Based in New Haven, William Speirs is a lifelong entrepreneur and currently the founder and CEO of Educated Solutions LLC, the company that launched Metrink (now in beta). He attended Rensselaer Polytechnic Institute as an undergraduate and received his PhD in Computer Science from Purdue University. His specialties include computer security and software architecture. William is also an active member of the Apache Software Foundation.
Metrink is a suite of monitoring tools that allows system administrators to easily pinpoint infrastructure issues, from hardware to software, resulting in decreased server downtime and improved network performance. Watch a video about Metrink, follow them on Twitter, and find them on Facebook.
As a part of the New Haven leg of the Startup Roadshow, we talked with William about his recent experiences and challenges in launching Metrink.
Why did you start Metrink? What problem are you solving, and for whom?
I started Metrink because I was frustrated with the monitoring tools I saw on the market. There is currently no single tool that pulls in all the information about your infrastructure, from hardware to software, and helps you find the root cause of an issue quickly. Plus, some tools are better for networks while others are better for applications.
Metrink helps server administrators answer the question: “What is wrong with my infrastructure?” Many of my competitors have dashboards to monitor network systems but they don’t help to answer that question.
There are a number of commodity monitoring tools that will show you basic dashboards with metrics that pertain to CPU, memory, disk, and network performance. With some of them you can configure an alert that will trigger when a metric is above or below a threshold. However, these tools are only useful if a metric isn’t supposed to go above or below the threshold at all times of the day. Without being able to specify different thresholds during different times of the day, you’re forced to set the threshold to the absolute maximum or minimum value.
Our clients are server administrators from a variety of industries. Ultimately, if you have more than a handful of servers, you can use Metrink. We also white label the product to segments like bandwidth brokers, agents who sell bandwidth wholesale, and hosting providers. These types of clients are interested in Metrink because they can expand the service they offer, and make more money in the process.
What stage of development is the product in, and what’s your key differentiator?
The product is currently in beta. We have about 5 beta customers, with some potential leads on paying customers. We are in the process of finding more beta customers and looking to convert them to paying customers by the end of the summer.
Our key differentiator is the use of Artificial Intelligence to automatically detect anomalies in metrics. This allows Metrink customers to not have to specify a static threshold for their alerts, but instead be notified when something anomalous has occurred. I should say that the AI component is in its very early stages, but it is like nothing else currently on the market.
The real value for any product comes down to its impact on the bottom line. As an example, one client told us, “Every kilowatt you show us being used provides us an opportunity to save money.”
What’s next for Metrink?
Anything that produces a number or metric is in scope for Metrink. By providing analysis on these metrics, we enable business leaders to make data-driven decisions. We will continue to build the product to serve the “big data” push.
We are also looking to move further and further up the stack to provide more detailed information about why an application is slow. One idea we have is to partner with universities to provide our tool free of charge to students to aid them while they learn to program. The hope being that when they graduate, they will ask for Metrink at their job. But, this is a long burn.
What are your biggest challenges?
Our biggest challenge is bringing differentiating features to the market. There are a number of open source projects and commercial products that provide similar functionality. With such a fragmented market, it is hard to get the attention of system administrators.
Hindsight is 20/20. We spent a lot of time building the product on our own, but we should have spent that time building it with the client. We’re now focused on answering the questions, “How do you fill the sales pipeline?” and “How do you make that a repeatable process?” I wish I had done more to answer these questions earlier on, but I only left my full-time job three months ago, so I feel we are moving quickly.
We’re moving fast and failing fast. But that’s the evolution of a startup. If you’re not failing, you’re not learning. One such failing is that as the CEO, I’m spending too much time on the technical side. While programming is my bread and butter (I have a PhD in Computer Science), it’s not the best use of my time. I need to be focusing more time on customer development and learning.
What advice do you have for first-time early stage entrepreneurs who want to break out of their current day job to start something?
Do as much as you can before you make the leap and focus on your core competency. If you enjoy building tools and playing with technology, but that is not your company’s core competency, you’ll need to examine closely where to spend your time. If you’re spending lots of time writing lots of code, that’s not a good use of your time. Customers don’t care about how much code you write. They only care about the product and its benefits to them.
As far as your day job goes, do as much as you can until it’s absolutely painful being there, wait a month and then consider leaving. For legal reasons, sometimes it’s not appropriate to start a company while working for someone else. You need to figure out how much is possible while still working a full-time job.
Another thing I’ve learned is to not start your interactions with a VC when you need money. You need to talk to them well in advance of needing money. Build a relationship so when the time comes that you need money, it will be a much easier pitch as they’ll know you.
What else have you learned along the way that has served you well as an entrepreneur?
Everyone says you’re supposed to have two founders. I haven’t found a person who is as passionate or who works as much as I do. I just haven’t found that person yet. I think it’s a recipe for disaster for me to partner with someone who is less enthusiastic than I am.
Also, I don’t get down when things don’t go right. For me, a challenge is more enjoyable than it is depressing. For example, we got some feedback from Indeed that our new user onboarding process was terrible. A lot of people might have been depressed by this, but I found it drove me to work harder and make the process better. Essentially, you really need to know yourself to know what the business needs, and what — and whom — you need in teammates.
What has it been like for you, as a Connecticut entrepreneur, to get this off the ground?
The challenge in Connecticut is the lack of people and density. In New York City for example, there are five other startups within a block. You have more access to programmers, sales people, VCs, etc. There’s more of everything in New York.
Yet, between The Grove and Stamford Innovation Center, we have a community here that you don’t find other places. However, I still wish there was more here: more VCs, angels, mentors, other entrepreneurs, talent, and grants. I should say that CT Next’s EIR program that allows me to talk with John Seiffer one hour per week for free has been amazing.
What is also frustrating is that there are other entrepreneurial programs in the area that we don’t have access to. Take for example the Yale Entrepreneurial Institute – it’s only for Yale students. I think it makes sense to try to open up programs like that as I think both sides could benefit.