Big impact innovations start as a small idea, a necessity, or a trivial situation. Imagine that, after receiving the compliments about the quality of the dessert you prepared for your guests, you wonder if the secret of how to make these pastries could please many more people and, who knows, serve as the basis for a profitable business. In other words, you want to know if it is possible to scale up the small experiment you carried out with your guests.

The first risk is that the result of the experiment is a false positive, that is, a misidentified effect that does not really exist. Could it be that what seemed to you as an approval of the quality of the pastries by your guests was just them being polite, knowing how proud you are of your culinary skills? It may also be the result of the so-called confirmation bias, which makes us give more importance to information that confirms our beliefs and less importance to the guest who was less enthusiastic about the quality of the pastries. If the repetition of the tasting experiment produces very different results, don’t be surprised, it happens to the best. Many results of  scientific studies were proven to be difficult or impossible to reproduce. A famous study tried to replicate experiments, published in relevant academic journals in the field of psychology, and only 36% of them gave the same results.

If you verify that the result of your experiment is not a false positive, you still need to confirm that the people involved adequately represent those you need to expand your idea. Will your pastries please in the same way people who live in another region or country? Will you be able to get help in the confection of the pastries to produce them in greater quantity? Remember that, in the case of teaching, many educational experiments fail to give good results when they are generalized to the entire education system because, generally, the teachers who volunteer for the experimental phase do not have the same characteristics as an average teacher.

The other validation you need to do is whether the circumstances in which you developed your experiment are generalizable. For example, your pastries were prepared just before your guests arrived. Will they have the same approval if they had been prepared earlier? What if I have to transport them from the cooking place to another one further away? This validation can lead a chef known for his elaborate dishes to decide not to open a chain of restaurants with his name, as he cannot guarantee the production of his dishes with the same quality. That’s why most restaurant chains ensure uniformity of quality by using simple recipes with standardized ingredients.

There is also a risk associated with the repercussions of expanding your experiment. These problems are difficult to test because they only occur after you scale up. For example, part of your product’s success may be due to the fact that it is handcrafted and challenging to acquire. Its greater accessibility may associate it with another category of products that are more common and less attractive to the audience of your original experiment. The expansion can lead to the emergence of competitors with similar products, increasing the offer and devaluing your pastries. 

There may be a risk that expanding your experiment will increase the cost of making your pastries. This could happen, for example, if its unique flavor depends on the use of unusual ingredients. This feature will make it difficult to copy the product, but expanding your business will increase demand for these ingredients and, consequently, their price.

Reducing the five scaling-up risks that I have described is essential for the second stage of startups when they are called scaleups. The knowledge of these five risks is useful for entrepreneurs and anyone who needs to decide whether or not an idea or activity is worth continuing or expanding. Many public policy proposals are based on the results of small experiments, without solid validation on whether their assumptions are generalized. Ensuring that ideas are capable of being scaled up is one of the goals of the book “The Voltage Effect”, by John A. List, a professor at the University of Chicago, where he states that there is no single recipe for the success of an idea, but that it will be difficult to achieve it lacking one of these five ingredients.

Returning to your idea, we can now assure you that, 185 years later, a similar idea had a huge impact. It remains to be seen whether the creators of pastel de Belém adequately assessed all the risks or if they simply did not think about it. In this case, the Anna Karenina principle may take the form: all ideas that have had an impact are equally good; every idea that failed, failed in its own way. 

Adapted from my article in i newspaper of April 26th, 2022