Milkshakes & Product Failures: A Lesson on Segmentation

I came across an article called ‘Milkshake Marketing’ by Clay Christensen – and a startling statistic jumped out at me. According to Christensen:

“Each year 30,000 new consumer products are launched—and 95 percent of them fail.”

That’s astounding! 28,500 failures and 1,500 successes every year. At 5%, you would have better rate of return by investing in stocks you randomly select – that average being 12%.


Blind Ideation

Christensen goes on to say that the major problem is ineffective market segmentation, a causation/correlation bias. For example, just because a consumer is the 18-35, caucasian male, college educated bracket, it doesn’t mean that consumer will buy the product. There may be a correlation in the demographic that does purchase the product, but being in that demographic alone doesn’t cause it.

What makes a successful startup is one that looks to solve a problem, therefore meeting the needs of a market. In Fact, part of the Lean Startup philosophy is to beta test your product, and if the idea doesn’t work then you need to tweak the business model, or pivot and find a market that matches it. The same is true in regards to launching a new product. What problem is the product trying to solve, and who is it trying to solve it for?

In Milkshake Marketing – Christensen describes the story of a fast food chain that is looking to improve its milkshake sales. The company did all the right things with segmentation. They segmented by:

  • Product: Milkshakes
  • Demographics: Typical milkshake drinkers profile
  • Characteristics: Surveyed demographic; describing characteristics of the ‘Ideal Milkshake’ (thick, thin, chunky, smooth, fruity, chocolaty, etc.)

The company responded to what their key demographic wanted – yet milkshake sales failed to increase. What the company failed to do was to understand what problem their milkshake was solving for their customers. What ‘job’ were the customers ‘hiring’  the milkshake to do?

Christensen’s team was hired on to help solve this problem. They spent a full day in one restaurant chain documenting:

  • Who bought the milkshakes
  • When they bought the milkshakes
  • Where they drank the milkshakes

His team discovered that 40% of the milkshakes were purchased first thing in the morning, by commuters who ordered them to go. The following day his team returned to the same store and interviewed customers who ordered milkshakes to go, asking “what job they had hired he milkshake to do.”

The results of his study were that most commuters bought the milkshakes for similar reasons. They were in a hurry, wearing work clothes, and had (at most) one free hand. They ‘hired’ the milkshake because they:

  • Staved boredom (drinking a thick milkshake through a thin straw)
  • Something to keep their one free hand busy
  • Not hungry, but could satisfy appetite for later
  • Tidy (no crumbs or grease)

In a TED presentation, Malcolm Gladwell tells the story of Howard Moskowitz, a psychophysicist who is famous for reinventing spaghetti sauce by realizing earlier in his career “There is no perfect Pepsi, but perfect Pepsi’s” – or simply that there is no single product for everyone, but there are multiple variations of products depending on wants/needs.

In the case of milkshakes: a regular milkshake can satisfy everyone, a thicker fruit shake can satisfy commuters, and a thinner non-fruit shake can satisfy children (or more likely, the child’s parent).

… there is also the rich oil tycoon demographic who will drink your milkshake …

Daniel Day Lewis in 'There Will Be Blood'

Take Home Message:

While it is important to know your demographics, it is more important to know what need you’re fulfilling. Most markets already have a product for everyone, what is needed are products for the sub groups.

The easiest way to know ‘what job your product was hired for’ – is ask yourself the 5 W’s & H for each product you have, or for any product you’re about to launch. From here look to see if there are any clusters of demographics. If you dont know what need you’re fulfilling, then you’ll probably have a better ROI by launching a random product than one without a clear direction.

Be An Expert – part 1

What I am about to say applies for both individuales and businesses … what are your strengths, what are the things that you do better than others? Take a moment to really think those questions over. This will take a lot of introspection, and looking at yourself from the outside as well, but there is nothing more powerful than knowing what your core is – and I’ll tell you why in a moment.

I first heard of Psychologist K Anders Ericcson’s study on what makes an expert back when I was training athletes in 2004. The study examined violinists from age 5 till 21, examining how much they practiced and what level of mastery they achieved. Here is a graph for it:

The premise of the graph is that around age 8, certain violinists practiced more, and those who by 21 were considered the best practiced a total of 10,000 hours. This brings about the rule of thumb that 10,000 hours of practice is the minimum you need to do to reach expert level in whatever you do – whether it is the violin, sports, or even academia. The study went on further to say that no prodigy effortlessly mastered the instrument, and no ‘normal’ child who worked harder than the others never made it to the top.

Our strengths are strengths for a reason – they are skills that we have spent more time nurturing because we enjoy using them. Whether your a business who has great customer service, or a person who is a hobbyist photographer, your skill level will be higher up than those who barely focus on those things.

Think about that for a second. If you enjoy doing something, who knows where you currently are on that graph. When I first heard of this study, it was described as training 3 hours a day, seven days a week, and within 10 years you will have your 10,000 hours and be at expert level. Most of us are at least 1/2 way there for most of our strengths, and even more will have already reached that level.

A word of caution though – just because you’re an expert, it doesn’t mean that you can stop learning. Ericcson and company have gone on to find that experience does not account for accuracy in blind tests, and that expertise declines with experience if you dont maintain training. That last part I dont fully agree with though because I believe you can become more of a niche expert instead of a well rounded expert. The example they gave is a physician having a hard time diagnosing unusual diseases – but I guarantee they can diagnose the things they see daily faster than those who can diagnose the unusual.

Part 2 will cover more in depth on why to be an expert, and traps to watch out for.