From the ground up

When Carol Latham started Thermagon Inc., she suffered from what she calls “the five no’s”—no facility, no customers, no products, no employees and no money. What she did have was an idea for a thermally conductive material that dissipated heat from computer equipment. In eight years, she built that idea into a $5 million company with more than 40 employees. In June, Latham was named an Ernst & Young Entrepreneur Of The Year in Northeast Ohio. Small Business News asked about the challenges of building a company from scratch.

What’s the first step needed to transform an idea into a business?

Product development. I didn’t have any products to take to customers, so I bought enough simple equipment to demonstrate the properties of the materials I was trying to produce.

What did you do once you had some sample materials?

Without any travel or marketing budget, or even any local customers, we had to find other ways to get them in the right people’s hands. So we started working with technical journals. Every month we’d send a little blurb about our products. We finally got a few published.

Did those journals spark much interest?

They’re the best medium. When an engineer gets a technical journal, he thumbs through it for what’s new. The new product releases are probably the most well-read section in the journal. Those blurbs led to the first leads.

How did you follow up on those leads?

We still didn’t have any money, but we needed to get those materials into the hands of as many electronic engineers as possible. Rather than send people information, we sent them the materials to test for themselves. You can’t really sell someone a product they’ve never tested. They have to use it in their applications and then say they need it.

So the best way to build a customer base is to let them try out the materials before they buy them?

Yes. You show them it isn’t just an idea. That it’s a product with hard numbers associated with how well it performs. You let them put your product in their product, under their conditions. Then they can test it and verify its performance. But that led us to another problem. Our material was so much better that it wasn’t a direct replacement for anything already out there. So most of our original business was in new design. But new design has a long time-line associated with it.

That goes back to the money issue. While you waited for the market to develop, how did you maximize what little revenues were trickling in?

I didn’t take a salary for the first two-and-a-half years. Then, in 1992, when sales were around $70,000, I started taking a little money out. Everything coming in was going back into the product development.

1992. Was that when the market began to materialize?

Well, we had made our first few sales. We knew who our customers were going to be. But it was a challenge—getting people to accept new ideas. We hired some manufacturer’s reps in Silicon Valley. In that cultural environment, they’re open to something new and better, and they are willing to try and design a new product in. But you go to other areas and other business cultures, particularly the East Coast, and you find an entirely different reception to what you have to offer… and a lot of resistance and questioning. The real issue then becomes credibility.

So how did you build credibility?

You have to try to create an image. Once you can afford to run an ad, it gives you the image of being bigger. But without any money, you have to find other ways, such as word of mouth. Someone from one company has to talk to someone else at another company and tell them they use your products and should try it.

Is name recognition that important in order to move forward?

Yes. That’s where Intel gave us a boost. If your name comes from a design guy at a major company as a way to solve a problem, it goes a long way. What that does is give you credibility.

When was that?

In 1995. Intel got a hold of our material and tested it. They actually came to us and said, We can take your materials and put it in our notebooks. They told us to get an agent in Taiwan and recommended us to all the computer manufacturer’s in Southeast Asia. They said, if you want this Pentium to work and there’s no space for fans, Thermagon’s the way to get the heat out.

Was that Thermagon’s first big break?

Yes. But even in 1996, I wasn’t known in the Cleveland business community at all. We were known in Southeast Asia before anyone had heard of us in Cleveland. In 1996, I walked into the Hewlett-Packard offices in Singapore, sat down in the lobby and pulled out one of my brochures in anticipation of a meeting. An engineer spotted the name Thermagon on the brochure, came racing over and took me into a conference room. Name recognition.

Once you built a customer base, how have you been able to keep them?

We continually develop new products. Then you have to get them in front of a customer that can use it. It’s an ongoing process. You have to constantly address what’s going on in your industry and adapt to it.

How quickly can Thermagon adapt?

Quickly. We’ll solicit feedback from our customers and find out what they want. I travel a lot. I go to the key shows and visit the customers. They tell us what they need. Then, we’ll look at a product and determine how to improve it. We can typically make changes in about a month.

What about larger customers? Do you find that you can become dependent on a few large clients?

No. Actually, that’s something I’m sensitive to. We try to stay as diversified as possible. We’ve never had to turn away business because it was too big of a chunk of us, but a manufacturer has to make a decision about how much of their business they’re willing to devote to one customer. We put limits on it. For instance, in the automotive industry, I told them up front that I didn’t want their business to be more than 10 percent of what we do.

Why is that?

A lot of what we do is with new products. If those products don’t fly the way manufacturer’s project them to, it can have a devastating effect on everyone.

What steps do you take to ensure that doesn’t happen?

We find as many different uses in different industries for the product as we can. We’re involved in telecommunications, computing, automotive and medical-test equipment. There’s a power supply in just about every computerized device that’s made.

What other issues may affect growth?

We have to stay competitive. If you price yourself out of the market, manufacturers will find a way to design you out. They may not find another material to use, but they may come up with something that’s mechanical or another trick. You just have to figure out what the market will bear and stay with it.