A little more than two years ago I was talking to Vijay Pande, a General Partner at Andreessen Horowitz, a professor at Stanford, and in his own right a great entrepreneur and thinker. We were discussing a “Grand Challenge” problem in healthcare to reduce type II diabetes in people by 50%. In that discussion, we outlined how a great breakthrough venture might work. In fact, many of the world’s greatest ventures followed a temporal pattern.
If you realize this is the likely pattern you’ll follow, it will give you more confidence to continue developing your idea. More than that, this pattern will give you deep insights useful in continuing the development. It goes like this:
First, when you have a truly breakthrough venture concept and begin to consider starting your venture, you will want to talk with some of your closest friends, peers, and industry experts. The more transformative the concept is, the more likely you are to hear from a good fraction of the people you talked with:
Stage 1: “What you are thinking about is crazy. No one in their right mind would even consider it.”
This response is natural. As I mentioned in previous posts, you’re going against conventional wisdom or even common sense. Your greatest challenge is to hear the critique and analyze it. Deeply understand it. Forget that someone of authority, knowledge, or influence was expressing a negative opinion. Turn that negativity into a statement of real value that gives you insight into your venture.
For example, when we first presented Siri as a concept to a close friend and great entrepreneur, his reaction was “no one will ever want to talk with a computer.” We left, somewhat despondent, and analyzed his concern. He caused us to think deeply about his critique. Was he right? Sometimes, for sure he was right:
- We hate making phone calls with a call center computer response to guide us through a maze of options
- We don’t like robotic voices
- We have an immediate and constant urge to talk with a human because of its inferior functionality
But what if Siri didn’t guide us through a maze but instead was conversational?
What if you asked an intelligent assistant for information and it provided immediate answers rather than providing links to be clicked on?
What if Siri had a human-like voice, even having a sense of emotion or humor, for example?
And what if Siri was so knowledgeable and efficient that, for our purposes, it was even better than talking with a human? And so we continued to develop Siri and incorporated these issues into our plan.
That brought us to the second stage of a breakthrough venture: we talked with some of our closest friends, peers, and industry experts. Many said:
Stage 2: “What you are thinking about might not be crazy, but it’s impossible.”
- No one has ever built a computer system that could have a conversational interaction.
- All computer voices are robotic.
- Computer systems will always have inadequate knowledge and be inferior to a human.
Some of the greatest researchers at SRI could demonstrate an interaction between a computer system and a human in the form of a conversational query and response.
Other researchers were able to demonstrate that a computer system could have a conversational tone, including the “prosody” or “sing-song” nature of a human voice.
And lastly, it was true that computer systems would almost certainly have inadequate knowledge if the query were unrestricted.
And so we continued to develop Siri and decided to limit the queries that Siri would have good answers to in the areas of travel and entertainment.
This brought us to Stage 3 of a breakthrough venture. We talked with some of our closest friends, peers, and industry experts.
Stage 3: “What you are thinking about is possible, but no one would invest in it since it’s much too difficult and expensive.”
We had worked for over 20 years on solving the difficult problems from Stage 2. We brought together a world-class team of researchers and engineers.
The technologies for conversational interaction required recognizing not just the words that were spoken but also recognizing the intent of the query. Then, knowing the intent, the system could identify the required web services to answer the query, and then construct the answer from those web services.
It would be possible but difficult to construct a voice for Siri to respond to human queries. We decided that the first Siri product would provide written responses to the queries.
We built a demo of how Siri would interact. We allowed users to ask questions in the space of travel and entertainment and showed Siri’s response. The demo proved that Siri could be built with our technology.
This brought us to Stage 4 of a breakthrough venture.
Stage 4: “What you are thinking about is possible and can be done with a great team within a few years and limited dollars.”
This was great feedback. We put together a plan for the development of Siri and brought it to multiple potential investors. The first Siri product would be ready within 18 months and less than $10M. Great investors including Gary Morgenthaler and Sean Carolan believed in our vision, our team, and our plan.
The point of this discussion is to recognize the stages that you will have to go through to create a truly breakthrough company. If you understand the stages, you’ll be much more likely to persevere rather than succumb. As a human, your DNA is tuned to make you want to be part of the social, family, and friends fabric of your world. Opposing this is hard to do.
More importantly, by going through these stages, we learned the issues of critical importance to creating our product and our venture.
And Siri was born.
