EFTA00711193.pdf

From: "Sinofsky, Steven" CIIIIIIOMIUMINIMPIIIIM To: Jeffrey Epstein Subject: tapping into software engineering / product development Date: Fri, 05 Jul 2013 19:42:47 +0000 J, Some thoughts on how to build out software assets in a country in a way that does not set you up for failure that has been seen so many time. Many places work hard to invest in capital to recreate silicon valley. This is not unlike the classic "industrial park" model -- build out a center, provide tax breaks, hire a few experts. The same thing goes for the universities that create extension offices in other countries. These all have the same failure points which is that they tend to undervalue the intangibles. This is a good article on silicon valley and how all the other places failed trying to recreate it, even within the US http://www.technologyreview.com/news/516506/silicon-valley-cant-be-co ied/ Software has a virtue of being able to scale as a global workforce to some degree. Israel has a strong locus of software and engineering people (hardware as well). But the market in country (and in Hebrew) is so small that they realized early on that they would never create a substantial business within the borders of the country. That was a wise observation as most countries see their home markets as "large" in a myopic way. Israel wanted to generate "outsized" returns on their investment in engineering. This makes sense because engineers are still a scarce resource and even for Israel they could always pay salaries and put them to work for IDF projects. So what Israel does (which is different than say UK, France, Germany who see their home markets as huge) is support people starting companies and going into software / hardware engineering at a few universities--much like the old industrial park model. BUT, they do one of two things differently: • For homegrown new companies -- that is places where the idea for the product is one they think up -- they immediately upon founding the company create a sales/marketing/product (and in fact the CEO) office in Silicon Valley. Essentially from the start the company is a multi-national and is guided by leadership in the US "ground zero" for the industry. These people are from silicon valley. This is a tough model and requires a lot of relationships but has worked for Israel. • For employing today engineers -- that is when you just want to invest money and employ your own engineers there is a second model. In this model you make connections to existing (new or mature) silicon valley companies and your engineers serve as engineers within these companies. Essentially you are outsourcing your work to US based companies. This is currently what we see a lot of from Czech, Hungary, Poland, and to some extent Berlin. This model does not generally have equity participation from the host country but it does start to create a network of engineers in the country and people in university start to see options in their home country for doing good work and having a connection to all they read about on the intemet. To get started on the second one you need simply have a locus of engineers and a connection to the US. What I see all the time are emigrants to the US who look to their own school friends and hire them. In just the past few months I've seen companies in the valley hiring most/all of their engineering from Poland, (old East) Germany, and even France. In all cases, the founders of the company are in the US and they run the product definition from here so the product is right/ready for the US market from the start. They focus little on the home market initially. That is key to getting attention, funding, support from the silicon valley world. EFTA00711193 Things are really set up for this now. For example, we see tons of companies that have apps for iPhones/iPads and need Android versions. The way things work technically it is just as easy to have a whole separate team do the Android version. being able to have an entire team just "copying" what is going on from the Valley team on iPhones is a big positive. I know a lot of companies that would hire an existing Android team right away if they could--or hire an iOS team to maintain their iOS version while they make the new Android one from here. There are only a couple of requirements: • Excellent software skills on modem technologies (open software stack including: http://en.wikipedia.org/wiki/OpenStack, PHP, Ruby, HTML5, and especially building apps on Android (first) and iOS (second) • Strong infrastructure (fast, reliable, secure intemet connectivity to the US west coast) • English skills (reading and writing proficiency, with verbal communication over Skype or Google Hangouts) My sense is there is a large frustration with India on at least two of these (skills and infrastructure). And the cost advantage is no there as well. In fact, cost is not a primary driver but if you're going to pay the same as the US then you at least want the same level of work. In many ways that is why the Chinese, Korea and Japanese companies fail as well -- they are different culturally and linguistically and by virtue of their market size they focus on the homeland first. It is why you almost never see a software product start in those countries and move outside of them, even if they make products in country which gain popularity in a large scale way. But you see the reverse all the time (Facebook, Instagram, Twitter, and more). Let me know if this makes sense. This doesn't employ 1000's of people but it does employ people in software. Steven Sinofsky Executive in Residence Harvard Business School ' Field Road EFTA00711194