Beating the demo gods
Despite the best efforts of the infamous demo gods, sometimes you can come out on top.
The Microsoft evangalist Scott Hanselman is renowned for his demos of technology. As well as being an incredible presenter he is also known for the level of preparation and planning he puts into them. Not only does he put in the effort practising, he tries to prepare for every eventuality. He has had demos where his main machine has failed and has had to move to a backup machine that he had prepared just for this eventuality.
In the book Fall in Love with the Problem, Not the Solution Uri Levine describes preparing for a meeting with some investors when he was pitching the navigation app Waze. He knew that it was not yet as good as he hoped but he realised that the investors would most likely ask to navigate to their own houses. With this in mind, he got hold of their home addresses and made sure that the navigation worked correctly to their houses and that additional information was available in the neighbourhood.
Sure enough, the lead investor asked to navigate to his house and they subsequently invested.
I had a similar situation early in my career. My first proper role was with a company that produced systems for public libraries. These were deployed all over Europe. This was in the days of DOS where most things were text based and doing graphics was possible but difficult.
We were asked to produce a demo where a user could query the platform for where the book they wanted was on the shelves and we would show them on a map where it was located. I wrote an initial version of this and then did a flying visit to Rotterdam for it to be demoed to senior librarians from around the country.
I flew in on the morning of the demo with the software on a floppy disk - this was in the days before the Internet took off. We installed the software and tried it, only to find that one of the requirements they had given me was wrong and it would not work.
The librarian who was helping me saved the day.
She worked out that the head of the library had written a book and that he would probably pick his own book to show where it was. So I spent the next hour making sure that the only book it worked for in the whole library was his book. We didn’t have the ability to tell him in advance so we just had to hope he would pick it - and he did. The demo was a success.
Off the back of that we wrote a full 3d model of the new 6 floor library that was being built in the Hague in Holland. This would not have happened without the inside knowledge of that librarian, being able to fudge the demo and the natural ego of the senior librarian.
Sometimes it is possible to beat the demo gods … or at least claim a draw!
Additional note
As a follow up, the system we ended up writing for the library in the Hague was really cutting edge at the time. We used a library called Renderware to produce a full 3d model. Each component of the model was modelled point by point - I had to manually work out the coordinates of each point in each component. In addition, depending on the order you defined them, the surface you created would either be transparent or solid.
Once the shape was in place we had to produce images for each of the surfaces to make them look like the bookshelf or desk they represented. Some of the models were quite complex including a piano, a spiral staircase and even a train that had childrens books in the back.
Once we had the model, an unfortunate “volunteer” had to go through and label each model with what books and media they contained so we could map the books in the catalog to the locations in the library.
Once we had done all this a user was able to ask for a book and it would then fly them through the library from where they were to where it was located in 3d. This was written in about 1995! Unfortunately I never got to see it in use in the library itself and I have no idea how much it was used.
Links