Creating a craze

by Peter Small


The first time I tried lookig at business in terms of biological strategies I had an amazing success. It resulted in a craze that swept Europe. This sounds as nutty as a fruit cake, but it actually happened this way.

In the world of an entrepreneur, random events and coincidences are as likely to lead to a profitable business as any rationally thought out plan. In 1977 I saw a hippie selling button badges in the Portobello Road market. Only a week later, I was at a trade show and saw a stand selling very cheap, button badge making machines. The technique was very simple. It needed a circular punch to cut out a picture and a special press that clamped a circular metal frame around it. Hey presto! A button badge, for just the few pennies it would cost for the components. Here was the basis of a business venture. So I bought a cutter, one of the presses and a few hundred components.

It was easy to make the button badges, but, I was faced with the problem of what to put on the badges. How could I know what kind of badges would sell? It was then that I thought of the way in which evolution finds successful genes: it tries a whole lot of them out and lets the results separate the winners from the losers. With this thought in mind, I used the cutter to cut out a random selection of pictures from various magazines. By putting them all on a board, and observing which sold and which didn't I could use this as a guide to making more. Those that didn't sell I could throw away.

I'd arranged with a friend of mine who had a shop in Carnaby Street to let me put a badge board outside his shop for a percentage of the takings. Pretty soon, this board was acting like an intelligent marketing device: showing me what kind of badges were selling and which were not. It was so successful that other shop owners started to notice how my board was always surrounded by customers and were asking me if I could supply them with a board and some badges.

Soon, with a wholesale trade increasing, I stopped using cut outs from magazines and started using a photocopying shop around the corner to print many copies of the best selling designs and employed outworkers to make up the badges for me.

Even with the photocopier and the outworkers, increasing demand was exceeding my ability to supply and I was forced to look for a new manufacturing solution. It was then I noticed that in some of the stores I was supplying there were a few button badges that had been around since the craze on the Beatles - some ten years previously. These were similar badges to mine except that they'd obviously been mass produced.

I tracked down the supplier and found it to be a company whose main business was making ornate metal cake tins. They'd perfected a technique of printing onto enamel coated steel sheeting and besides pressing these into cake tins, they'd branched out in the 1960's to use the technique to make button badges for the Beatle mania.

It was a solution to my problem. They had the facilities to produce any number of button badges I could possibly need. The snag was that their technique and machinery was based upon printing very large sheets of steel - which printed 340 badges on every sheet. As it was not viable to print less than 100 sheets at a time it meant that my minimum order had to be at least 34,000 badges.

A few calculations told me that I could sell this number of badges but the problem then became which designs to choose to be printed. My evolutionary system was great for responding quickly to customer demand, but, as this demand was constantly changing with the rise and fall of the popularity of different pop stars, it didn't seem to offer a solution to help me predetermine an advance order of many thousands.

I then thought of the way in which evolution works by putting many genes into an ecosystem, letting the unsuccessful die out and the successful survive and multiply. How could I apply that principle to a button badge making system? It then occurred to me to think of each button badge design as a gene. If I printed 340 designs on a sheet, it would be the equivalent of testing the market with 100 each of 340 genes.

As this method of manufacturing was cheaper and there was a high markup on the cost to sale price, I figured I could work like nature and allow for some redundancy. A back of the envelope calculation told me that the business would still be viable if the manufacturing costs were doubled. This would allow me to work with a fifty percent redundancy rate. In other words, if like nature I created a surplus and allowed for half of the badges to die (not sell), I could still be profitable.

Effectively, this allowance for waste halved the number of badges I needed to sell. Instead of having to sell all of the 34,000 badges in a print run, I'd only need to sell 17,000. As by this time I was selling around two thousand badges a week, it meant that within two months I would need to print more. Looking at this in terms of an evolutionary strategy, this second print run would in effect be a second generation.

By copying nature's strategy, I could arrange for this second generation to repeat on the designs that had sold from the first print run and leave out the others. This equates with the strategy of nature, which increases the successful genes at each generation while leaving out those that are less successful. This could be repeated over any number of new generations (print runs), ensuring I always had a regular stock of the best selling badges.

Certainly some of the designs turned out to be duds; sometimes best sellers were over estimated in some generations; but, overall this method of working in generations and adjusting the designs to respond to the constantly changing demand for button badges was surprisingly efficient. There were many badges that remained unsold but these turned out to be far less than expected such that this method of creating a "living" and "adapting" product resulted in only a ten percent increase in costs through wastage. This was easily absorbed by the profit margin.

Another advantage was that it was easily scalable. I soon realized the potential of supplying shops all over the country, I could use exactly the same method of production but increase the frequency of the print runs.

There then arose the problem of getting other shops to sell the badges. How could I persuade shops located miles from London to take a board of badges, when they hadn't been to Carnaby Street to see how successfully they were selling? Again the biological strategy of planned redundancy came to my aid. I realized that I could give shop keepers a board full of badges on a sale or return basis - allowing them to pay only for those that sold and returning those that didn't.

At the beginning I had a lot of returns, but quickly re-arranged the manufacturing to match the sales so that these returns became less and less as the contents of the boards became more efficiently stocked to cater for customer demands.

I'm sure a conventional business marketing strategy might have seen this in quite a different light, but by likening all the boards to living organisms made up of genes, I had a similar strategy to Mother Nature - who tests her genes out by averaging the effectiveness of each gene over a number of organisms. I know this sounds bizarre, but it gave me a very clear conceptual model that made a lot of sense and was actually working.

My printing schedules were the equivalent of Nature creating new generations of genes. These kept all the badge boards alive, adapting them to continuously changing trends in much the same way that nature maintains the survival of living organisms by keep changing their genes to allow them to cope with changes in their eco-systems.

Once this system was running successfully, I came to realize the value of another of Nature's clever tricks: the introduction of random mutations. Effectively, the whole system - of supplying badges and taking back the returns - was acting as an excellent information system. It was telling me what badges were selling and which weren't. This meant I could introduce test designs into my manufacturing cycle at each new print run and be able to quickly determine how well they were selling. Those that sold well I could repeat. Those that didn't I could afford to throw away. It was the same strategy as I'd used with the first badge board, but now being applied to a complete system.

It occured to me that Nature doesn't work out what new gene variations to try. She just introduces random changes. This gave me an idea. I went along to a company that supplied stock pictures to newspapers and magazines. They had hundreds of thousands of photographs of pop stars. I explained to the owner of the company that I wanted to use some of his pictures for badges but I had no idea which would sell and the only way I could find out was to try them out.

I proposed that instead of paying for selected designs, I gave him a regular sum of money each week to be able to experiment with his photographs to find out which worked and which didn't. This arrangement was accepted, so, for a relatively small overhead, I had access to a large gene pool of button badge designs that I could regularly introduce into my evolving system - popping a few into each new generation (print run) to test them out. Those that proved successful I repeated in further generations.

This simple application of nature's evolutionary strategy allowed me to create a successful and efficient business that fed a runaway craze with the utmost of simplicity and efficiency. That happened over a quarter of a century ago. I've been using variations of this strategy ever since. Look at the example of the Kempelen Box in the research section of this site. Do you recognize the imprint of this strategy there?

Perhaps you won't. At the time I was creating this business, I tried to explain what I was doing to my friends in the pub where we used to meet in Carnaby Street. They just thought I was weird and asked me what kind of dope I was on.