ZIPF’S LAW: STOP WHIPPING LAME HORSES

Recently I was working with a client who owns a café in Auckland, and we were discussing pricing and sales volumes for items on his menu.

He mentioned to me that a customer had told him that his scone was overpriced at $6.50.

I asked him how many he sold a week.

He replied 35.

I then asked how many items are on his menu.

He replied 30.

We then brought up how many food items he sells in a week.

The answer was roughly 450 at an average of $17 per item.

To save you the math, his scone was selling more than its fair share and it was too cheap.

On average each item should sell only 15 in a week, the scone is selling 233% more than what it should be, and at a third of the price of the average.

My reply to this was the scone is either to cheap or you should stop selling it altogether.

Not what he was expecting.

Obviously, it’s not so black and white as this but it did illustrate an interesting topic where people are over-emphasising the impact of a product and its performance based on offhand comments or gut feel rather than data.

What we did next was demonstrate the distribution of sales per item on a graph and what we found was a little mathematical law playing out perfectly.

The law I’m referring to is Zipf’s law.

What is Zipf’s Law?

Zipf’s Law demonstrates the frequency of a certain event is inversely proportional to it’s rank in a data set.

If you have a data set that shows how often events occur, the event that ranks first will occur twice as often as the second ranked option and three times the third option and so on.

You can see it in so many different data sets.

For example, In the English language using the Brown Corpus, ‘the’ is the most common used word which occurs 7% of the time, second place is ‘of’ which occurs 3.6% followed by ‘and’ which occurs 2.8%. If you want to check this out go to wordcount.org, a cool way to visualise it.

It also shows up in populations of cities, occurrences of last names, which keys are pressed the most on your keyboard, and pertinent to this article how often any given menu item is sold relative to all others.

It is one of those weird laws or phenomena’s that show up in the world that are hard to explain exactly why.

So how did it look in this café?

Well it was right on the money. For a dataset of sales of only 3 months it was crazy how in line it was with the law.

The reason this is important is because it highlights how little of the menu actually sells, which then provides a solid argument as to why you would even have most of these items on the menu if they rarely sell.

A lot of business owners don’t look at this data and then make business decisions based on gut feel and ‘industry experience’. They keep pushing products that customers don’t want, or they haven’t marketed correctly. The bigger the company, the more this usually happens.

However, when a company understands this at a fundamental level, they completely dominate the market.

The best example of this is Apple. How many Apple products are there? Compare that to Dell, Samsung, Sony who all have crap loads. Love them or hate them, Apple dominates the tech world, they do this by focusing on what works and what integrates into their eco system and they make it the best in the world, and they know a thing or two about marketing and creating a tribe.

Now that’s not to say their competitors are doing badly, I would be more than happy to own any of them. It’s just Apple really is the market leader, which is hard to say as I type this on my Microsoft Surface and will read it again on my Samsung Galaxy.

Anyway, back to cafes. When you run this model in any restaurant you will get a similar distribution. My advice would be to chop 60-70% of your items and concentrate on having a few items that are the best in the world.

Anyone who has focused on the Hospitality industry in recent years has seen the shift towards casual dining restaurants and away from all day dining restaurants that cater for the broad market. Instead we have burger joints that offer 3 burgers and chips, instead of bakeries or patisseries we have specialty donut shops that only offer coffee to go with their round sugary ambrosia, in Melbourne there is a guy that sells nothing but a few types of Gnocchi and so many more examples that are thriving in todays’ competitive Hospitality scene where businesses are lucky to last out the year.

The reason these outliers thrive is simple, you already probably get it but to spell it out to you, it’s because they concentrate on being the best in the world at that one thing, they have lower overall costs as they aren’t keeping stock on hand for menu items that don’t sell, they can charge a premium for what they serve because it is the best, they can scale and do volume with their resources concentrated on minimal products, the service is improved because the product knowledge of their staff is absolute which in turn means less training costs to get recruits up to speed, and the list of benefits go on.

What’s the takeaway from this then?

Run some data on your business and see what comes out from it.

Is Zipf’s law at play in your business? Are you whipping lame horses that are never going to produce for you wasting resources that could be better used for growing the products or services that you are the best in the world at?

If you don’t have that data, then start collecting it. Stop making decisions based on your gut feel. It doesn’t work anymore, because your competitors aren’t. They have hard data and they are using it every day to make business decisions.

Cheers,

Stanley Henry

Misfit | Dreamer | CEO

Wallace & Henry House