NCA Microelectronics

1990 had not been a great year for NCA Microelectronics. Some good products had been developed, but financial success still seemed as far away as ever. Losses had forced severe staff cutbacks. The management group met in a planning session, in 1991, knowing that they had only one more chance to get it right. Harvey Nickerson looked at his partner Lewis Cobb and said, 'If we don't avoid the problems we had with the R1000 and R2000, then the Chameleon project will 10 NCA Microelectronics. We'll lose the company and Saint John will have one less high-tech startup." "Relax," said Lewis, "we know more about technology marketing since 1990, and anyway, we don't have a federal lab to complicate the development."

The Company

NCA Microelectronics was formed in 1987 by Harvey Nickerson and Lewis Cobb, two young electrical engineering graduates of the University of New Brunswick. Each had become disenchanted with the idea of being a "salaryman" in a large organisation and felt excited by the challenge of creating a successful electronics design company in their home province of New Brunswick.

This case was prepared by Professor Alan McLean of the University of New Brunswick for the Atlantic Entrepreneurial Institute as a basis for classroom discussion, and is not meant to illustrate either effective or ineffective management.

Copyright 1992, the Atlantic Entrepreneurial Institute. Reproduction of this case is allowed without permission for educational purposes, but all such reproduction must acknowledge the copyright. This permission does not include publication.

The first step in raising investment capital to augment their own resources was the preparation of a business plan. As a guide, they used the handbook Making Technology Happen by the well-known technological entrepreneur Denny Doyle. Denny, among other achievements, founded the Control Data Corporation operation in Canada. Eventually their plan and their personal enthusiasm attracted the interest of Bill Stanley of Cable Management, a company which controlled a group of companies involved in the cable TV industry. Stanley saw some potential synergy between his companies and a young high-tech microelectronics company. In January 1987, after an exchange of shares for equity, NCA Microelectronics was incorporated and set up operations in Saint John to provide high-tech electronic engineering services to the rest of the world. There were few, if any, other electronics design companies in New Brunswick.

The company's mission statement was "to generate customer satisfaction and corporate profitability through the design and manufacture of quality electronic products."

Initially, the company's only expertise was that of Harvey and Lewis and they chased energetically after any work which was within their technical competence. This strategy was reasonably successful, and by 1989, NCA had a staff of 15 and an annual revenue well in excess of $1,000,000.

Harvey Nickerson and Lewis Cobb

Both Harvey and Lewis were native New Brunswickers and like most Atlantic Canadians, had a strong attachment to their home province. After graduating with his masters degree in electrical engineering in 1982, Harvey worked as a research engineer at the newly formed CADMI Microelectronics Inc, a technology development facility on the campus of the University of New Brunswick (UNB). Subsequently, he took a position as a design engineer with Process Technologies Limited in Oromocto, New Brunswick. Concern with the uncertain future of this company contributed to his desire to establish his own enterprise.

Lewis who graduated with his master's in electrical engineering in the fall of 1983, worked for a year with MacDonald, Detweiler and Associates in Richmond, British Columbia, on satellite image processing. He returned to New Brunswick, also to work with CADMI, and renewed his acquaintance with Harvey. They shared their desires to own and operate a business, and decided to branch out together.

Products and Services of NCA

Initially, the work Which Harvey and Lewis brought in was vaned. It included consulting, contract research, conceptual design, equipment development and modifications. Clients came from the private sector and government, including the military. Most projects involved some aspect of control systems, gradually the company began to carve out a niche in the design and development of microprocessor-based monitoring and control systems

NCA was able to secure several contracts from Saint John Shipbuilding which was engaged in the construction of a series of ultra-modem frigates for the Canadian Armed Forces. This work led NCA straight into the issue of quality assurance. The very demanding military standards forced them to examine existing quality control systems very critically and the company was obliged to become qualified to the Z2993 standard of the Canadian Standards Association. This attention to quality, originally driven by the military contracts, helped the company build an excellent reputation with private sector customers.

A typical project for NCA was the Chat-R-Box developed under subcontract for Edmonton Transit. This product was originally designed to replace the endless-loop taped messages used in the transit industry to announce stops and other relevant information. Chat-R-Box could store up to four minutes of digitised as opposed to synthesised speech, accessible as a number of small phrases or as continuous audio text. Chat-R-Box could be used for anything from talking signs and traffic signals to talking garbage cans.

The company had electronics manufacturing facilities suitable for custom construction or for small runs of product. Other manufacturing requirements such as housing fabrication could be contracted out.

The R1000 Story

In the spring of 1988, NCA won a $10,000 contract from the Department of Communications (DOC) to replicate a prototype of the Interact system. DOC was mandated to monitor and prevent illegal use of the airwaves. Continuous, routine monitoring of radio traffic, especially short range traffic in remote locations, was expensive and impractical. The Interact system, then used in prototype form by DOC, had been designed by a DOC engineer, Mike LeBlanc, also a UNB graduate, to help with this problem. The unit could be transported to location and could be interrogated remotely using ordinary telephone lines. DOC was interested in having a second copy of their prototype made and was also prepared to transfer the technology to any company who was interested in and capable of marketing it commercially.

This seemed a perfect opportunity to Harvey and Lewis and they entered enthusiastically into discussions with DOC personnel, principally Marty MacLellan and Mike LeBlanc of the Moncton Office of DOC. The DOC technology transfer expert from Ottawa was brought in and, according to Marty, "the project seemed do-able." Negotiations, which were never actually completed, began on the terms of the technology licensing agreement. The Interact system, although effective, was IBM-PC based and bulky. The NCA team saw two tasks before them: further development of the technology to make the system more acceptable to the market place and development of a marketing strategy.

A $100,000 contribution was made to the technology transfer process from the National Research Council, under the Industrial Research Assistance Program, (IRAP) and NCA budgeted a further $100,000 from their own resources. Additional staff were hued and the technology development team began work, in April 1989, on miniaturising the system and making it more robust. The process was planned to take twelve months. It was decided that the final product would be named the R1000.

Harvey and Lewis, realising that their marketing expertise lay well short of their technical expertise, enlisted the help of a marketing expert from Fundy Cable, a company controlled by Cable Management, to help them assess the market prospects for the product.

The Interact system had been the subject of a paper presented by DOC personnel at an international communications conference in Europe and the technology had attracted a lot of interest. Also, the Ottawa office of DOC was actively involved in the provision of technical assistance to developing countries who wished to improve their ability to regulate and control their airwaves. The Interact/R1000 technology was demonstrated at a trade show to participants in a training program and was received enthusiastically. Influenced by the DOC's enthusiasm the marketing expert from Fundy Cable became excited by the prospects not only for North American sales but also for overseas sales.

A marketing strategy was developed which involved the appointment of Canadian and US distributors. A demonstration unit was shipped to a distributor in the Far East and was exhibited at a trade fair in Singapore.

A key to the overseas sales was the purchase by DOC of a unit for their own use. This would then become a reference point for potential overseas customers. Unfortunately, as the technical modifications developed, a difference developed between NCA and DOC over what features the final version of the R1000 should have. The NCA project team felt that the features desired by the DOC could not be achieved within the development budget and would make the final product unnecessarily expensive. This would detract from the marketability of the product. According to Harvey, "they wanted a Cadillac but would only pay for a Ford." Eventually, a compromise was reached and NCA agreed to develop an R900 version which could be upgraded at a later date to the "Cadillac"i.e., the fun-feature R1000. The main marketing effort would be directed to promoting the R900 with its lower cost.

The main competition for the R900 seemed to come from the German company, Rohde and Schwartz, whose product sold for about $1,000,000, well over ten times the projected cost of the R900. This product had many more features than the R900, although it could not be operated remotely.

The marketing prospects still looked good for the R900, not only to NCA, but to a second marketing consultant hired to evaluate the situation. Quotes were requested by the governments of India and Uganda and visits to the plant were made by prospective buyers.

Unfortunately, by the spring of 1990, the DOC began to feel the brunt of budgetary restraint and informed NCA that they would no longer be in a position to purchase any R900's. It also emerged that the regional office of DOC, with whom NCA had been dealing, was the only one interested in the R900. Other regions appeared to have different ideas as to how to tackle the problem of remotely controlled radio spectrum monitoring. "It seemed that the NIH (Not Invented Here) syndrome was working to the detriment of NCA," claimed Harvey.

Realising immediately that sales to foreign vernments would be seriously Jeopardized if the Canadian government did not purchase the R900 or the R1000, Harvey and Lewis reviewed their options. They could either drop the project completely, develop other markets for the R900, or try to find other applications for the core technology. Because they had faith in the technology, they decided to pursue the second and third options.

In their efforts to develop other markets for the R900, NCA contacted several agencies and government departments. 'We realized that, while DOC and similar government departments overseas are interested in regulating the use of airwaves, law enforcement agencies may be interested in what is being said on the air," sad Harvey. "We approached the RCMP and the DEA in the US, thinking that our equipment could be used to gain information on drug smuggling. Each organisation seemed interested and we eventually lent a unit to the RCMP for a counter-smuggling operation in Newfoundland. Although the use of our equipment led to arrests, or so we understand, it unfortunately did not lead to a sale. Our experience with the DEA was also frustrating. The key to a successful sale of sophisticated technology depends very much on personal contact. We would laboriously build a relationship with someone and then they would 'disappear'. They would even refuse to acknowledge that our former contact had ever existed. I suppose this is the nature of law enforcement agencies, particularly those engaged in the war against drugs, however they are extremely hard for a small company to do business with. "Perhaps they were concerned about our security," said Harvey ruefully.

With respect to finding another application for the unique aspects of the R900, Harvey and Lewis agreed that the most important feature to pursue was the ability to use ordinary telephone lines to carry out interrogations of remote facilities.

The transmission towers used in the broadcasting industry are usually situated on hilltops, often in remote locations. Since the facilities are automatic, maintenance by direct inspection is inconvenient and expensive. The NCA staff concluded that this would be a natural opportunity to apply their expertise in remote monitoring. The potential market was studied by examining competitors' products and by talking to equipment suppliers and distributors. There seemed to be an opportunity and a development plan was struck for a new product. Since about 50 percent of the R900/1000 technology was transferable, it was decided to call the new product the R2000. Work began on the development of the R2000 before the R900 project was finally terminated.

By September 1990, the R2000 was ready The system was user-friendly, modular and therefore expanded easily, and programmable by the user, meaning it could be customized for a wide variety of situations. A marketing manager had been hired and the marketing campaign, Which had been carried out in parallel with the technology development, seemed successful. There had been good response from trade shows and from customers who had agreed to test preliminary versions (beta products). The team had also created a telephone demonstration. A potential customer could phone, using an 800 number, a simulated transmission tower and experiment with the interrogation features of the R2000. This was extremely popular and the phone hardly stopped ringing About 12 units were sold, at about $10,000 per unit.

Unfortunately, sales stopped abruptly. Radio stations, hit by a sudden drop in advertising revenues brought about by the recession, drastically cut back on their capital spending. Although there was still a certain amount of business from the defence industry, the NCA had tied its future to the success of the R2000. The negative cash flow brought about by the nonperformance of the R900 and now the collapse of the market for the R2000, was unsupportable. By November 1990, the company had no alternative but reduce the staff down to three: Harvey, Lewis and Mark Sanford, a production engineer who had been one of the first employees. The drastic decline in staff meant that the company could no longer provide technical support to the R2000, and the product had to be taken off the market.

The Chameleon Project

Because of their association with Fundy Cable, Lewis and Harvey had already considered developing products for the cable TV industry. The cable TV industry was however, not technology conscious but very cost conscious, and did not seem a fruitful market for high-technology products. The new reality, brought about by the collapse in demand for the "R" series of products, forced a second look at the situation and a niche seemed to emerge.

The theft of Pay-TV signal was a growing problem for cable television companies, especially in the US. In some regions, the theft rate was estimated to be as high as 50 percent. The theft was accomplished by bypassing or otherwise defeating the signal scrambler supplied by the TV company. The construction and sale of bootleg signal descramblers was a thriving cottage industry.

Sophisticated scramblers we're available at a cost of about $150 per subscriber, however small-to-medium cable companies with a large investment in older, more easily defeated, scramblers could not afford this investment. NCA reckoned that the solution to the problem these companies faced was their new product, the Chameleon. This was a circuit board which would fit made existing scramblers and would sell for under $50. A feature of the board was a special chip which contained about 100 scrambling algorithms which used a random number generator. The great advantage was that the scrambling algorithm could be changed by a signal from the -head end,- i.e. from equipment operated by the cable company itself.

By far, the most common of the older style scramblers was made by General Instruments and the Chameleon was designed to be easily fitted into this equipment. Millions of these units were in service in North America and would not be completely replaced until 1997. Even then, the equipment would be in use in South America for many more years which would further extend the market window.

NCA expected to sell about 160,000 units in the first year of full production of which 35,000 would be sold in the Maritimes. At an estimated wholesale price of $50, the first year's gross revenue would be $8.0 million. Future revenues would depend on the degree of difficulty of penetrating the various sub-segments of the market, but the potential existed for sales of as many as half a million units per year for the following four years.

The Planning Session

"You're right, Lewis," said Harvey, "the situation is simpler and we have learned a lot, but selling to the cable TV industry is different again. These guys are not from an R and D culture. Everything depends on proof-of-payback and =proved service is not something they stay up late worrying about."

"OK, OK, but we can show them proof-of-payback! That's not the problem! The problem is, how do we sell them a $50 item? We are used to thinking in terms of marketing relatively small numbers of highly specialised products. The Chamelon is virtually a consumer item with large volumes and low markups."

"And speaking of low markups," interrupted Mark Sanford, "how will we handle the manufacture of this little goody? We will have to decide whether to set up our own line or contract out the manufacture once the design is frozen. If we make it ourselves, we can control the quality better, but we will get stuck with setup and inventory costs. If we contract it out, it will be easier on our capital, but we will have to take a smaller markup."

By this time, Harvey was staring out the window over the empty parking lot. "We must figure out the best way of marketing the little suckers. If we've learned anything, we've learned how difficult it is to market technically complex systems. The professional marketers don't understand the technology and we don't know enough about the science of marketing. The cable companies are interested, at least according to their reaction at the trade shows, but who will make the decision or push our product for us? We have to work through distributors. We cant market directly. I was talking to the head of Fundy Cable's repair section about it. She reckons that most small cable TV companies in the US contract out their repair and installation work. She reckons we should market to the repair companies through their distributors. They are the ones with the knowhow, not the cable companies themselves.

"Well, Louise knows how many beans make five alright, but it's the cable companies' money we are trying to protect," Lewis pointed out. 'Surely they have the motivation to tell the repairers to install Chamelon? Another problem we have to face is whether we should be betting the company on one product again."

'Wait a minute!" cried Harvey, turning back from the window, "it's not our only product. We still have Chat-R-Box and how about the work we are doing for Thomas Equipment? If we can get the bugs out of the onboard computer we are developing for their skid loader, things could work out well."

Lewis grudgingly agreed, "...but there is a danger in developing a product line for only one client. We are too small to play hardball with them."

"We won't be small for Iong," grinned Harvey. "However, we have a number of issues to straighten out before we get rich. We need to develop a marketing game plan for the Chameleon and as Mark pointed out, we need to decide whether to make or buy. I guess we should really look at other work we want to go after. It's not smart to bet the company, but there are only the three of us and we don't have the money to invest in all the opportunities...I would dearly love to be able to get back to the R2000 game though..." "Yes," sighed Lewis. "It was great technology- Pity nobody wanted to buy it."