Some leading companies arrived at the top without taking early venture capital funding. I sat down with Ajay Shah, co-founder of Smart Modular Technologies, to learn how he bootstrapped his startup to create a billion-dollar revenues business in just 11 years with only $80,000 of initial capital.
Headquartered in Newark, Calif, Smart Modular Technologies was founded in 1988 and designs and manufactures dynamic random access memory, computer memory modules, solid state drives, compact Flash cards, flash-based storage, and embedded storage systems for original equipment manufacturers. It sells products directly and through independent sales representatives primarily to major OEMs that compete in computing, networking, communications, printer, storage, and industrial markets. Communications companies comprise the bulk of the sales.
M&A Exit/IPO Valuation: IPO valuation of $500 million; $2 billion purchase by Solectron
M&A Exit/IPO Return Multiple: Very High/Not Meaningful Figure
I have known Ajay Shah for more than 15 years. His success runs counter to the approach of most technology companies in Silicon Valley and provides valuable insights in today’s technology market.
- Rather than rely on a technology breakthrough to create a large market, he created an innovative business model in a rapidly growing existing market: device memory systems. Investors call these businesses “execution plays” because they require great execution instead of innovative technology.
- Identifying a customer problem that his employer, Samsung, did not want to address, he jumped into the market with existing customers and support from a large corporate partner.
- Through excellent execution and continued product expansion, he grew the business to $1 billion in revenues.
Javier Rojas: How did Smart Modular get started?
Ajay Shah: In 1989, I was director of marketing for Samsung’s memory business. I noticed customers kept asking for custom memory devices. We would send them to other companies to design and manufacture the product, but they wanted a more integrated solution from a dedicated provider. Together with Samsung’s Head of Application Engineering, Mukesh Patel, I proposed a new internal business to meet the requirement, but after several trips to Korea, Samsung’s management said, “no thanks.” They saw the memory market as a high volume business poised for rapid growth and did not want to create and manage custom designed solutions within it.
So, Mukesh Patel, my wife Lata Krishnan (who left a great position at Montgomery Securities to become our CFO), and I started a company ourselves. Samsung supported us on the condition we committed to a six-month transition period. This worked great for us. We used this time to develop the business plan and put together some of the first products, while working to meet their customers’ needs. We didn’t fully prototype the products, but designed and developed them with a little bit of our own money.
JR: How long did it take you to break even?
AS: Six weeks after leaving Samsung in 1989, we had a product, two potential customers designing our product into their devices, and an agreement from Samsung to supply memory in a tight market. Eight weeks after starting we had a customer, Network Equipment Technologies, and enough revenue to break even within the first quarter without taking a salary. We also had about 10 employees.
JR: Sounds like an amazing start. How did you handle the need for a manufacturing facility? Those aren’t cheap.
AS: This was in the middle of a tech downturn, so there was a lot of excess capacity. We went from outsourcing to manufacturing the product ourselves within four months by sub-leasing real estate on terms that didn’t require us to pay until the company generated cash. Manufacturing equipment vendors had new products sitting on the shelf, so Lata negotiated with them to lend us the money to buy their new equipment and rent space to make our product. This made a good first impression when customers came to visit. It was key bootstrap creative financing that got us going; early in the process, we took $80,000 from a friend of my father-in-law to build inventory.
JR: How long did it take for revenues to ramp?
AS: In 1989, our first year, we generated $3 million in revenues and became profitable. Back then margins were between 40 and 50 percent, but that didn’t last. Memory devices became plentiful the following year, pushing Samsung to increase business with us, which drove up revenues to more than $10 million.
IBM came in shortly after. They wanted to purchase our product, but their payment terms were lousy and we didn’t have enough money to fund inventory requirements. So, when we turned them down, Big Blue offered to guarantee specific purchases, which pushed Smart Modular Technologies’ revenues to $80 million in 1992. They prepaid before Smart Modular Technologies paid the vendor. IBM became a big customer and the positive working capital cycle allowed us to extend business elsewhere.
Apple, Cisco, Hewlett-Packard, and Sun Microsystems became customers after we landed IBM. It really shows how fast things can come together after being validated by a major vendor. Cisco was very small then. Operations Manager Tom Fallon (now Infinera CEO) and Randy Pond (now Cisco EVP of Operations, Processes, and Systems), who reported to Fallon, spent time at our manufacturing facility monitoring operations. Both men spent lots of time hanging around our operations.
As Apple grew into a major company in 1993 and 1994, they became nervous with our ability to supply the company with product because we were so small. They convinced Samsung to reverse the order of the business. Rather than sell us memory so we could build the device into the product, they sold the memory with our modules directly to Apple. This meant less revenue, but we were still paid for our value-added services. As a result margins looked much healthier and we didn’t need as much cash. Revenues in this period were:
JR: That’s amazing growth. Why did you decide to purse an IPO in 1995?
AS: That’s an interesting question. As founders, we weren’t going to get liquid, at least not initially, so the primary drivers were other shareholders and customers. First, employees had now been with us for a number of years and wanted liquidity for their equity, or at least a clear path to getting it. Also, it helped with customers. Compaq and Hewlett Packard were prospective customers that wanted to see us having the financial resources and transparency that comes with being a public company. Obviously it would make us more secure, but I suspect they also wanted transparency to have a better view on what our margins were.
Finally, we also acquired a company called Apex Data to enter the data communications market. That product acquisition wasn’t successful but the team, technology, and credibility it provided allowed us to form a new division that grew to more than $250 million. The investors in that company also looked for liquidity, which also drove us toward an IPO.
The stock was rocky after the IPO. It first went very high, then collapsed. We kept focused on the business and kept growing.
JR: Sounds like you were all set with financial resources, what happened next?
AS: The next five years we grew very fast. Our memory business was strong, but we realized that we needed to get into some new areas for additional growth and to diversify. The first was communications, as I mentioned before. The Apex team and their capabilities helped us land HP in this market. We then signed Qualcomm and Toshiba. We grew from virtually zero revenue in 1994 to a $250 million business by 1999. This market was driven by the rapid growth of telecom company investments in new communication equipment.
The other big market for us was embedded systems. These were basically racks of computer boards loaded with customer software and custom chips to deliver a custom appliance. We grew this to about $30 million in revenue organically and then we acquired Compaq’s (formerly DEC’s) operations adding $100 million in revenue in 1998. This market was driven by enterprise networking requirements with Cisco being a major customer. By 1999, this was a $140 million business. Memory made up for the rest of the growth, more than doubling our growth from 1995 to 1999. Here was the annual breakdown:
JR: Why did you merge with Solectron in 1999?
AS: By this time, we had grown the business to a sizable level and would need a partner to keep expanding or enter new markets. As founders, we were also ready for some liquidity and the $2.2 billion acquisition was attractive for $120 million in earnings.
Solectron, a contract manufacturer for the electronics industry, wanted to expand their offerings services by becoming a product company. They had an embedded systems group, Force Systems, like ours with $150 million in revenue, so combined we had a $300 million division.
Prior to the merger, Smart Modular Technologies generated about $1 billion in revenue. The newly formed Solectron Technology Solutions Business grew to about $3 billion during the next two years. In that time we acquired a few other businesses.
In 2002, when new management came in, I left Solectron’s board. At the time, margins dropped and manufacturing continued its move to China. The CEM divested the solutions group in 2004. I led a group that bought it back with another partner because we liked the business and believed we could revive margins.
JR: What lessons did you learn from the experience?
AS: Most people ask if I’m glad that I didn’t take VC funds other than money from small investors. It worked out well for us, but as I look at what others accomplished in this period, I feel a partner could have helped us accomplish more. Institutional money allowed others to take more risks and professionalize sooner. Smart Modular Technologies professionalized but not in the same way or timetable. This may have lead to a bigger outcome, but we will never know.
Two key lessons I learned:
- When you see a customer problem not being addressed by your company, look to see if that might be an entrepreneurial opportunity. The opportunity that was presented to us required great execution, not technology innovation, but it was sizable just the same and well suited to our skill set. This brings me to my second point.
- To pursue an execution-focused business it helps to have some large company management skills—they certainly favored us—and you need to execute flawlessly while you are growing rapidly. That’s not easy to do but it was what led customers to grow their business with us, providing strong references to others.
Javier Rojas is Managing Director of Kennet Partners. For more articles on entrepreneurship visit Javier’s blog at seekinggrowth.typepad.com.