by Tim Cose, Hal Leonard Corporation Vilfredo Frederico Damaso Pareto was born in Paris to an exiled Genoese family in the summer of 1848 and he started growing peas after moving back to Italy in his mid-forties. Pareto, an engineer by trade, observed (among other things) that 20% of the pea pods in his garden produced 80% of the peas. That observation begat many others as Pareto delved into sociology and economics throughout later life, most famously that twenty percent of the population of Italy owned eighty percent of the land. This phenomenon does not stop with peas and wealth though. Modern science has found that approximately twenty percent of oil wells in an oil field account for eighty percent of the value. Twenty percent of files transferred via TCP account for eighty percent of the overall data flow. Twenty percent of software programming bugs cause eighty percent of computer crashes. The list goes on, but twenty percent of the relevant examples should make up eighty percent of the impact. If it sounds familiar, that is because the 80/20 rule (eponymously known as the Pareto Law or Pareto Principle) is a common heuristic in business today. If you have been told that twenty percent of your customers make up eighty percent of your revenue, you now see that there really is a history of statistics to back up the claim. In Pareto’s case, the essence of the relationship is eighty percent of effects coming from twenty percent of causes. While not everything follows exactly an 80/20 ratio, the implication that mankind’s statistically comparable events can be spoken of and relied upon to follow a mathematical trend holds a lot of promise for more than just pea farmers. Phenomena that conform to such a ratio are said to have Power Law relationships. Here’s a graph illustrating such a relationship (don’t worry the calculus is just there to look fancy): In business, and especially in retail, arguably our most essential ‘effect’ to affect our success is revenue (overhead too is crucial, but without revenue, all the cost savings in the world won’t turn a profit). Thus if revenue is the effect, it follows that transactions are the ‘cause’. Many variables conspire to influence transactions: some of them within your control and some of them not; some of them intrinsic to products and some of them universal to your store. Some of them have nothing to do with you or your business. But of all the variables shaping the transactions that cause your revenue, perhaps the most impactful and yet easiest to control is what you have on the shelf. That brings us to print music. Power Law and Publishing I don’t want to let you down, but I should admit up front that I have no specific silver bullet product-mix advice for you (or at least not any advice that will still be accurate by the time you read this). The ideal inventory mix is a rapidly moving target and is very dependant on who you are as a retailer and who your customers are as consumers. The ratio of revenue to product performance, on the other hand, does not change nearly so much. In fact, month to month, it likely remains within a couple of percent of average. Said ratio is also reasonably independent of the sample size. That is to say: it doesn’t matter if you have 100 skus or 10,000 skus, the results are fairly consistent (and, more importantly, the concepts they illuminate are valid). This point in particular is central to the explanation that follows. If all the talk of ratios and product performance was confusing, don’t worry: another graph is approaching (and no calculus this time). To begin, let’s look at the very real national sales of a leading publisher as a model for an ideal store inventory (ie everything in stock at all times), this will serve as a basis for further analysis. We’ll also look at how close this inventory ratio model is for some successful retailers and why that’s important. Note that the following story is true, but the names have been changed to protect…well me, actually. Here are the songbook sales for January 2013 at a leading print music publisher: You’ll have to take my word that the rest of the months from 2013 all look just like this (though the items making it up change very rapidly). Specific dollar amounts are not included because they aren’t especially important. What is important are the ratios the surface area of this curve represent. Instead of Pareto’s Law of an 80/20 relationship, we have something closer to 70% of sales coming from 30% of skus. If we look at just that top 20%, we find it makes up approximately 60% of sales (instead of the 80% Pareto would predict). So where’s the “missing” 20%? It’s in the much discussed and little understood Long Tail. Songbooks are not peas. Contrary to what you may have been lead to believe by salespeople, cute blog articles, or eCom service providers, the Long Tail has very little to do with the products themselves. The Long Tail has everything to do with the preferences of the consumer purchasing those products. That is the difference between peas and print music (or at least one of the differences). Peas need soil, water, and sunlight. Small genetic differences in the plants, along with a long list of variables that we usually call ‘chance’, determine how many peas the plant will produce. Products, usually, do not have genes. However, we can treat products’ marketing, and in turn individuals’ perception of said products, as something similar. Consumer perception and association are the unseen and constantly evolving forces in the shopping experience. They are what favor one product over another seemingly similar product in the same setting. Perhaps more importantly, these perceptions have a sort of memory over time. Taking into account the human elements of tradition and fashion, and the culture-flattening effect of the Internet, you see why all books, just like all pea plants, are not created equal. So, the Power Law distribution in sheet music relies not on magic, but on people’s preference for X over Y. It’s actually a very simple concept. You have just so many customers poo-pooing modern pop and just so many pre-teen lesson students, just so many classical enthusiasts, and just so many honky-tonkers. Each of these customers is Z% likely to prefer product X over product Y. The demographic distribution of your customers, not the artists or the books in particular, are what determine sales ratios. Your most popular product is your most popular product because you cater to a specific type of customer, not because of something intrinsic to it. That brings us back to the Long Tail. Like your most popular sku, the legacy products, niche products, and small-market products that make up the Long Tail are also determined by consumer preference. But because every member of what seem to be homogenous demographics still weighs their preference just a little differently, and because niche products are by definition smaller markets, there’s a fuzzy area of less popular items that, for the consumer purchasing them, are still hugely worthwhile. In that respect, a Long Tail should be seen as something of a necessary evil if your goal is to cater to as many customers as you can. So the more X’s and Y’s there are, the longer the tail stretches and the higher the ‘slope’ of the most popular items reaches. The persistence of culture and the ubiquity of information and exposure provided by the media and the internet skews print music sales ratios from pea plant production, but data from ‘A Leading Print Music Publisher’ suggests that ratios do exist and they do remain consistent despite the passage of time or changes in scale. If you’re looking for a ‘missing’ 20%, look no further than the ever evolving and diversifying marketplace of human culture. Watch for Part 2 of Tim’s article, “On Peas and Print Music” next week on the RPMDA blog. It’s it as wonderfully informative as it is entertaining!