PrEtotypes vs PrOtotypes

What's the difference

The term "pretotype" was first mentioned in Alberto Savoy’s book "The Right It: Why So Many Ideas Fail". I will review this work sometime in a separate post. For now – about the difference between the prototype and the pretotype:

• A prototype is when we have a very stripped down but working version of something. What startups call "MVP" is almost always a prototype of a future product.
• A pretotype is when we don't have anything at all, even a stripped-down version of the product. Pretotype is from the word "pretend", that is, we only imagine that we have something.

Prototypes help determine the market need for a product and get baseline metrics at the lowest possible cost. The later we find out that no one needs the product, the more expensive such an experiment will cost.

Real life examples

Palm

Jeff Hawkins, the creator of the popular Palm pocket computer, had a wooden plank in his pocket for a long time. Whenever there was a need to make a note, look at a calendar or check an email, Jeff would pull out a piece of wood and imagine that he was using the future Palm. So the creator was able to test the main convenience hypothesis "to get something out of a pocket to execute a specific scenario". Some scripts were handy and later entered the first version of Palm, and some disappointed Jeff, which saved him a lot of money in creating the prototype.

McDonald's

When McDonald’s plans to add a new burger to its menu, the company is in no hurry to spend millions of dollars on real recipe development, writing technological processes for cooking, etc. To test the main hypothesis "Will people ever buy this?" McDonald’s simply adds photos and descriptions of the new burger to a menu of several dozen restaurants. If visitors order this position, the staff apologizes, says that at the moment there are no relevant ingredients, and the resulting order statistics fly to the McDonald’s office, where managers finally whether to launch a new burger or not.

Bookstore

Norwegian girl Anna is a real bookworm. One day, Anna decided that the Oslo area she lived in needed a bookstore. To save money on testing this hypothesis, the girl made her pretotype. On the ground floor of a suitable building, she rented a small room for three days, hung a handmade sign at the entrance, put some advertising signs on the street, closed herself inside an improvised store and started counting buyers, who approached the pretotype and tried to open the door. In 3 days only 34 people were interested. Having estimated the economics of such a business, Anna realized that it was not profitable to open a bookstore in that place, so her hypothesis was not confirmed.

How I use pretotypes

In some cases it is impossible to create a good pretotype at all. In other cases it is more economically feasible to start with a prototype from the very beginning. But, nevertheless, I always study the possibility of creating a pretotype first and only if it is impossible, I move to prototyping (MVP). Several examples:

Vivafit

When we started Vivafit, even before we wrote the first line of code, we were able to measure the interest of the potential audience, gather early followers, determine the region to run the MVP and the optimal subscription price.

To do this, we made a cool landing, which told about the product as if it already existed and was successful. There was a "subscribe" button, next to which the price tag with the value taken from the active A/B-test. When a visitor decided to buy a subscription, he had a dialog in which we gently explained that Vivafit had not yet entered the market, but was about to release its Beta version. The dialog further suggested leaving an email where we could contact a potential user and notify him of the product release.

By analyzing the conversion by region and A/B-test prices, we were able to approximate the above metrics. By the way, after the launch, most of the data obtained this way was confirmed. In the later stages of Vivavit, unfortunately, was not a successful product, but it is a topic for a separate post.

ESP

About a month ago we made a soft launch of English Speaking Practice – application for training of spoken English. The idea of the product is very simple: ESP invites a user to talk in English on a random topic, after which AI analyzes the resulting speech and sets grades on several criteria: fluency of speech, vocabulary, lack of accent and grammar.

However, we do not yet have any real AI (OMG, who would have thought!). We just imagine that we have it. Estimates are calculated on the basis of various behavioral factors (the length of speech, the evaluation of the user, etc.), rather than expensive in the development of machine learning algorithms.

If the basic hypothesis is confirmed and users are really interested to develop their  spoken English in this way, we will certainly make a real AI (R&D in this direction is already underway). But we managed to avoid spending a lot of money at the start. And all thanks to the approach of pretotyping 😉

So guys: