How To Fail in Your Startup Business – The Data Center Failure


so the first mistake that you can make: I
see a lot of people writing down a good business plan with all the details on the market on
the business on the technology with actually no idea of what they are talking. This is
actually a story for raising money because most of the people and actually me too maybe
i wasn’t realizing that well at that time I was thinking that writing down a good business
plan was a good story for getting money and once i had money it was easier to success.
I know it sounds crazy I m gonna tell this story about the data center failure we were
PhD students and the summer of 2015 23rd of June we decided to open HotBlack Robotics
SRL we were working as Phds in Telecom Italia and Politecnico di Torino in Telecom Italia
(TIM) they have data centers. Data center is like imagine a room full of computers with
your very important data and when you put a lot of computers into a room the temperature
is getting very high and you can risk to damage your computers so your data and that’s very
risky so what they are doing usually it’s putting a cooling system to cool down the
temperature and maintain all the structure safe and your data safe. The problem is that
with no monitoring system of the status of the infrastructure you consume a lot of money
and we had a real test case in Rozzano near Milano and we calculated that 8 Millions per
year were spent for the cooling system and 60% of this amount of money could be reduced
because of the i mean reducing the unefficiency having a good monitoring system. In fact so
far some of the top technology in the field they provide monitoring system with a technology
called Wireless Sensor Networks. Those are sensors in any computer and into the room
they build a mesh of sensors in order to give a real time monitoring system to tell the
energy manager how much they are consuming how much is efficient the structure in that
particular zone or where are the unefficiency and so on. This technology is very cool it
s the leader in the market but it has the problem of maintenance because can you imagine
a huge room with 1000 and 1000 of sensors and you need to find a way to maintain all
this sensor mesh in a perfect way. So our idea was to have one robot with one sensor
going around and taking the monitoring measurements and building the same and actually more efficient
3D map for monitoring the status of the data center. So the robot were going around with
thermal camera and thermal probes and creating the same energy monitoring map. So the first
mistake that I made was to think that Wireless Sensor Networks has a real problem of maintenance
and I was closed into my lab thinking “of course 1000 of sensors they would have a maintenance
problem and my robot is gonna be the right solution” while the market is a lot different.
Second huge mistake was to develop the technology straightforward so instead of looking at the
market and taking as assumption that one data center in Rozzano was enough to validate the
market we deep down so bad building this robot that was going around and making thermal analysis
and it was working! And what happen?It happened that I wrote down a business plan the third
mistake that i made was that instead of knowing by person energy managers and customers and
knowing exactly how energy manager works and how the data center world works I thought
that was enough to googling and I said of 48 data center in Italy it s fine and googling
again I found out that the energy problem it s very important so I thought “ok it s
written in Google so it s true” and then I ve learned also that they were thousands of
data centers in Italy so I thought “wow it s a huge market and Im gonna write down my
business plan” then I made another mistake that was to write down in details the technology
about to use and the revenue model how to sell this product to the people without knowing
actually the people I was just knowing one guy!It was ok for me. What I find out? That
when I started to make a deeper market analysis I found that the data center world it’s huge
and my assumption that one data center is the same as 48 in Italy and other thousands
as internal data centers was completely wrong. I mean you think “ok I’m selling to one data
center and it’s the same for everyone” it’s a huge assumption and it’s not concrete actually.
The data center world is a world it’s a huge world and it works like this they have services
of housing hosting and cloud that more or less are the same big providers but knowing
a little bit better energy managers and speaking especially to Aruba Aruba told me “look what
you’re saying it’s very strange because we in Aruba redesign everything we redesign the
entire infrastructure to make it more efficient and to avoid services of energy monitoring
because it’s a cost!” I thought “Ok I’m f***” and then I decided to go to internal data
centers so like for example (big) supermarkets that are holding data centers into the company
or General Electrics that they have their own data center but they answered speaking
with the CTO of GE he told me well look it’s a problem we know it in fact we are building
a new data center centralized worldwide and make it more efficient because it’s a new
infrastructure so ok and the third one was audit that was a perfect suite for my robot
those are guys that are going around and measurements they check the status of the efficiency in
any infrastructure they are working also in the architecture field but it was a small
market. So in conclusion I mean making too many plans it’s not working I mean you really
need to know your market in deep and make friends in your market and know everything
about your customers and what they are doing so next video it’s how to understand a problem
and how to find a need!

, , ,

Post navigation

Leave a Reply

Your email address will not be published. Required fields are marked *