Lexicographic Choice and Trains (Don't leave! This is interesting! Honestly!)

In my previous blog on Rory Sutherland, I talked about how when we make a choice with multiple factors, we approach it in a linear fashion, rather than weighing up all options.

What does this mean? Simply that if there are 2 factors in a decision, let's call them A and B, then we look for the option that scores the highest in A, and choose on that basis alone. We only look at the scores for B if there is a tie in A. If one option had a massive advantage in the B column, it wouldn't matter, even if it was only just beaten on factor A.

A quick example - choosing wine. The first factor is very often "red or white". Then, having eliminated the unwanted reds, we move onto other factors (price, familiarity, recommendations, etc) to choose between the white wine. The decision to have white wine is stuck to, even if they are a little pricey and there is a great bargain to be had in the red wines.

This is Lexicographic Choice. And it suddenly occurred to me that you can apply the theory to our old friends London Midland. I previously talked about how their tendency to promote trains from London to Birmingham on the basis of 3 factors rather than 1 broke the basic rules of positioning in marketing.

But we can see they break the rule of Lexicographic choice, too. Look at the page. There seems to be an implicit need to put ALL the information in front of the customer, so when they weigh up and combine these factors, they would surely conclude that London Midland is the train to choose.

But people don't think and act like that. The factor may vary, but most passengers will have just one value that they use to select which train to get to London. It might be price, speed, or frequency. But once they've found a winner in their chosen category, they are statistically unlikely to let other factors change their mind.

So saying that they have trains up to every 20 minutes is pointless, because other companies offer a more regular service. Mentioning the 20 minutes they've shaved off the journey time is pointless, because it is still slower than their competitors. These things are only worth mentioning as supporting factors to their one true advantage - their price.

The only people they are ever realistically going to attract are those who are looking for the low-cost travel. Anyone looking for, say, high speed travel is unlikely to say "well London Midland is slower, but it's cheaper, so maybe I'll put up with the journey time".

But then train companies are very often run by engineers, so what can you expect?

The best of Rory Sutherland

A bit of a lazy post, and as much as anything for my own future reference - a list of interesting and occasionally connected things said by Rory Sutherland in just one hour-long talk the other day at the School of Communication Arts.

If you're going to see a talk by him, he may well say similar things - so consider this your spoiler warning.

- Always remember that the problem you're trying to solve might just be a symptom of a bigger problem. So why not fix that instead? See the Inventor's Paradox
- Advertising has been under attack from a market-driven neo-classical economics approach to solving problems. Human insight is no longer valued as a way to create a competitive advantage.
- BUT science, research, economics, and logic can give weak answers. They tell us that 35 units is a safe amount of alcohol to drink, so we stick to it. Behavioural economics tells us that keeping a tally is difficult, self-deception is rife, it's hard to police or get others to police, etc. Instead you should just set a rule not to drink on 2 or 3 days of each week - which is a absolute binary choice, easy to remember, and can be done socially with friends or relations, each easily policing each others' drinking habits.
- God was the first behavioural economist, by creating a seventh day of rest.
- Procrustes was the first neo-classical economist.
- Binary absolutes are almost always the easiest rule to follow - "No work on Sundays" > "Only work 40 hours maximum", and "Don't eat carbs" > "Eat x calories per week".
- Eurostar spent £6bn making a marginal improvement on journey times (an engineer's logical improvement) when they could have spent a fraction of that installing WiFi on their trains, rendering the journey more enjoyable.
- "The invention of Angry Birds made Concorde irrelevant"
- "The annoyance of a wait is not duration, it's the uncertainty". London Underground's greatest innovation was dot matrix displays with countdown to next train. South Korea now has red light signals with a countdown to when the lights turn green which has apparently reduced accidents. See also the loading / progress bar in software downloads/installs.
- The best ideas combine technology, economics, and psychology.
- Giving people control and choice is more likely to get them to part with their money.
- Be aware of Heuristics - the rules by which humans come to a decision where a completely scientific and rational analysis of the choices and the situation is impossible (NB this is pretty much all of the time).
- A good heuristic example is our tendency to buy the second or third cheapest wine on the list. Or that in a group of 5 with varying price, the 2nd and 4th most expensive will be the most popular.
- We see things relatively - therefore changing the frame of reference can increase sales. People choosing between a £2 and £1 can of lager will choose the more expensive by 2 to 1. But if you add a 50p can of lager to the choices, the £1 can suddenly becomes more acceptable and more popular. Nespresso works by making people compare it to the cost of a cup from a coffee shop, not the cost from homemade ground or instant coffee.
- When making a decision with many factors, we tend to make it in a linear fashion. When buying a house, people often choose location, then choose the best they can find in that location. They don't think "well I'll keep an open mind in case there's one I love so much in location B that I don't mind not living in location A" - they just rule location B out. See Lexicographic Preferences.
- Behaviour drives attitudes, not the other way round. “When a man says, ‘my wife doesn’t understand me,’ it doesn’t mean he’s planning an affair. He’s already had one.” We post-rationalise. This is why it is easier to change behaviour (make recycling habitual and easy) than attitudes (make people care about the environment so they'll recycle).
- We are happier with limitations than infinite possibility. Spotify offers us unlimited downloads for a certain price, but how much is unlimited worth? If you said 500 tracks, you can make a comparison and feel like you're getting a good deal.
- Anyone engaged in a market research survey is unreliable, because they are acting as a "maximiser" all the time - consciously trying to decide what is best. In the real world, we are "satisficers" - happy with something that will not kill or maim us, break, or be embarrassing. This is why we go to MacDonald's - though you wouldn't take your wife there on a wedding anniversary, when you revert to maximiser.
- Therefore a focus group would analyse Red Bull and tell you that they would never pay 3 times more for it than coke costs. But in the real world, they just changed the size of the can, so people thought of it as a different type of product (see changing frames of reference above), and happily paid the extra.
- Statistically the best way to get sex is to constantly proposition people until you find one (on average, about 15 tries later) who agrees. However very few people, even if aware of the success rates, would be willing to put up with the rejection and social embarrassment in order to try it.
- English football fans are only marginally better than a foreigner with no knowledge of the English game at predicting scores. This is partly because the foreigner will just pick the team he/she has heard of to win. For the same reason, 62% of Americans know that San Diego is bigger than San Antonio. But 100% of Germans select San Diego as the bigger, because it's the only one they've heard of. It's the Fame Heuristic - recognition is a massive factor in prediction and choice.

And that's just the bits I managed to jot down legibly. You can't say he's not good value.