this example, probably why you made a certain purchase or decided to tour the park a certain way) and you’re saying the answer may not be true because you don’t know why you did that something. Given the planning required to visit Disney nowadays plus the money spent people don’t blindly buy a Mickey ice cream bar or a Goofy hat and not know why they did it.

Okay, that came out weirder sounding then it was in my head. Let me try it a different way:
Survey: I noticed you didn't purchase anything while you were in the store just now. Do you mind if I ask why?
Me: I didn't see anything I liked.
Survey: Thank you.

The reality is, I didn't know what I was looking for, or maybe I was just getting out of the heat, or maybe I just wandered in because it was crowded in the street and I was tired of the crowds. In the first and third scenario, what was really happening in my head isn't what I told the person. Maybe that was on purpose, and maybe it really just didn't cross my mind that those were the reasons I went in the store. That's all I meant - people that act initially on an impulse don't always cognitively know why they did something, and if you ask them, all they can give is a general cursory answer, because they really don't know why. From your statement above, we know why we purchased something, you are correct there. We don't always know why our behavior was a certain way because it was an impulse, and impulses can be mapped, tracked, and eventually be accurately predicted for any type of person anywhere.

Several people were calling this scary in some of these posts. I guess it can be taken that way, but it is what all marketers and retail stores try to do on a daily basis. Disney just has access to a much higher number of people to pull data from that will likely make their behavior predictions much more accurate than most companies. It is very complex dealing with the number of people and data points, but as they learn to manage the information, it could help them run a substantially better company if they really listen to what they can learn.

Of course, maybe they will use it to go cheap because most people don't care about new attractions because they only go once to WDW, which will make us repeat visitors even more annoyed than we may already be.

 

[Mouse of Mystery]

Actually it is. There is no personal data that you are obliged to provide Disney other than what’s been provided for years when you make reservations.

Again, to get the feedback you seek you need to directly interface with the guests. Most of the conversation has been around passive scanning of thousands of people all doing different things and somehow that’s going to generate valuable data almost immediately that they can use for financial gain.

And I’d be curious as to how you can take information from a small handful of people and somehow extrapolate that into forecasting behavior for millions. Complex systems simply don’t work that way. If you know of one please provide the example.

And I’m not following you regarding someone asking you why you do something (in this example, probably why you made a certain purchase or decided to tour the park a certain way) and you’re saying the answer may not be true because you don’t know why you did that something. Given the planning required to visit Disney nowadays plus the money spent people don’t blindly buy a Mickey ice cream bar or a Goofy hat and not know why they did it.

The difference, you see, between data "they've been getting for years" and now, is that it can be put into meaningful context. Paper tickets only tell Disney you entered a park, they don't follow you around all day, relaying your every move. If Magic Bands can allow them to pinpoint your family in a park to take your picture, then it can darn sure lay out a virtual map of what you did in the park all day, what you bought and where you ate (non-cash), where you stopped (and didn't) -- heck, how many times you went to the restroom. Now imagine hundreds of thousands of those maps and they can all be broken down by age, sex, country/state of origin, family size, how much/little was spent, etc.

Without knowing exactly what is being done with the data, no one can really explain how it can be used to predict, but we do know that it CAN be used to see what people are and aren't doing. It's undoubtedly much easier to experiment with new ideas and measure the success.

There is a science called human behavior and it can be (and most definitely is) applied to the retail space. The most famous examples are grocery stores and their complex layouts. They've managed to do that through years of studying data -- imagine what they could do if they could track your movement and speed throughout the store. Studies are even being done on how to capture the length of time people spend looking at things and learning how to tweak that. If they figure that out, you can bet Magic Bands will become Magic Glasses some day.