Hardware · Data center

How do you break through sales for enterprise hardware?

Mario Facusse Founder and CEO of Xyber Technologies

March 10th, 2015

I created a server with a cooling system that requires no power and has no moving parts. I was able to invest in the concept and create prototypes. After an angel investment, I was able to extend the concept to the rack as well. Unfortunately, I got sick and was MIA for four years.

Now, I have prototypes with aging components, very little money, and angel investor doing his best to help. Our goal is to license the technology and we've met with OEMs, data center operators, cloud service providers, etc. The reaction we usually get is: "That's not possible, cooling without fans..." or just absolute silence. We show them the lab data, thermal and CFD analysis, thermal videos, etc. We even got FCC, UL, CE, and CSA certification for our design and prototypes. Yet, they never give us a chance and test a unit in their environment.

We have been awarded seven patents worldwide and more are in progress. We have a great solution to an exceedingly difficult problem. How do we get anyone to pay attention?

Every business needs an eye-catching website, and building one shouldn’t be expensive or time-intensive. Learn how to set up a Wordpress site, pick a stunning theme, add a blog and newsletter, sell products, and monitor analytics — in just one hour.

Travis Russi

March 10th, 2015

To state the obvious, you are either 1) trying to sell to the wrong customer or 2) not communicating the value in terms the customer cares about.

You need alignment of value proposition with the target market. I'm sure you think you do (doesn't everybody want to save money?!), but, based on the lack of sales, you don't have that. 

Re-think who your customer is (non-trivial; think outside the box; think low-hanging fruit).  Once you accurately define your customer, nail the value proposition as it relates to them.  Each customer type will have a different value proposition.  Once you get a message that resonates with a target market, find force multipliers to get your message in front of them as quick and cheap as possible.

Do some research into 'customer discovery' to get more insight into how this process works.

Rob G

March 11th, 2015

"our system would cost approximately 20% higher that comparable servers".  Are you selling serves or a cooling system?  i guess i misread you post - you did say "i created a server with a cooling system...".   So what you are selling these DC operators is new servers not a new cooling system.  This complicates the sale by orders of magnitude.  Risk is a huge issue.  Is there a way you can separate your cooling system such that it can be purchased separately, preferably by the DC operator then later by the server manufacturer? 

Rob G

March 10th, 2015

Something doesn't add up.  What does the cost/benefit analysis show?  What does the risk analysis show? Obviously cooling is not the only cost to operating a data center, but i'm sure it is a significant cost.  Cost reduction projects aren't typically as attractive as revenue generating projects (in the eyes of the C suite), but if the costs and risks to at least trial your solution are low and the savings high then there is no reason why you should not be at least getting trials.  Trials lead to sales.  How much time and money would it cost to trial run your solution and what are the risks (to the customer)? I could see a customer hesitating simply because they don't want to take the time and $$ and risk on a small company that can't afford to go into production even if the trial succeeds, but if the solution is compelling enough they will figure out a way to work with you.  Assuming the cost savings are some significant % of operating costs and there's little you can do to affect that (i.e. you can't eliminate more than 100% of their cooling costs) then the focus comes down to cost and risk.  Get those numbers low enough and something has to budge.  If the cost savings just aren't compelling enough then you need to look at other benefits.  Some DC operators (goodle, apple, MS come to mind) may pay attention if you can show a compelling "green" story, i.e. reducing greenhouse gases or some other environmental benefit. 

Mario Facusse Founder and CEO of Xyber Technologies

March 10th, 2015

Research shows that on average 30% of the OPEX of a data center is cooling and 30% is servers (estimates based on averages). Our data, verified by an independent lab, shows that we can cut the energy consumption of the server itself by up to 30%. Furthermore, our system does not require cold air, just air flow so we can cut DC cooling costs significantly, up to 50%.

This equates to a total reduction in OPEX of up to 25%. In our current short-run manufacturing capacity our system would cost approximately 20% higher that comparable servers. Even taking into account the higher CAPEX, the ROI would be less than one year. We have an MVP for trials and yet we keep running into this inertia.

We have met with some of the companies you've mentioned and our point of failure seems to be the fact that we don't have complete racks of systems for them to try, just individual servers.

I think we have a strong financial argument, and a rather compelling "green" story. Am I not seeing the problem correctly?

Mario Facusse Founder and CEO of Xyber Technologies

March 10th, 2015

Well, at least we weren't looking at it incorrectly. Travis that's exactly what we thought. We hunkered down and realized that a potentially niche segment that could use our tech were Bitcoin mining OEMs, reworked our message to reflect the advantages and went off to the races. We met with two large players in the space and were excited with what we heard. Fast forward a couple weeks into the conversations and one of them (here in Austin) declared bankruptcy and the other (in Europe) bought an oil submersion company to solve their issue.

Now we're back to square one and the frustration, as you can imagine, is mounting.

Joanan Hernandez CEO & Founder at Mollejuo

March 11th, 2015

Hello,

The complete initial message of this thread fit perfectly with a crowdfunding campaign. I'm not saying is easy, but the basics are there.

Now, is it to difficult to sell by yourself this solutions instead of relying on big business?

Maybe if you adapt it to be compatible with Open Compute other doors will open. Just a comment.

Good luck!

Cheers!

Mario Facusse Founder and CEO of Xyber Technologies

March 12th, 2015

Joanan, we are actively working on the joining open compute and share some designs to try to spark some community engagement. Thank you for the suggestion.

Mario Facusse Founder and CEO of Xyber Technologies

March 12th, 2015

Rob,

the system is based on a complete redesign of traditional server chassis. We can adapt to pretty much any electronic component but selling as an after-market add on is not an option at the moment. I agree that this is a difficult task. Which is part of the reason why our business model is to license the IP.